diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000000000000000000000000000000000000..52373fe24473b1aa44333d318f578ae6bf04b49b --- /dev/null +++ b/.gitattributes @@ -0,0 +1,36 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.ftz filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.h5 filter=lfs diff=lfs merge=lfs -text +*.joblib filter=lfs diff=lfs merge=lfs -text +*.lfs.* filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.onnx filter=lfs diff=lfs merge=lfs -text +*.ot filter=lfs diff=lfs merge=lfs -text +*.parquet filter=lfs diff=lfs merge=lfs -text +*.pb filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +saved_model/**/* filter=lfs diff=lfs merge=lfs -text +*.tar.* filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.tflite filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text +tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/args.json b/args.json new file mode 100644 index 0000000000000000000000000000000000000000..fc81db4043b465559f54e9567542b1da599d0c98 --- /dev/null +++ b/args.json @@ -0,0 +1,393 @@ +{ + "output_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_planA_wm_joint-20260531-113013/v0-20260531-113038", + "per_device_train_batch_size": 1, + "num_train_epochs": 1.0, + "max_steps": -1, + "learning_rate": 1e-05, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_steps": 200.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "optim_target_modules": null, + "gradient_accumulation_steps": 16, + "average_tokens_across_devices": true, + "max_grad_norm": 1.0, + "label_smoothing_factor": 0.0, + "bf16": true, + "fp16": false, + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": true, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "use_liger_kernel": false, + "liger_kernel_config": null, + "use_cache": false, + "neftune_noise_alpha": null, + "torch_empty_cache_steps": null, + "auto_find_batch_size": false, + "logging_strategy": "steps", + "logging_steps": 10, + "logging_first_step": true, + "log_on_each_node": true, + "logging_nan_inf_filter": true, + "include_num_input_tokens_seen": false, + "log_level": "passive", + "log_level_replica": "warning", + "disable_tqdm": null, + "report_to": [ + "wandb" + ], + "run_name": "qwen35_4b_planA_wm_joint-20260531-113013", + "project": "huggingface", + "trackio_space_id": null, + "trackio_bucket_id": null, + "trackio_static_space_id": null, + "eval_strategy": "steps", + "eval_steps": 500.0, + "eval_delay": 0, + "per_device_eval_batch_size": 1, + "prediction_loss_only": false, + "eval_on_start": false, + "eval_do_concat_batches": true, + "eval_use_gather_object": false, + "eval_accumulation_steps": null, + "include_for_metrics": [], + "batch_eval_metrics": false, + "save_only_model": false, + "save_strategy": "steps", + "save_steps": 100.0, + "save_on_each_node": false, + "save_total_limit": 3, + "enable_jit_checkpoint": false, + "push_to_hub": false, + "hub_token": null, + "hub_private_repo": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_always_push": false, + "hub_revision": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "restore_callback_states_from_checkpoint": false, + "full_determinism": false, + "seed": 42, + "data_seed": 42, + "use_cpu": false, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "dataloader_drop_last": false, + "dataloader_num_workers": 8, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 4, + "remove_unused_columns": true, + "label_names": null, + "train_sampling_strategy": "random", + "length_column_name": "length", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "ddp_static_graph": null, + "ddp_backend": null, + "ddp_timeout": 18000000, + "fsdp": [], + "fsdp_config": null, + "deepspeed": { + "bf16": { + "enabled": true + }, + "fp16": { + "enabled": false + }, + "zero_optimization": { + "stage": 2, + "overlap_comm": true, + "contiguous_gradients": true, + "reduce_bucket_size": 500000000.0, + "allgather_bucket_size": 500000000.0, + "reduce_scatter": true, + "round_robin_gradients": true + }, + "gradient_clipping": "auto", + "gradient_accumulation_steps": "auto", + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "steps_per_print": 100, + "wall_clock_breakdown": false, + "_comment_bf16_safety": "DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives." + }, + "debug": null, + "skip_memory_metrics": true, + "do_train": false, + "do_eval": false, + "do_predict": false, + "resume_from_checkpoint": null, + "warmup_ratio": null, + "logging_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_planA_wm_joint-20260531-113013/v0-20260531-113038/runs", + "local_rank": 0, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "group_by_length": false, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", + "model_type": "qwen3_5", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attention_2", + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_5", + "system": null, + "max_length": 32768, + "truncation_strategy": "delete", + "max_pixels": 16777216, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": true, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/ubuntu/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/ubuntu/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l1/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l2/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l3/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/OS-Atlas/linux/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/OS-Atlas/macos/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/OS-Atlas/windows/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/ShowUI/desktop/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/ShowUI-web/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GroundCUA/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/ScreenSpot/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/UGround/web_direct_150k_description_filtered/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/UGround/web_direct_258k_function_filtered/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/UGround/web_hybrid_773k_max_25qa_filtered/train.jsonl#80000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/Multimodal-Mind2Web/train.jsonl" + ], + "val_dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/val_wm.jsonl" + ], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 4, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": true, + "packing_length": 32768, + "packing_num_proc": 1, + "lazy_tokenize": false, + "use_hf": false, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual", + "model.visual.merger" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3.5-4B", + "model_info": "ModelInfo(model_type='qwen3_5', model_dir='/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling=None, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_5', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.5-0.8B', hf_model_id='Qwen/Qwen3.5-0.8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B', hf_model_id='Qwen/Qwen3.5-2B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B', hf_model_id='Qwen/Qwen3.5-4B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B', hf_model_id='Qwen/Qwen3.5-9B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B', hf_model_id='Qwen/Qwen3.5-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B-FP8', hf_model_id='Qwen/Qwen3.5-27B-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-0.8B-Base', hf_model_id='Qwen/Qwen3.5-0.8B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B-Base', hf_model_id='Qwen/Qwen3.5-2B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B-Base', hf_model_id='Qwen/Qwen3.5-4B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B-Base', hf_model_id='Qwen/Qwen3.5-9B-Base', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.6-27B', hf_model_id='Qwen/Qwen3.6-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.6-27B-FP8', hf_model_id='Qwen/Qwen3.6-27B-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[])], loader=, template='qwen3_5', model_arch=MultiModelKeys(arch_name='qwen2_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3_5ForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=5.0.0.dev', 'qwen_vl_utils>=0.0.14', 'decord'], tags=[])", + "model_dir": "/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", + "template_meta": "QwenTemplateMeta(template_type='qwen3_5', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='qwen3_5', is_thinking=True, thinking_prefix='\\n', non_thinking_prefix='\\n\\n\\n\\n', history_thinking_prefix='')", + "_val_dataset_exists": true, + "hub": "", + "evaluation_strategy": "steps", + "training_args": "Seq2SeqTrainingArguments(output_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_planA_wm_joint-20260531-113013/v0-20260531-113038', per_device_train_batch_size=1, num_train_epochs=1.0, max_steps=-1, learning_rate=1e-05, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_steps=200.0, optim=, optim_args=None, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=16, average_tokens_across_devices=None, max_grad_norm=1.0, label_smoothing_factor=0.0, bf16=True, fp16=False, bf16_full_eval=False, fp16_full_eval=False, tf32=True, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, use_liger_kernel=False, liger_kernel_config=None, use_cache=False, neftune_noise_alpha=None, torch_empty_cache_steps=None, auto_find_batch_size=False, logging_strategy=, logging_steps=10, logging_first_step=True, log_on_each_node=True, logging_nan_inf_filter=True, include_num_input_tokens_seen=None, log_level='passive', log_level_replica='warning', disable_tqdm=False, report_to=['wandb'], run_name='qwen35_4b_planA_wm_joint-20260531-113013', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=, eval_steps=500, eval_delay=0, per_device_eval_batch_size=1, prediction_loss_only=False, eval_on_start=False, eval_do_concat_batches=True, eval_use_gather_object=False, eval_accumulation_steps=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=, save_steps=100, save_on_each_node=False, save_total_limit=3, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=, hub_always_push=False, hub_revision=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, restore_callback_states_from_checkpoint=False, full_determinism=False, seed=42, data_seed=42, use_cpu=False, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, dataloader_drop_last=False, dataloader_num_workers=8, dataloader_pin_memory=True, dataloader_persistent_workers=True, dataloader_prefetch_factor=4, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, deepspeed={'bf16': {'enabled': True}, 'fp16': {'enabled': False}, 'zero_optimization': {'stage': 2, 'overlap_comm': True, 'contiguous_gradients': True, 'reduce_bucket_size': 500000000.0, 'allgather_bucket_size': 500000000.0, 'reduce_scatter': True, 'round_robin_gradients': True}, 'gradient_clipping': 'auto', 'gradient_accumulation_steps': 'auto', 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'steps_per_print': 100, 'wall_clock_breakdown': False, '_comment_bf16_safety': 'DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives.'}, debug=[], skip_memory_metrics=True, do_train=False, do_eval=True, do_predict=False, resume_from_checkpoint=None, warmup_ratio=None, logging_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_planA_wm_joint-20260531-113013/v0-20260531-113038/runs', local_rank=0, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, group_by_length=False, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/chat_template.jinja b/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..a585dec894e63da457d9440ec6aa7caa16d20860 --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,154 @@ +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- macro render_content(content, do_vision_count, is_system_content=false) %} + {%- if content is string %} + {{- content }} + {%- elif content is iterable and content is not mapping %} + {%- for item in content %} + {%- if 'image' in item or 'image_url' in item or item.type == 'image' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain images.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set image_count.value = image_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Picture ' ~ image_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|image_pad|><|vision_end|>' }} + {%- elif 'video' in item or item.type == 'video' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain videos.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set video_count.value = video_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Video ' ~ video_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|video_pad|><|vision_end|>' }} + {%- elif 'text' in item %} + {{- item.text }} + {%- else %} + {{- raise_exception('Unexpected item type in content.') }} + {%- endif %} + {%- endfor %} + {%- elif content is none or content is undefined %} + {{- '' }} + {%- else %} + {{- raise_exception('Unexpected content type.') }} + {%- endif %} +{%- endmacro %} +{%- if not messages %} + {{- raise_exception('No messages provided.') }} +{%- endif %} +{%- if tools and tools is iterable and tools is not mapping %} + {{- '<|im_start|>system\n' }} + {{- "# Tools\n\nYou have access to the following functions:\n\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n" }} + {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {%- if content %} + {{- '\n\n' + content }} + {%- endif %} + {%- endif %} + {{- '<|im_end|>\n' }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" %} + {%- set content = render_content(message.content, false)|trim %} + {%- if not(content.startswith('') and content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if ns.multi_step_tool %} + {{- raise_exception('No user query found in messages.') }} +{%- endif %} +{%- for message in messages %} + {%- set content = render_content(message.content, true)|trim %} + {%- if message.role == "system" %} + {%- if not loop.first %} + {{- raise_exception('System message must be at the beginning.') }} + {%- endif %} + {%- elif message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = '' %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- set reasoning_content = reasoning_content|trim %} + {%- if loop.index0 > ns.last_query_index %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {%- if loop.first %} + {%- if content|trim %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n\n' }} + {%- endif %} + {%- else %} + {{- '\n\n\n' }} + {%- endif %} + {%- if tool_call.arguments is defined %} + {%- for args_name, args_value in tool_call.arguments|items %} + {{- '\n' }} + {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %} + {{- args_value }} + {{- '\n\n' }} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.previtem and loop.previtem.role != "tool" %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if not loop.last and loop.nextitem.role != "tool" %} + {{- '<|im_end|>\n' }} + {%- elif loop.last %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- else %} + {{- raise_exception('Unexpected message role.') }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n' }} + {%- endif %} +{%- endif %} \ No newline at end of file diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b1ade33543aaa3d299835283e0e20968262a8a0e --- /dev/null +++ b/config.json @@ -0,0 +1,113 @@ +{ + "architectures": [ + "Qwen3_5ForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 248046, + "hidden_size": 2560, + "image_token_id": 248056, + "model_type": "qwen3_5", + "pad_token_id": 248044, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "attn_output_gate": true, + "bos_token_id": null, + "dtype": "bfloat16", + "eos_token_id": 248044, + "full_attention_interval": 4, + "head_dim": 256, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9216, + "layer_types": [ + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention" + ], + "linear_conv_kernel_dim": 4, + "linear_key_head_dim": 128, + "linear_num_key_heads": 16, + "linear_num_value_heads": 32, + "linear_value_head_dim": 128, + "mamba_ssm_dtype": "float32", + "max_position_embeddings": 262144, + "mlp_only_layers": [], + "model_type": "qwen3_5_text", + "mtp_num_hidden_layers": 1, + "mtp_use_dedicated_embeddings": false, + "num_attention_heads": 16, + "num_hidden_layers": 32, + "num_key_value_heads": 4, + "pad_token_id": 248044, + "partial_rotary_factor": 0.25, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "mrope_interleaved": true, + "mrope_section": [ + 11, + 11, + 10 + ], + "partial_rotary_factor": 0.25, + "rope_theta": 10000000, + "rope_type": "default" + }, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 248320 + }, + "tie_word_embeddings": true, + "transformers_version": "5.8.1", + "use_cache": false, + "video_token_id": 248057, + "vision_config": { + "deepstack_visual_indexes": [], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_5_vision", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 248054, + "vision_start_token_id": 248053 +} diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1f1ab970fed81cb788c2945bed33833d247b9091 --- /dev/null +++ b/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "eos_token_id": [ + 248044, + 248046 + ], + "transformers_version": "5.8.1", + "use_cache": true +} diff --git a/model-00001-of-00003.safetensors b/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8c7b5c60e2b4cf5b34a0fab1ecc834619ba724ab --- /dev/null +++ b/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eafbddf48dd2889539f340d150076297ea5bdfec759a0e0099b77d35da20881d +size 4958192864 diff --git a/model-00002-of-00003.safetensors b/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c2361ea355caeb95b494555747c52e2dac347c97 --- /dev/null +++ b/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8cfaf74f186f8396316694f72d60c62529bce98ad4e5bcf433a7a69fa980d270 +size 4993498424 diff --git a/model-00003-of-00003.safetensors b/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ab8023a740649342f1bc7e224320898de73054c --- /dev/null +++ b/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9132529e66cac3079a828b70ff720888e735b62dcc8a839f7c42730ef896d42c +size 398327248 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..17ab266552b397e797c55ed9da38d7b9120febd8 --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,732 @@ +{ + "metadata": { + "total_parameters": 4553957376, + "total_size": 10350018536 + }, + "weight_map": { + "lm_head.weight": "model-00001-of-00003.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00003.safetensors" + } +} \ No newline at end of file diff --git a/preprocessor_config.json b/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/processor_config.json b/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..33818c7f9e991ad735fd240209f4fa73e6c28c50 --- /dev/null +++ b/processor_config.json @@ -0,0 +1,60 @@ +{ + "image_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/args.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/args.json new file mode 100644 index 0000000000000000000000000000000000000000..7a2622c7b2495916780dca7e70197446cf7797ba --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/args.json @@ -0,0 +1,382 @@ +{ + "output_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr/v3-20260618-000502", + "per_device_train_batch_size": 1, + "num_train_epochs": 1.0, + "max_steps": -1, + "learning_rate": 1e-05, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_steps": 200.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "optim_target_modules": null, + "gradient_accumulation_steps": 16, + "average_tokens_across_devices": true, + "max_grad_norm": 1.0, + "label_smoothing_factor": 0.0, + "bf16": true, + "fp16": false, + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": true, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "use_liger_kernel": false, + "liger_kernel_config": null, + "use_cache": false, + "neftune_noise_alpha": null, + "torch_empty_cache_steps": null, + "auto_find_batch_size": false, + "logging_strategy": "steps", + "logging_steps": 1, + "logging_first_step": true, + "log_on_each_node": true, + "logging_nan_inf_filter": true, + "include_num_input_tokens_seen": false, + "log_level": "passive", + "log_level_replica": "warning", + "disable_tqdm": null, + "report_to": [ + "wandb" + ], + "run_name": "wm_pool32x32_phase2_baseline_wmlr", + "project": "huggingface", + "trackio_space_id": null, + "trackio_bucket_id": null, + "trackio_static_space_id": null, + "eval_strategy": "steps", + "eval_steps": 100.0, + "eval_delay": 0, + "per_device_eval_batch_size": 1, + "prediction_loss_only": false, + "eval_on_start": false, + "eval_do_concat_batches": true, + "eval_use_gather_object": false, + "eval_accumulation_steps": null, + "include_for_metrics": [], + "batch_eval_metrics": false, + "save_only_model": false, + "save_strategy": "steps", + "save_steps": 100.0, + "save_on_each_node": false, + "save_total_limit": 3, + "enable_jit_checkpoint": false, + "push_to_hub": false, + "hub_token": null, + "hub_private_repo": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_always_push": false, + "hub_revision": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "restore_callback_states_from_checkpoint": false, + "full_determinism": false, + "seed": 42, + "data_seed": 42, + "use_cpu": false, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "dataloader_drop_last": false, + "dataloader_num_workers": 8, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 4, + "remove_unused_columns": true, + "label_names": null, + "train_sampling_strategy": "random", + "length_column_name": "length", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "ddp_static_graph": null, + "ddp_backend": null, + "ddp_timeout": 18000000, + "fsdp": [], + "fsdp_config": null, + "deepspeed": { + "bf16": { + "enabled": true + }, + "fp16": { + "enabled": false + }, + "zero_optimization": { + "stage": 2, + "overlap_comm": true, + "contiguous_gradients": true, + "reduce_bucket_size": 500000000.0, + "allgather_bucket_size": 500000000.0, + "reduce_scatter": true, + "round_robin_gradients": true + }, + "gradient_clipping": "auto", + "gradient_accumulation_steps": "auto", + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "steps_per_print": 100, + "wall_clock_breakdown": false, + "_comment_bf16_safety": "DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives." + }, + "debug": null, + "skip_memory_metrics": true, + "do_train": false, + "do_eval": false, + "do_predict": false, + "resume_from_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr/v2-20260617-102539/checkpoint-3800", + "warmup_ratio": null, + "logging_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr/v3-20260618-000502/runs", + "local_rank": 0, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "group_by_length": false, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "model_type": "qwen3_5", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attention_2", + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_5", + "system": null, + "max_length": 12288, + "truncation_strategy": "delete", + "max_pixels": 16777216, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": false, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l1/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l2/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l3/train_wm.jsonl#50000" + ], + "val_dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/val_wm.jsonl" + ], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 4, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "use_hf": false, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual", + "model.visual.merger" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "checkpoint-4336", + "model_info": "ModelInfo(model_type='qwen3_5', model_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling=None, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_5', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.5-0.8B', hf_model_id='Qwen/Qwen3.5-0.8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B', hf_model_id='Qwen/Qwen3.5-2B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B', hf_model_id='Qwen/Qwen3.5-4B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B', hf_model_id='Qwen/Qwen3.5-9B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B', hf_model_id='Qwen/Qwen3.5-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B-FP8', hf_model_id='Qwen/Qwen3.5-27B-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-0.8B-Base', hf_model_id='Qwen/Qwen3.5-0.8B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B-Base', hf_model_id='Qwen/Qwen3.5-2B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B-Base', hf_model_id='Qwen/Qwen3.5-4B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B-Base', hf_model_id='Qwen/Qwen3.5-9B-Base', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.6-27B', hf_model_id='Qwen/Qwen3.6-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.6-27B-FP8', hf_model_id='Qwen/Qwen3.6-27B-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[])], loader=, template=None, model_arch=MultiModelKeys(arch_name='qwen2_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3_5ForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=5.0.0.dev', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "template_meta": "QwenTemplateMeta(template_type='qwen3_5', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='qwen3_5', is_thinking=True, thinking_prefix='\\n', non_thinking_prefix='\\n\\n\\n\\n', history_thinking_prefix='')", + "_val_dataset_exists": true, + "hub": "", + "evaluation_strategy": "steps", + "training_args": "Seq2SeqTrainingArguments(output_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr/v3-20260618-000502', per_device_train_batch_size=1, num_train_epochs=1.0, max_steps=-1, learning_rate=1e-05, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_steps=200.0, optim=, optim_args=None, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=16, average_tokens_across_devices=None, max_grad_norm=1.0, label_smoothing_factor=0.0, bf16=True, fp16=False, bf16_full_eval=False, fp16_full_eval=False, tf32=True, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, use_liger_kernel=False, liger_kernel_config=None, use_cache=False, neftune_noise_alpha=None, torch_empty_cache_steps=None, auto_find_batch_size=False, logging_strategy=, logging_steps=1, logging_first_step=True, log_on_each_node=True, logging_nan_inf_filter=True, include_num_input_tokens_seen=None, log_level='passive', log_level_replica='warning', disable_tqdm=False, report_to=['wandb'], run_name='wm_pool32x32_phase2_baseline_wmlr', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=, eval_steps=100, eval_delay=0, per_device_eval_batch_size=1, prediction_loss_only=False, eval_on_start=False, eval_do_concat_batches=True, eval_use_gather_object=False, eval_accumulation_steps=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=, save_steps=100, save_on_each_node=False, save_total_limit=3, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=, hub_always_push=False, hub_revision=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, restore_callback_states_from_checkpoint=False, full_determinism=False, seed=42, data_seed=42, use_cpu=False, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, dataloader_drop_last=False, dataloader_num_workers=8, dataloader_pin_memory=True, dataloader_persistent_workers=True, dataloader_prefetch_factor=4, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, deepspeed={'bf16': {'enabled': True}, 'fp16': {'enabled': False}, 'zero_optimization': {'stage': 2, 'overlap_comm': True, 'contiguous_gradients': True, 'reduce_bucket_size': 500000000.0, 'allgather_bucket_size': 500000000.0, 'reduce_scatter': True, 'round_robin_gradients': True}, 'gradient_clipping': 'auto', 'gradient_accumulation_steps': 'auto', 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'steps_per_print': 100, 'wall_clock_breakdown': False, '_comment_bf16_safety': 'DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives.'}, debug=[], skip_memory_metrics=True, do_train=False, do_eval=True, do_predict=False, resume_from_checkpoint='/NHNHOME/data/sanghyeok/qwen-cua/runs/opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr/v2-20260617-102539/checkpoint-3800', warmup_ratio=None, logging_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr/v3-20260618-000502/runs', local_rank=0, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, group_by_length=False, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/chat_template.jinja b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..a585dec894e63da457d9440ec6aa7caa16d20860 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/chat_template.jinja @@ -0,0 +1,154 @@ +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- macro render_content(content, do_vision_count, is_system_content=false) %} + {%- if content is string %} + {{- content }} + {%- elif content is iterable and content is not mapping %} + {%- for item in content %} + {%- if 'image' in item or 'image_url' in item or item.type == 'image' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain images.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set image_count.value = image_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Picture ' ~ image_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|image_pad|><|vision_end|>' }} + {%- elif 'video' in item or item.type == 'video' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain videos.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set video_count.value = video_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Video ' ~ video_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|video_pad|><|vision_end|>' }} + {%- elif 'text' in item %} + {{- item.text }} + {%- else %} + {{- raise_exception('Unexpected item type in content.') }} + {%- endif %} + {%- endfor %} + {%- elif content is none or content is undefined %} + {{- '' }} + {%- else %} + {{- raise_exception('Unexpected content type.') }} + {%- endif %} +{%- endmacro %} +{%- if not messages %} + {{- raise_exception('No messages provided.') }} +{%- endif %} +{%- if tools and tools is iterable and tools is not mapping %} + {{- '<|im_start|>system\n' }} + {{- "# Tools\n\nYou have access to the following functions:\n\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n" }} + {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {%- if content %} + {{- '\n\n' + content }} + {%- endif %} + {%- endif %} + {{- '<|im_end|>\n' }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" %} + {%- set content = render_content(message.content, false)|trim %} + {%- if not(content.startswith('') and content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if ns.multi_step_tool %} + {{- raise_exception('No user query found in messages.') }} +{%- endif %} +{%- for message in messages %} + {%- set content = render_content(message.content, true)|trim %} + {%- if message.role == "system" %} + {%- if not loop.first %} + {{- raise_exception('System message must be at the beginning.') }} + {%- endif %} + {%- elif message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = '' %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- set reasoning_content = reasoning_content|trim %} + {%- if loop.index0 > ns.last_query_index %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {%- if loop.first %} + {%- if content|trim %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n\n' }} + {%- endif %} + {%- else %} + {{- '\n\n\n' }} + {%- endif %} + {%- if tool_call.arguments is defined %} + {%- for args_name, args_value in tool_call.arguments|items %} + {{- '\n' }} + {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %} + {{- args_value }} + {{- '\n\n' }} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.previtem and loop.previtem.role != "tool" %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if not loop.last and loop.nextitem.role != "tool" %} + {{- '<|im_end|>\n' }} + {%- elif loop.last %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- else %} + {{- raise_exception('Unexpected message role.') }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n' }} + {%- endif %} +{%- endif %} \ No newline at end of file diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/config.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/config.json new file mode 100644 index 0000000000000000000000000000000000000000..314297e9ae3a6f7ca90b01d29ee0a46cfbf53c2a --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/config.json @@ -0,0 +1,119 @@ +{ + "architectures": [ + "Qwen35VLLatentForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 248046, + "hidden_size": 2560, + "image_token_id": 248056, + "latent_end_id": 248078, + "latent_pad_id": 248079, + "latent_start_id": 248077, + "model_type": "qwen3_5", + "pad_token_id": 248044, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "attn_output_gate": true, + "bos_token_id": null, + "dtype": "bfloat16", + "eos_token_id": 248044, + "full_attention_interval": 4, + "head_dim": 256, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9216, + "latent_end_id": 248078, + "latent_pad_id": 248079, + "latent_start_id": 248077, + "layer_types": [ + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention" + ], + "linear_conv_kernel_dim": 4, + "linear_key_head_dim": 128, + "linear_num_key_heads": 16, + "linear_num_value_heads": 32, + "linear_value_head_dim": 128, + "mamba_ssm_dtype": "float32", + "max_position_embeddings": 262144, + "mlp_only_layers": [], + "model_type": "qwen3_5_text", + "mtp_num_hidden_layers": 1, + "mtp_use_dedicated_embeddings": false, + "num_attention_heads": 16, + "num_hidden_layers": 32, + "num_key_value_heads": 4, + "pad_token_id": 248044, + "partial_rotary_factor": 0.25, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "mrope_interleaved": true, + "mrope_section": [ + 11, + 11, + 10 + ], + "partial_rotary_factor": 0.25, + "rope_theta": 10000000, + "rope_type": "default" + }, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 248080 + }, + "tie_word_embeddings": true, + "transformers_version": "5.8.1", + "use_cache": false, + "video_token_id": 248057, + "vision_config": { + "deepstack_visual_indexes": [], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_5_vision", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 248054, + "vision_start_token_id": 248053 +} diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/generation_config.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1f1ab970fed81cb788c2945bed33833d247b9091 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "eos_token_id": [ + 248044, + 248046 + ], + "transformers_version": "5.8.1", + "use_cache": true +} diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..5520a4c1b35a8fa21311e18ae92de96b1e47111d --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bba262c41b51e4a1dc02efbc93b2de371230e43ea61125b743720b8542609026 +size 6333306959 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..adefc753f5a6868800b52be57c146ea1bde3eecb --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:737e53b3e74e28c82c469a75555e956ff4c081656416b269d7f09d88d58f11e4 +size 6333311375 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..b842f53e1cdb8a93c3675e6a2855ea68aa88dc9f --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0b07a47cad629cefedfb2131a281525f796f1ea8f104cbde23c81f401330943d +size 6333312911 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..56e8befdc72fc7dd540342d495a9016493f15169 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b0b3afae9095f578763cc08868125fe7cf158640a774def8e35c0377894cc792 +size 6333312591 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..1d2f6f032a180d93e65aa07a044aebe5130a04bd --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b97d71bee2efc8a8fa2c4d3ebb25c60c08e2d1e618972f3adab7d569659f04f +size 6333312463 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..59e96e5b5803c5c7d28193c5bb171b896231a3af --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52f10fcac34cc787456dfb322073f9cc39ffddefbff6448b9a76b4b35ca5fe7c +size 6333312783 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..8ef00de699151cdf35299bcac3e341c505b67729 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06fa2bf6499b4c9f42a337c2d2805f8af8595ec8d484f821af55e1df9b8aa198 +size 6333312591 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..473c7a5e3de2ebf4dd43d9727f8f9752f9046a6d --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ad744d73b1bf72241550639f881096eca07a7fabfd9c6607c560bc7546f9fe3b +size 6333312527 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/mp_rank_00_model_states.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/mp_rank_00_model_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..b58bdb12e3af0584a7fcf3d41adfb5b8572f0c24 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/global_step4336/mp_rank_00_model_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:704e908ac3067060f3e8f7465c4a335452fdce75d86d7d4b0844f153b9ae1e03 +size 9778831179 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/latest b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/latest new file mode 100644 index 0000000000000000000000000000000000000000..8b8d2c63230a2474a05594980ba040feb389e875 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/latest @@ -0,0 +1 @@ +global_step4336 \ No newline at end of file diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00001-of-00003.safetensors b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..022ac7450119ed4ba5854ec7acc1ad908688a81f --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ffcde7038829569e92762a15a987ab936d60df83eff5b67d75266dde71d875c5 +size 4989860336 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00002-of-00003.safetensors b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..10b752f77dfbf608b9066c52a477509dbbc6d237 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e418025055ab7cf25ecc8c001476955250477a171075fad5b991dadb97e58dd +size 4993498424 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00003-of-00003.safetensors b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ab8023a740649342f1bc7e224320898de73054c --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9132529e66cac3079a828b70ff720888e735b62dcc8a839f7c42730ef896d42c +size 398327248 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model.safetensors.index.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..6a37ec4c02c77592002094503fd812d32a3c3c13 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/model.safetensors.index.json @@ -0,0 +1,747 @@ +{ + "metadata": { + "total_parameters": 4555712896, + "total_size": 10381595392 + }, + "weight_map": { + "alignment_projector.action_proj.bias": "model-00001-of-00003.safetensors", + "alignment_projector.action_proj.weight": "model-00001-of-00003.safetensors", + "alignment_projector.ln_action.bias": "model-00001-of-00003.safetensors", + "alignment_projector.ln_action.weight": "model-00001-of-00003.safetensors", + "alignment_projector.ln_state.bias": "model-00001-of-00003.safetensors", + "alignment_projector.ln_state.weight": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.1.bias": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.1.weight": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.3.bias": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.3.weight": "model-00001-of-00003.safetensors", + "alignment_projector.norm.bias": "model-00001-of-00003.safetensors", + "alignment_projector.norm.weight": "model-00001-of-00003.safetensors", + "alignment_projector.state_proj.bias": "model-00001-of-00003.safetensors", + "alignment_projector.state_proj.weight": "model-00001-of-00003.safetensors", + "latent_init_embedding": "model-00001-of-00003.safetensors", + "lm_head.weight": "model-00001-of-00003.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00003.safetensors" + } +} diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/preprocessor_config.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/processor_config.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..33818c7f9e991ad735fd240209f4fa73e6c28c50 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/processor_config.json @@ -0,0 +1,60 @@ +{ + "image_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_0.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..9c50bad4284aae9df2c12d94478e1fb07ad24f93 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d27399d48b3d16af30568e0145e4d09fb06420ccd997a54fe9ae37218d5280f +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_1.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..7e53f8310f11d7d966d1480e3b14003ee24999ed --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea33de107440eb2ba7f9351bce3d6204d85c65f0640592d82b30a61520357246 +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_2.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..5bdf167fb63ba28844f68a43b1e5b68a7117b5d2 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fac2509f18a8c3e1f81744d91dc5d3185f7233b6b1ec08b00b510a6fa84ade2 +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_3.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..5c9d0a1683cd9c83637e9d61e60c883a897a215c --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:343704e630d035e911942381df8c936e665db91f72d3cce99c82384764d7b3af +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_4.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_4.pth new file mode 100644 index 0000000000000000000000000000000000000000..03d6d187a07f08a1a516eaa66fe0c881b00998f7 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_4.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92bbb8beb1d76bd7e604ebdd77c432076af92db3fa12af94c6da6f81633ef291 +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_5.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_5.pth new file mode 100644 index 0000000000000000000000000000000000000000..b32d9a2cd99cd28c7898bcb675348c40a748f4a5 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_5.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8dcd37c8d4d5be10c02ecbcbbbe0e510998be9a425e88986b1b7280cdb4351fa +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_6.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_6.pth new file mode 100644 index 0000000000000000000000000000000000000000..d0f7edc5467e6d6c5f9879eb7f194f1167f46c92 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_6.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fca419a8ca44081663b31485e6bf41afc745e067c403e9b95897d1fa44945da0 +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_7.pth b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_7.pth new file mode 100644 index 0000000000000000000000000000000000000000..a760749b7cd38dd23e8d155a561b03c28a57a2ec --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/rng_state_7.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:82825b18c674de95162208cad3eef4e0b1946370f5af54ba40d8d8632d1e253c +size 16389 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/scheduler.pt b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..b4e0bad54a5c2197db3a74fa5cc7a92648b624c6 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a586e158b848c9f998312c0b44c1c34d4a47a8ffdb0724aca165b9133b269d15 +size 1529 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/tokenizer.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..9e8ac8634b7c90a05a65182d43516ddf7371c521 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4012eb888c8edaca05866241a3a8319ca02d45c367eef3da25b8ba4ef002f5b3 +size 19989885 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/tokenizer_config.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1d134cd298be1e3be25db393d93a1cefe80e3214 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/tokenizer_config.json @@ -0,0 +1,33 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "image_token": "<|image_pad|>", + "is_local": true, + "local_files_only": false, + "model_max_length": 262144, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +} diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/trainer_state.json b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..51f08bcc4ad3e32328ee2bd910c9f39d0b590cca --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/trainer_state.json @@ -0,0 +1,61354 @@ +{ + "best_global_step": 4336, + "best_metric": 0.56188542, + "best_model_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr/v3-20260618-000502/checkpoint-4336", + "epoch": 1.0, + "eval_steps": 100, + "global_step": 4336, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.00023063063063063062, + "grad_norm": 309.599853515625, + "learning_rate": 5.0000000000000004e-08, + "loss": 19.15211296081543, + "loss_alignment": 1.01953125, + "loss_alignment_w": 0.509765625, + "loss_sft": 0.686509020626545, + "loss_total_v6": 1.1970070600509644, + "step": 1 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.00046126126126126124, + "grad_norm": 354.2319030761719, + "learning_rate": 1.0000000000000001e-07, + "loss": 19.568588256835938, + "loss_alignment": 1.0302734375, + "loss_alignment_w": 0.51513671875, + "loss_sft": 0.7080221176147461, + "loss_total_v6": 1.2230368256568909, + "step": 2 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0006918918918918919, + "grad_norm": 388.2864990234375, + "learning_rate": 1.5000000000000002e-07, + "loss": 20.205001831054688, + "loss_alignment": 1.0322265625, + "loss_alignment_w": 0.51611328125, + "loss_sft": 0.7468671947717667, + "loss_total_v6": 1.2628126293420792, + "step": 3 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0009225225225225225, + "grad_norm": 371.87725830078125, + "learning_rate": 2.0000000000000002e-07, + "loss": 19.96813201904297, + "loss_alignment": 1.033203125, + "loss_alignment_w": 0.5166015625, + "loss_sft": 0.7314219400286674, + "loss_total_v6": 1.2480083107948303, + "step": 4 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0011531531531531532, + "grad_norm": 345.0340576171875, + "learning_rate": 2.5000000000000004e-07, + "loss": 19.540603637695312, + "loss_alignment": 1.033203125, + "loss_alignment_w": 0.5166015625, + "loss_sft": 0.7059221044182777, + "loss_total_v6": 1.2212877124547958, + "step": 5 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0013837837837837837, + "grad_norm": 378.2585144042969, + "learning_rate": 3.0000000000000004e-07, + "loss": 20.0699462890625, + "loss_alignment": 1.025390625, + "loss_alignment_w": 0.5126953125, + "loss_sft": 0.7396315485239029, + "loss_total_v6": 1.254371553659439, + "step": 6 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0016144144144144145, + "grad_norm": 367.7576599121094, + "learning_rate": 3.5000000000000004e-07, + "loss": 19.794151306152344, + "loss_alignment": 1.01953125, + "loss_alignment_w": 0.509765625, + "loss_sft": 0.7257818430662155, + "loss_total_v6": 1.237134411931038, + "step": 7 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.001845045045045045, + "grad_norm": 397.68310546875, + "learning_rate": 4.0000000000000003e-07, + "loss": 19.756786346435547, + "loss_alignment": 1.03125, + "loss_alignment_w": 0.515625, + "loss_sft": 0.719158910214901, + "loss_total_v6": 1.2347992062568665, + "step": 8 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0020756756756756755, + "grad_norm": 343.6162109375, + "learning_rate": 4.5000000000000003e-07, + "loss": 19.077781677246094, + "loss_alignment": 1.025390625, + "loss_alignment_w": 0.5126953125, + "loss_sft": 0.6778959110379219, + "loss_total_v6": 1.1923612356185913, + "step": 9 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0023063063063063064, + "grad_norm": 348.3703918457031, + "learning_rate": 5.000000000000001e-07, + "loss": 18.910926818847656, + "loss_alignment": 1.02734375, + "loss_alignment_w": 0.513671875, + "loss_sft": 0.6682763323187828, + "loss_total_v6": 1.1819329261779785, + "step": 10 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.002536936936936937, + "grad_norm": 312.861328125, + "learning_rate": 5.5e-07, + "loss": 18.488842010498047, + "loss_alignment": 1.01416015625, + "loss_alignment_w": 0.507080078125, + "loss_sft": 0.6468857601284981, + "loss_total_v6": 1.1555527448654175, + "step": 11 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0027675675675675675, + "grad_norm": 256.996337890625, + "learning_rate": 6.000000000000001e-07, + "loss": 18.30491828918457, + "loss_alignment": 1.0224609375, + "loss_alignment_w": 0.51123046875, + "loss_sft": 0.6310264095664024, + "loss_total_v6": 1.1440574079751968, + "step": 12 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0029981981981981984, + "grad_norm": 224.00747680664062, + "learning_rate": 6.5e-07, + "loss": 17.422832489013672, + "loss_alignment": 1.0048828125, + "loss_alignment_w": 0.50244140625, + "loss_sft": 0.5839220210909843, + "loss_total_v6": 1.0889269262552261, + "step": 13 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.003228828828828829, + "grad_norm": 162.4442596435547, + "learning_rate": 7.000000000000001e-07, + "loss": 16.427976608276367, + "loss_alignment": 1.0126953125, + "loss_alignment_w": 0.50634765625, + "loss_sft": 0.5176542773842812, + "loss_total_v6": 1.0267485231161118, + "step": 14 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0034594594594594594, + "grad_norm": 134.44862365722656, + "learning_rate": 7.5e-07, + "loss": 16.13473129272461, + "loss_alignment": 1.0078125, + "loss_alignment_w": 0.50390625, + "loss_sft": 0.5026529207825661, + "loss_total_v6": 1.0084207355976105, + "step": 15 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.00369009009009009, + "grad_norm": 129.0150909423828, + "learning_rate": 8.000000000000001e-07, + "loss": 15.670186996459961, + "loss_alignment": 1.0087890625, + "loss_alignment_w": 0.50439453125, + "loss_sft": 0.47204722464084625, + "loss_total_v6": 0.979386679828167, + "step": 16 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.003920720720720721, + "grad_norm": 105.92912292480469, + "learning_rate": 8.500000000000001e-07, + "loss": 15.822564125061035, + "loss_alignment": 0.99853515625, + "loss_alignment_w": 0.499267578125, + "loss_sft": 0.4878879189491272, + "loss_total_v6": 0.9889102652668953, + "step": 17 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.004151351351351351, + "grad_norm": 98.16565704345703, + "learning_rate": 9.000000000000001e-07, + "loss": 15.392788887023926, + "loss_alignment": 0.98974609375, + "loss_alignment_w": 0.494873046875, + "loss_sft": 0.46966341510415077, + "loss_total_v6": 0.9620493203401566, + "step": 18 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.004381981981981982, + "grad_norm": 73.03960418701172, + "learning_rate": 9.500000000000001e-07, + "loss": 14.874799728393555, + "loss_alignment": 0.9853515625, + "loss_alignment_w": 0.49267578125, + "loss_sft": 0.43931853771209717, + "loss_total_v6": 0.9296749606728554, + "step": 19 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.004612612612612613, + "grad_norm": 95.68405151367188, + "learning_rate": 1.0000000000000002e-06, + "loss": 14.764375686645508, + "loss_alignment": 0.93896484375, + "loss_alignment_w": 0.469482421875, + "loss_sft": 0.45173468068242073, + "loss_total_v6": 0.9227734804153442, + "step": 20 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.004843243243243243, + "grad_norm": 80.20225524902344, + "learning_rate": 1.0500000000000001e-06, + "loss": 14.682424545288086, + "loss_alignment": 0.94140625, + "loss_alignment_w": 0.470703125, + "loss_sft": 0.4447206221520901, + "loss_total_v6": 0.917651541531086, + "step": 21 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.005073873873873874, + "grad_norm": 73.7504653930664, + "learning_rate": 1.1e-06, + "loss": 14.50346565246582, + "loss_alignment": 0.9287109375, + "loss_alignment_w": 0.46435546875, + "loss_sft": 0.44108882918953896, + "loss_total_v6": 0.9064666405320168, + "step": 22 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.005304504504504505, + "grad_norm": 66.2354965209961, + "learning_rate": 1.1500000000000002e-06, + "loss": 14.457942962646484, + "loss_alignment": 0.93310546875, + "loss_alignment_w": 0.466552734375, + "loss_sft": 0.43629052117466927, + "loss_total_v6": 0.9036214053630829, + "step": 23 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.005535135135135135, + "grad_norm": 54.1392707824707, + "learning_rate": 1.2000000000000002e-06, + "loss": 13.57803726196289, + "loss_alignment": 0.86474609375, + "loss_alignment_w": 0.432373046875, + "loss_sft": 0.41871094331145287, + "loss_total_v6": 0.8486273139715195, + "step": 24 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.005765765765765766, + "grad_norm": 50.91067123413086, + "learning_rate": 1.25e-06, + "loss": 13.286900520324707, + "loss_alignment": 0.84375, + "loss_alignment_w": 0.421875, + "loss_sft": 0.40799177065491676, + "loss_total_v6": 0.8304313123226166, + "step": 25 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.005996396396396397, + "grad_norm": 51.64557647705078, + "learning_rate": 1.3e-06, + "loss": 13.14913558959961, + "loss_alignment": 0.82470703125, + "loss_alignment_w": 0.412353515625, + "loss_sft": 0.4074991047382355, + "loss_total_v6": 0.8218209594488144, + "step": 26 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.006227027027027027, + "grad_norm": 49.315956115722656, + "learning_rate": 1.3500000000000002e-06, + "loss": 13.03608512878418, + "loss_alignment": 0.814453125, + "loss_alignment_w": 0.4072265625, + "loss_sft": 0.40792546421289444, + "loss_total_v6": 0.8147553205490112, + "step": 27 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.006457657657657658, + "grad_norm": 47.43415451049805, + "learning_rate": 1.4000000000000001e-06, + "loss": 12.539966583251953, + "loss_alignment": 0.798828125, + "loss_alignment_w": 0.3994140625, + "loss_sft": 0.3861496187746525, + "loss_total_v6": 0.7837478891015053, + "step": 28 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.006688288288288288, + "grad_norm": 47.50248336791992, + "learning_rate": 1.45e-06, + "loss": 12.494140625, + "loss_alignment": 0.7705078125, + "loss_alignment_w": 0.38525390625, + "loss_sft": 0.3955230116844177, + "loss_total_v6": 0.7808837443590164, + "step": 29 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.006918918918918919, + "grad_norm": 47.45895767211914, + "learning_rate": 1.5e-06, + "loss": 12.16197395324707, + "loss_alignment": 0.74755859375, + "loss_alignment_w": 0.373779296875, + "loss_sft": 0.38747328892350197, + "loss_total_v6": 0.7601233944296837, + "step": 30 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.00714954954954955, + "grad_norm": 50.05726623535156, + "learning_rate": 1.5500000000000002e-06, + "loss": 11.912944793701172, + "loss_alignment": 0.74169921875, + "loss_alignment_w": 0.370849609375, + "loss_sft": 0.37424346432089806, + "loss_total_v6": 0.7445590272545815, + "step": 31 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.00738018018018018, + "grad_norm": 42.83274841308594, + "learning_rate": 1.6000000000000001e-06, + "loss": 11.831995010375977, + "loss_alignment": 0.72412109375, + "loss_alignment_w": 0.362060546875, + "loss_sft": 0.3777596093714237, + "loss_total_v6": 0.7394997403025627, + "step": 32 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.007610810810810811, + "grad_norm": 42.3469352722168, + "learning_rate": 1.6500000000000003e-06, + "loss": 11.597698211669922, + "loss_alignment": 0.6875, + "loss_alignment_w": 0.34375, + "loss_sft": 0.3817012682557106, + "loss_total_v6": 0.7248561754822731, + "step": 33 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.007841441441441442, + "grad_norm": 39.23835754394531, + "learning_rate": 1.7000000000000002e-06, + "loss": 11.317506790161133, + "loss_alignment": 0.66552734375, + "loss_alignment_w": 0.332763671875, + "loss_sft": 0.3744431994855404, + "loss_total_v6": 0.7073442116379738, + "step": 34 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.008072072072072072, + "grad_norm": 40.98651885986328, + "learning_rate": 1.75e-06, + "loss": 11.074630737304688, + "loss_alignment": 0.64990234375, + "loss_alignment_w": 0.324951171875, + "loss_sft": 0.36689287051558495, + "loss_total_v6": 0.6921644434332848, + "step": 35 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.008302702702702702, + "grad_norm": 39.303863525390625, + "learning_rate": 1.8000000000000001e-06, + "loss": 10.906831741333008, + "loss_alignment": 0.6435546875, + "loss_alignment_w": 0.32177734375, + "loss_sft": 0.3595792129635811, + "loss_total_v6": 0.6816769912838936, + "step": 36 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.008533333333333334, + "grad_norm": 37.62526321411133, + "learning_rate": 1.85e-06, + "loss": 11.052016258239746, + "loss_alignment": 0.6328125, + "loss_alignment_w": 0.31640625, + "loss_sft": 0.37555019557476044, + "loss_total_v6": 0.6907510235905647, + "step": 37 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.008763963963963964, + "grad_norm": 38.192291259765625, + "learning_rate": 1.9000000000000002e-06, + "loss": 10.980457305908203, + "loss_alignment": 0.6162109375, + "loss_alignment_w": 0.30810546875, + "loss_sft": 0.3788292407989502, + "loss_total_v6": 0.6862786039710045, + "step": 38 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.008994594594594594, + "grad_norm": 45.19548797607422, + "learning_rate": 1.9500000000000004e-06, + "loss": 10.602665901184082, + "loss_alignment": 0.6083984375, + "loss_alignment_w": 0.30419921875, + "loss_sft": 0.3588031008839607, + "loss_total_v6": 0.6626666262745857, + "step": 39 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.009225225225225226, + "grad_norm": 40.3022575378418, + "learning_rate": 2.0000000000000003e-06, + "loss": 10.657114028930664, + "loss_alignment": 0.578125, + "loss_alignment_w": 0.2890625, + "loss_sft": 0.37630515918135643, + "loss_total_v6": 0.6660696044564247, + "step": 40 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.009455855855855856, + "grad_norm": 32.95872116088867, + "learning_rate": 2.05e-06, + "loss": 10.208895683288574, + "loss_alignment": 0.564453125, + "loss_alignment_w": 0.2822265625, + "loss_sft": 0.355615820735693, + "loss_total_v6": 0.6380560025572777, + "step": 41 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.009686486486486486, + "grad_norm": 35.36344528198242, + "learning_rate": 2.1000000000000002e-06, + "loss": 10.409223556518555, + "loss_alignment": 0.55322265625, + "loss_alignment_w": 0.276611328125, + "loss_sft": 0.37451445311307907, + "loss_total_v6": 0.6505764946341515, + "step": 42 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.009917117117117118, + "grad_norm": 35.61820983886719, + "learning_rate": 2.15e-06, + "loss": 9.948071479797363, + "loss_alignment": 0.541015625, + "loss_alignment_w": 0.2705078125, + "loss_sft": 0.3513534553349018, + "loss_total_v6": 0.6217544674873352, + "step": 43 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.010147747747747748, + "grad_norm": 38.396976470947266, + "learning_rate": 2.2e-06, + "loss": 10.070068359375, + "loss_alignment": 0.5361328125, + "loss_alignment_w": 0.26806640625, + "loss_sft": 0.3612671233713627, + "loss_total_v6": 0.6293793022632599, + "step": 44 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.010378378378378378, + "grad_norm": 37.754451751708984, + "learning_rate": 2.25e-06, + "loss": 9.731096267700195, + "loss_alignment": 0.52392578125, + "loss_alignment_w": 0.261962890625, + "loss_sft": 0.3459712006151676, + "loss_total_v6": 0.6081935316324234, + "step": 45 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01060900900900901, + "grad_norm": 38.72932434082031, + "learning_rate": 2.3000000000000004e-06, + "loss": 9.693443298339844, + "loss_alignment": 0.52001953125, + "loss_alignment_w": 0.260009765625, + "loss_sft": 0.34569310769438744, + "loss_total_v6": 0.6058401986956596, + "step": 46 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01083963963963964, + "grad_norm": 32.00844192504883, + "learning_rate": 2.35e-06, + "loss": 10.195390701293945, + "loss_alignment": 0.5166015625, + "loss_alignment_w": 0.25830078125, + "loss_sft": 0.3788805715739727, + "loss_total_v6": 0.637211874127388, + "step": 47 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01107027027027027, + "grad_norm": 37.83436584472656, + "learning_rate": 2.4000000000000003e-06, + "loss": 9.682966232299805, + "loss_alignment": 0.497802734375, + "loss_alignment_w": 0.2489013671875, + "loss_sft": 0.3557804860174656, + "loss_total_v6": 0.6051853522658348, + "step": 48 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.011300900900900902, + "grad_norm": 35.604854583740234, + "learning_rate": 2.4500000000000003e-06, + "loss": 9.279752731323242, + "loss_alignment": 0.496337890625, + "loss_alignment_w": 0.2481689453125, + "loss_sft": 0.33157144114375114, + "loss_total_v6": 0.5799845457077026, + "step": 49 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.011531531531531532, + "grad_norm": 36.18699645996094, + "learning_rate": 2.5e-06, + "loss": 9.231813430786133, + "loss_alignment": 0.477783203125, + "loss_alignment_w": 0.2388916015625, + "loss_sft": 0.33763896860182285, + "loss_total_v6": 0.5769883282482624, + "step": 50 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.011762162162162162, + "grad_norm": 39.88041305541992, + "learning_rate": 2.55e-06, + "loss": 8.902301788330078, + "loss_alignment": 0.466064453125, + "loss_alignment_w": 0.2330322265625, + "loss_sft": 0.3235600218176842, + "loss_total_v6": 0.5563938990235329, + "step": 51 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.011992792792792794, + "grad_norm": 34.067806243896484, + "learning_rate": 2.6e-06, + "loss": 9.628734588623047, + "loss_alignment": 0.468505859375, + "loss_alignment_w": 0.2342529296875, + "loss_sft": 0.3677718862891197, + "loss_total_v6": 0.6017959080636501, + "step": 52 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.012223423423423424, + "grad_norm": 39.24044418334961, + "learning_rate": 2.6500000000000005e-06, + "loss": 9.394180297851562, + "loss_alignment": 0.458740234375, + "loss_alignment_w": 0.2293701171875, + "loss_sft": 0.35791876539587975, + "loss_total_v6": 0.5871363133192062, + "step": 53 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.012454054054054054, + "grad_norm": 37.47954177856445, + "learning_rate": 2.7000000000000004e-06, + "loss": 9.038666725158691, + "loss_alignment": 0.440673828125, + "loss_alignment_w": 0.2203369140625, + "loss_sft": 0.3444729670882225, + "loss_total_v6": 0.56491669267416, + "step": 54 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.012684684684684684, + "grad_norm": 35.83920669555664, + "learning_rate": 2.7500000000000004e-06, + "loss": 9.01968765258789, + "loss_alignment": 0.4345703125, + "loss_alignment_w": 0.21728515625, + "loss_sft": 0.34647584334015846, + "loss_total_v6": 0.5637304708361626, + "step": 55 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.012915315315315316, + "grad_norm": 36.870018005371094, + "learning_rate": 2.8000000000000003e-06, + "loss": 9.044709205627441, + "loss_alignment": 0.42724609375, + "loss_alignment_w": 0.213623046875, + "loss_sft": 0.35183916613459587, + "loss_total_v6": 0.5652943551540375, + "step": 56 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.013145945945945946, + "grad_norm": 32.982051849365234, + "learning_rate": 2.85e-06, + "loss": 8.958125114440918, + "loss_alignment": 0.42138671875, + "loss_alignment_w": 0.210693359375, + "loss_sft": 0.348991084843874, + "loss_total_v6": 0.5598828122019768, + "step": 57 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.013376576576576576, + "grad_norm": 37.6244010925293, + "learning_rate": 2.9e-06, + "loss": 8.609430313110352, + "loss_alignment": 0.411376953125, + "loss_alignment_w": 0.2056884765625, + "loss_sft": 0.33220258355140686, + "loss_total_v6": 0.5380894057452679, + "step": 58 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.013607207207207208, + "grad_norm": 31.17098045349121, + "learning_rate": 2.95e-06, + "loss": 9.27241039276123, + "loss_alignment": 0.404541015625, + "loss_alignment_w": 0.2022705078125, + "loss_sft": 0.3778654970228672, + "loss_total_v6": 0.5795256495475769, + "step": 59 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.013837837837837838, + "grad_norm": 31.500625610351562, + "learning_rate": 3e-06, + "loss": 8.63952922821045, + "loss_alignment": 0.394775390625, + "loss_alignment_w": 0.1973876953125, + "loss_sft": 0.3429948315024376, + "loss_total_v6": 0.5399705469608307, + "step": 60 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.014068468468468468, + "grad_norm": 33.50215148925781, + "learning_rate": 3.05e-06, + "loss": 8.188718795776367, + "loss_alignment": 0.393310546875, + "loss_alignment_w": 0.1966552734375, + "loss_sft": 0.31518546119332314, + "loss_total_v6": 0.5117949657142162, + "step": 61 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0142990990990991, + "grad_norm": 34.041107177734375, + "learning_rate": 3.1000000000000004e-06, + "loss": 8.704641342163086, + "loss_alignment": 0.391845703125, + "loss_alignment_w": 0.1959228515625, + "loss_sft": 0.34831557981669903, + "loss_total_v6": 0.5440400801599026, + "step": 62 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01452972972972973, + "grad_norm": 38.78754425048828, + "learning_rate": 3.1500000000000003e-06, + "loss": 8.267170906066895, + "loss_alignment": 0.380126953125, + "loss_alignment_w": 0.1900634765625, + "loss_sft": 0.3265736699104309, + "loss_total_v6": 0.5166981630027294, + "step": 63 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01476036036036036, + "grad_norm": 40.594970703125, + "learning_rate": 3.2000000000000003e-06, + "loss": 8.342784881591797, + "loss_alignment": 0.384033203125, + "loss_alignment_w": 0.1920166015625, + "loss_sft": 0.3295600078999996, + "loss_total_v6": 0.5214240550994873, + "step": 64 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.014990990990990992, + "grad_norm": 37.221656799316406, + "learning_rate": 3.2500000000000002e-06, + "loss": 8.769034385681152, + "loss_alignment": 0.375732421875, + "loss_alignment_w": 0.1878662109375, + "loss_sft": 0.3597712144255638, + "loss_total_v6": 0.5480646565556526, + "step": 65 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.015221621621621622, + "grad_norm": 38.77353286743164, + "learning_rate": 3.3000000000000006e-06, + "loss": 8.230737686157227, + "loss_alignment": 0.372314453125, + "loss_alignment_w": 0.1861572265625, + "loss_sft": 0.32815706729888916, + "loss_total_v6": 0.514421109110117, + "step": 66 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.015452252252252252, + "grad_norm": 35.560733795166016, + "learning_rate": 3.3500000000000005e-06, + "loss": 8.320768356323242, + "loss_alignment": 0.361328125, + "loss_alignment_w": 0.1806640625, + "loss_sft": 0.33958227559924126, + "loss_total_v6": 0.5200479589402676, + "step": 67 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.015682882882882884, + "grad_norm": 36.12428283691406, + "learning_rate": 3.4000000000000005e-06, + "loss": 8.599283218383789, + "loss_alignment": 0.359130859375, + "loss_alignment_w": 0.1795654296875, + "loss_sft": 0.35758454725146294, + "loss_total_v6": 0.5374551713466644, + "step": 68 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.015913513513513514, + "grad_norm": 37.68132019042969, + "learning_rate": 3.45e-06, + "loss": 7.7437052726745605, + "loss_alignment": 0.352783203125, + "loss_alignment_w": 0.1763916015625, + "loss_sft": 0.30786462128162384, + "loss_total_v6": 0.48398157209157944, + "step": 69 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.016144144144144144, + "grad_norm": 36.67192840576172, + "learning_rate": 3.5e-06, + "loss": 8.300985336303711, + "loss_alignment": 0.352294921875, + "loss_alignment_w": 0.1761474609375, + "loss_sft": 0.34257254749536514, + "loss_total_v6": 0.5188115537166595, + "step": 70 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.016374774774774774, + "grad_norm": 31.640045166015625, + "learning_rate": 3.5500000000000003e-06, + "loss": 8.316624641418457, + "loss_alignment": 0.347412109375, + "loss_alignment_w": 0.1737060546875, + "loss_sft": 0.34600669890642166, + "loss_total_v6": 0.5197890587151051, + "step": 71 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.016605405405405404, + "grad_norm": 33.716617584228516, + "learning_rate": 3.6000000000000003e-06, + "loss": 7.893542289733887, + "loss_alignment": 0.343505859375, + "loss_alignment_w": 0.1717529296875, + "loss_sft": 0.3215629681944847, + "loss_total_v6": 0.4933464154601097, + "step": 72 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.016836036036036037, + "grad_norm": 35.522979736328125, + "learning_rate": 3.65e-06, + "loss": 7.841757774353027, + "loss_alignment": 0.3408203125, + "loss_alignment_w": 0.17041015625, + "loss_sft": 0.3195776529610157, + "loss_total_v6": 0.490109883248806, + "step": 73 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.017066666666666667, + "grad_norm": 34.37610626220703, + "learning_rate": 3.7e-06, + "loss": 8.354498863220215, + "loss_alignment": 0.343994140625, + "loss_alignment_w": 0.1719970703125, + "loss_sft": 0.350220151245594, + "loss_total_v6": 0.5221561715006828, + "step": 74 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.017297297297297298, + "grad_norm": 31.361726760864258, + "learning_rate": 3.7500000000000005e-06, + "loss": 8.283336639404297, + "loss_alignment": 0.334228515625, + "loss_alignment_w": 0.1671142578125, + "loss_sft": 0.35053325071930885, + "loss_total_v6": 0.517708532512188, + "step": 75 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.017527927927927928, + "grad_norm": 36.72067642211914, + "learning_rate": 3.8000000000000005e-06, + "loss": 7.893118381500244, + "loss_alignment": 0.339111328125, + "loss_alignment_w": 0.1695556640625, + "loss_sft": 0.32347428798675537, + "loss_total_v6": 0.49331988766789436, + "step": 76 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.017758558558558558, + "grad_norm": 34.83134460449219, + "learning_rate": 3.85e-06, + "loss": 8.383413314819336, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.3520883210003376, + "loss_total_v6": 0.5239633210003376, + "step": 77 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.017989189189189188, + "grad_norm": 38.60700988769531, + "learning_rate": 3.900000000000001e-06, + "loss": 7.803718566894531, + "loss_alignment": 0.335205078125, + "loss_alignment_w": 0.1676025390625, + "loss_sft": 0.3199772387742996, + "loss_total_v6": 0.4877323806285858, + "step": 78 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01821981981981982, + "grad_norm": 34.75847625732422, + "learning_rate": 3.95e-06, + "loss": 8.186202049255371, + "loss_alignment": 0.334228515625, + "loss_alignment_w": 0.1671142578125, + "loss_sft": 0.3440808802843094, + "loss_total_v6": 0.5116376578807831, + "step": 79 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01845045045045045, + "grad_norm": 33.00989532470703, + "learning_rate": 4.000000000000001e-06, + "loss": 8.06666374206543, + "loss_alignment": 0.338623046875, + "loss_alignment_w": 0.1693115234375, + "loss_sft": 0.33506854623556137, + "loss_total_v6": 0.5041664615273476, + "step": 80 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01868108108108108, + "grad_norm": 33.75870895385742, + "learning_rate": 4.05e-06, + "loss": 8.265729904174805, + "loss_alignment": 0.33349609375, + "loss_alignment_w": 0.166748046875, + "loss_sft": 0.3498295210301876, + "loss_total_v6": 0.5166081041097641, + "step": 81 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01891171171171171, + "grad_norm": 30.48691749572754, + "learning_rate": 4.1e-06, + "loss": 7.970099449157715, + "loss_alignment": 0.3369140625, + "loss_alignment_w": 0.16845703125, + "loss_sft": 0.3296741843223572, + "loss_total_v6": 0.4981312043964863, + "step": 82 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.01914234234234234, + "grad_norm": 32.76646041870117, + "learning_rate": 4.15e-06, + "loss": 8.306663513183594, + "loss_alignment": 0.339599609375, + "loss_alignment_w": 0.1697998046875, + "loss_sft": 0.3491988703608513, + "loss_total_v6": 0.5191665217280388, + "step": 83 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.019372972972972972, + "grad_norm": 34.553035736083984, + "learning_rate": 4.2000000000000004e-06, + "loss": 8.072857856750488, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.3385684862732887, + "loss_total_v6": 0.504553597420454, + "step": 84 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.019603603603603605, + "grad_norm": 33.30038070678711, + "learning_rate": 4.25e-06, + "loss": 7.977458477020264, + "loss_alignment": 0.333984375, + "loss_alignment_w": 0.1669921875, + "loss_sft": 0.3317973390221596, + "loss_total_v6": 0.4985911585390568, + "step": 85 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.019834234234234235, + "grad_norm": 28.97978973388672, + "learning_rate": 4.3e-06, + "loss": 8.198236465454102, + "loss_alignment": 0.3408203125, + "loss_alignment_w": 0.17041015625, + "loss_sft": 0.34191855415701866, + "loss_total_v6": 0.5123897567391396, + "step": 86 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.020064864864864865, + "grad_norm": 32.486236572265625, + "learning_rate": 4.350000000000001e-06, + "loss": 8.047284126281738, + "loss_alignment": 0.327880859375, + "loss_alignment_w": 0.1639404296875, + "loss_sft": 0.33877066522836685, + "loss_total_v6": 0.5029552355408669, + "step": 87 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.020295495495495496, + "grad_norm": 36.17887496948242, + "learning_rate": 4.4e-06, + "loss": 8.102680206298828, + "loss_alignment": 0.328857421875, + "loss_alignment_w": 0.1644287109375, + "loss_sft": 0.34249233454465866, + "loss_total_v6": 0.5064174830913544, + "step": 88 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.020526126126126126, + "grad_norm": 40.74455642700195, + "learning_rate": 4.450000000000001e-06, + "loss": 7.810851097106934, + "loss_alignment": 0.327392578125, + "loss_alignment_w": 0.1636962890625, + "loss_sft": 0.3241614755243063, + "loss_total_v6": 0.48817818611860275, + "step": 89 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.020756756756756756, + "grad_norm": 36.004112243652344, + "learning_rate": 4.5e-06, + "loss": 7.791347980499268, + "loss_alignment": 0.329345703125, + "loss_alignment_w": 0.1646728515625, + "loss_sft": 0.3222711402922869, + "loss_total_v6": 0.4869592674076557, + "step": 90 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.020987387387387386, + "grad_norm": 39.171138763427734, + "learning_rate": 4.5500000000000005e-06, + "loss": 7.865126132965088, + "loss_alignment": 0.32958984375, + "loss_alignment_w": 0.164794921875, + "loss_sft": 0.326393973082304, + "loss_total_v6": 0.4915703870356083, + "step": 91 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02121801801801802, + "grad_norm": 39.14467239379883, + "learning_rate": 4.600000000000001e-06, + "loss": 7.773711204528809, + "loss_alignment": 0.3232421875, + "loss_alignment_w": 0.16162109375, + "loss_sft": 0.32429689541459084, + "loss_total_v6": 0.48585696518421173, + "step": 92 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02144864864864865, + "grad_norm": 37.92268753051758, + "learning_rate": 4.65e-06, + "loss": 7.970284938812256, + "loss_alignment": 0.330322265625, + "loss_alignment_w": 0.1651611328125, + "loss_sft": 0.332645982503891, + "loss_total_v6": 0.4981427974998951, + "step": 93 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02167927927927928, + "grad_norm": 34.39594650268555, + "learning_rate": 4.7e-06, + "loss": 7.9906206130981445, + "loss_alignment": 0.32861328125, + "loss_alignment_w": 0.164306640625, + "loss_sft": 0.33477143570780754, + "loss_total_v6": 0.49941377714276314, + "step": 94 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02190990990990991, + "grad_norm": 32.180885314941406, + "learning_rate": 4.75e-06, + "loss": 8.160640716552734, + "loss_alignment": 0.327880859375, + "loss_alignment_w": 0.1639404296875, + "loss_sft": 0.3460233323276043, + "loss_total_v6": 0.5100400447845459, + "step": 95 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02214054054054054, + "grad_norm": 37.33108139038086, + "learning_rate": 4.800000000000001e-06, + "loss": 7.862780570983887, + "loss_alignment": 0.32470703125, + "loss_alignment_w": 0.162353515625, + "loss_sft": 0.32943649031221867, + "loss_total_v6": 0.4914237968623638, + "step": 96 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02237117117117117, + "grad_norm": 37.376380920410156, + "learning_rate": 4.85e-06, + "loss": 7.852429389953613, + "loss_alignment": 0.323486328125, + "loss_alignment_w": 0.1617431640625, + "loss_sft": 0.32904891669750214, + "loss_total_v6": 0.49077681452035904, + "step": 97 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.022601801801801803, + "grad_norm": 33.00474166870117, + "learning_rate": 4.9000000000000005e-06, + "loss": 7.470846176147461, + "loss_alignment": 0.330322265625, + "loss_alignment_w": 0.1651611328125, + "loss_sft": 0.3016447089612484, + "loss_total_v6": 0.4669279083609581, + "step": 98 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.022832432432432433, + "grad_norm": 32.49423599243164, + "learning_rate": 4.95e-06, + "loss": 8.20756721496582, + "loss_alignment": 0.32763671875, + "loss_alignment_w": 0.163818359375, + "loss_sft": 0.34871210157871246, + "loss_total_v6": 0.512972965836525, + "step": 99 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.023063063063063063, + "grad_norm": 32.885990142822266, + "learning_rate": 5e-06, + "loss": 8.452499389648438, + "loss_alignment": 0.326416015625, + "loss_alignment_w": 0.1632080078125, + "loss_sft": 0.36501218751072884, + "loss_total_v6": 0.5282812230288982, + "step": 100 + }, + { + "epoch": 0.023063063063063063, + "eval_loss": 0.5970373749732971, + "eval_loss_alignment": 0.2991692529965753, + "eval_loss_alignment_w": 0.14958462649828766, + "eval_loss_sft": 0.44745285464578294, + "eval_loss_total_v6": 0.5970374810675235, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 95.0889, + "eval_samples_per_second": 18.351, + "eval_steps_per_second": 2.303, + "step": 100 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.023293693693693693, + "grad_norm": 32.03620910644531, + "learning_rate": 5.050000000000001e-06, + "loss": 7.92031717300415, + "loss_alignment": 0.32568359375, + "loss_alignment_w": 0.162841796875, + "loss_sft": 0.3320102132856846, + "loss_total_v6": 0.4950198531150818, + "step": 101 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.023524324324324324, + "grad_norm": 35.95601272583008, + "learning_rate": 5.1e-06, + "loss": 8.257457733154297, + "loss_alignment": 0.327880859375, + "loss_alignment_w": 0.1639404296875, + "loss_sft": 0.35193705931305885, + "loss_total_v6": 0.5160911045968533, + "step": 102 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.023754954954954954, + "grad_norm": 31.77120590209961, + "learning_rate": 5.150000000000001e-06, + "loss": 7.454949378967285, + "loss_alignment": 0.324951171875, + "loss_alignment_w": 0.1624755859375, + "loss_sft": 0.3034129813313484, + "loss_total_v6": 0.4659343399107456, + "step": 103 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.023985585585585587, + "grad_norm": 39.318241119384766, + "learning_rate": 5.2e-06, + "loss": 8.13177490234375, + "loss_alignment": 0.330322265625, + "loss_alignment_w": 0.1651611328125, + "loss_sft": 0.34316635876893997, + "loss_total_v6": 0.5082359425723553, + "step": 104 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.024216216216216217, + "grad_norm": 37.94911193847656, + "learning_rate": 5.2500000000000006e-06, + "loss": 7.638897895812988, + "loss_alignment": 0.31982421875, + "loss_alignment_w": 0.159912109375, + "loss_sft": 0.31785472109913826, + "loss_total_v6": 0.47743112221360207, + "step": 105 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.024446846846846847, + "grad_norm": 32.5563850402832, + "learning_rate": 5.300000000000001e-06, + "loss": 7.460195541381836, + "loss_alignment": 0.32275390625, + "loss_alignment_w": 0.161376953125, + "loss_sft": 0.3050226103514433, + "loss_total_v6": 0.46626223251223564, + "step": 106 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.024677477477477477, + "grad_norm": 34.01363754272461, + "learning_rate": 5.3500000000000004e-06, + "loss": 7.945626735687256, + "loss_alignment": 0.327392578125, + "loss_alignment_w": 0.1636962890625, + "loss_sft": 0.3326612561941147, + "loss_total_v6": 0.4966016784310341, + "step": 107 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.024908108108108108, + "grad_norm": 31.223146438598633, + "learning_rate": 5.400000000000001e-06, + "loss": 8.072397232055664, + "loss_alignment": 0.325439453125, + "loss_alignment_w": 0.1627197265625, + "loss_sft": 0.341560922563076, + "loss_total_v6": 0.5045248083770275, + "step": 108 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.025138738738738738, + "grad_norm": 34.79131317138672, + "learning_rate": 5.450000000000001e-06, + "loss": 7.757287979125977, + "loss_alignment": 0.322998046875, + "loss_alignment_w": 0.1614990234375, + "loss_sft": 0.32348407432436943, + "loss_total_v6": 0.4848305210471153, + "step": 109 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.025369369369369368, + "grad_norm": 36.02422332763672, + "learning_rate": 5.500000000000001e-06, + "loss": 7.641190528869629, + "loss_alignment": 0.32373046875, + "loss_alignment_w": 0.161865234375, + "loss_sft": 0.3158617727458477, + "loss_total_v6": 0.4775744155049324, + "step": 110 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0256, + "grad_norm": 33.08688735961914, + "learning_rate": 5.550000000000001e-06, + "loss": 8.136154174804688, + "loss_alignment": 0.329345703125, + "loss_alignment_w": 0.1646728515625, + "loss_sft": 0.3439130298793316, + "loss_total_v6": 0.5085095949470997, + "step": 111 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02583063063063063, + "grad_norm": 32.938236236572266, + "learning_rate": 5.600000000000001e-06, + "loss": 7.527066707611084, + "loss_alignment": 0.322265625, + "loss_alignment_w": 0.1611328125, + "loss_sft": 0.30962927266955376, + "loss_total_v6": 0.47044166922569275, + "step": 112 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02606126126126126, + "grad_norm": 33.89869689941406, + "learning_rate": 5.65e-06, + "loss": 7.833620071411133, + "loss_alignment": 0.32373046875, + "loss_alignment_w": 0.161865234375, + "loss_sft": 0.3278123326599598, + "loss_total_v6": 0.4896012656390667, + "step": 113 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02629189189189189, + "grad_norm": 40.58633804321289, + "learning_rate": 5.7e-06, + "loss": 7.3299407958984375, + "loss_alignment": 0.32275390625, + "loss_alignment_w": 0.161376953125, + "loss_sft": 0.29669853672385216, + "loss_total_v6": 0.45812128484249115, + "step": 114 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02652252252252252, + "grad_norm": 31.13019561767578, + "learning_rate": 5.75e-06, + "loss": 8.021743774414062, + "loss_alignment": 0.322265625, + "loss_alignment_w": 0.1611328125, + "loss_sft": 0.3402567058801651, + "loss_total_v6": 0.5013589933514595, + "step": 115 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02675315315315315, + "grad_norm": 34.4667854309082, + "learning_rate": 5.8e-06, + "loss": 7.5923309326171875, + "loss_alignment": 0.32568359375, + "loss_alignment_w": 0.162841796875, + "loss_sft": 0.3116941321641207, + "loss_total_v6": 0.4745206832885742, + "step": 116 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.026983783783783785, + "grad_norm": 37.40016174316406, + "learning_rate": 5.85e-06, + "loss": 7.4090576171875, + "loss_alignment": 0.3212890625, + "loss_alignment_w": 0.16064453125, + "loss_sft": 0.30216215550899506, + "loss_total_v6": 0.46306609734892845, + "step": 117 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.027214414414414415, + "grad_norm": 39.51665496826172, + "learning_rate": 5.9e-06, + "loss": 7.541941165924072, + "loss_alignment": 0.32470703125, + "loss_alignment_w": 0.162353515625, + "loss_sft": 0.30929247103631496, + "loss_total_v6": 0.4713713303208351, + "step": 118 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.027445045045045045, + "grad_norm": 33.59859085083008, + "learning_rate": 5.950000000000001e-06, + "loss": 7.176734924316406, + "loss_alignment": 0.3271484375, + "loss_alignment_w": 0.16357421875, + "loss_sft": 0.2852921560406685, + "loss_total_v6": 0.4485459476709366, + "step": 119 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.027675675675675675, + "grad_norm": 30.752450942993164, + "learning_rate": 6e-06, + "loss": 7.346385478973389, + "loss_alignment": 0.319580078125, + "loss_alignment_w": 0.1597900390625, + "loss_sft": 0.2996947281062603, + "loss_total_v6": 0.4591490961611271, + "step": 120 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.027906306306306305, + "grad_norm": 31.12807846069336, + "learning_rate": 6.0500000000000005e-06, + "loss": 7.300016403198242, + "loss_alignment": 0.323974609375, + "loss_alignment_w": 0.1619873046875, + "loss_sft": 0.29450785741209984, + "loss_total_v6": 0.45625101402401924, + "step": 121 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.028136936936936936, + "grad_norm": 31.823476791381836, + "learning_rate": 6.1e-06, + "loss": 7.997175216674805, + "loss_alignment": 0.32470703125, + "loss_alignment_w": 0.162353515625, + "loss_sft": 0.3373326361179352, + "loss_total_v6": 0.4998234696686268, + "step": 122 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02836756756756757, + "grad_norm": 32.06503677368164, + "learning_rate": 6.15e-06, + "loss": 7.392229080200195, + "loss_alignment": 0.318359375, + "loss_alignment_w": 0.1591796875, + "loss_sft": 0.302697267383337, + "loss_total_v6": 0.4620143137872219, + "step": 123 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0285981981981982, + "grad_norm": 32.64785385131836, + "learning_rate": 6.200000000000001e-06, + "loss": 7.835288047790527, + "loss_alignment": 0.314697265625, + "loss_alignment_w": 0.1573486328125, + "loss_sft": 0.33202118054032326, + "loss_total_v6": 0.48970550298690796, + "step": 124 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02882882882882883, + "grad_norm": 31.97476577758789, + "learning_rate": 6.25e-06, + "loss": 7.461186408996582, + "loss_alignment": 0.3193359375, + "loss_alignment_w": 0.15966796875, + "loss_sft": 0.3063357584178448, + "loss_total_v6": 0.46632416546344757, + "step": 125 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02905945945945946, + "grad_norm": 29.72202491760254, + "learning_rate": 6.300000000000001e-06, + "loss": 7.397162914276123, + "loss_alignment": 0.33251953125, + "loss_alignment_w": 0.166259765625, + "loss_sft": 0.29575773142278194, + "loss_total_v6": 0.4623226746916771, + "step": 126 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02929009009009009, + "grad_norm": 33.242916107177734, + "learning_rate": 6.35e-06, + "loss": 7.541790962219238, + "loss_alignment": 0.322998046875, + "loss_alignment_w": 0.1614990234375, + "loss_sft": 0.3095577135682106, + "loss_total_v6": 0.4713619016110897, + "step": 127 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.02952072072072072, + "grad_norm": 31.30267333984375, + "learning_rate": 6.4000000000000006e-06, + "loss": 7.625205039978027, + "loss_alignment": 0.323486328125, + "loss_alignment_w": 0.1617431640625, + "loss_sft": 0.3149084784090519, + "loss_total_v6": 0.4765753298997879, + "step": 128 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.029751351351351353, + "grad_norm": 31.909730911254883, + "learning_rate": 6.450000000000001e-06, + "loss": 7.33542537689209, + "loss_alignment": 0.316650390625, + "loss_alignment_w": 0.1583251953125, + "loss_sft": 0.2999252825975418, + "loss_total_v6": 0.4584640972316265, + "step": 129 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.029981981981981983, + "grad_norm": 30.961273193359375, + "learning_rate": 6.5000000000000004e-06, + "loss": 7.6855573654174805, + "loss_alignment": 0.317138671875, + "loss_alignment_w": 0.1585693359375, + "loss_sft": 0.32190007343888283, + "loss_total_v6": 0.4803473576903343, + "step": 130 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.030212612612612613, + "grad_norm": 31.44903564453125, + "learning_rate": 6.550000000000001e-06, + "loss": 7.387978553771973, + "loss_alignment": 0.31884765625, + "loss_alignment_w": 0.159423828125, + "loss_sft": 0.3018975891172886, + "loss_total_v6": 0.4617486670613289, + "step": 131 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.030443243243243243, + "grad_norm": 32.281368255615234, + "learning_rate": 6.600000000000001e-06, + "loss": 7.836825370788574, + "loss_alignment": 0.321044921875, + "loss_alignment_w": 0.1605224609375, + "loss_sft": 0.3294011764228344, + "loss_total_v6": 0.4898015782237053, + "step": 132 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.030673873873873873, + "grad_norm": 35.65143966674805, + "learning_rate": 6.650000000000001e-06, + "loss": 7.2022786140441895, + "loss_alignment": 0.315185546875, + "loss_alignment_w": 0.1575927734375, + "loss_sft": 0.2922444436699152, + "loss_total_v6": 0.45014240220189095, + "step": 133 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.030904504504504503, + "grad_norm": 39.8371696472168, + "learning_rate": 6.700000000000001e-06, + "loss": 7.508930206298828, + "loss_alignment": 0.318115234375, + "loss_alignment_w": 0.1590576171875, + "loss_sft": 0.31031152606010437, + "loss_total_v6": 0.46930813044309616, + "step": 134 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.031135135135135134, + "grad_norm": 34.65803146362305, + "learning_rate": 6.750000000000001e-06, + "loss": 7.2480010986328125, + "loss_alignment": 0.323486328125, + "loss_alignment_w": 0.1617431640625, + "loss_sft": 0.2914247512817383, + "loss_total_v6": 0.4530000761151314, + "step": 135 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03136576576576577, + "grad_norm": 36.77183532714844, + "learning_rate": 6.800000000000001e-06, + "loss": 7.2597455978393555, + "loss_alignment": 0.317138671875, + "loss_alignment_w": 0.1585693359375, + "loss_sft": 0.2951189950108528, + "loss_total_v6": 0.4537341110408306, + "step": 136 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.031596396396396394, + "grad_norm": 29.987255096435547, + "learning_rate": 6.850000000000001e-06, + "loss": 7.860541343688965, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3324856199324131, + "loss_total_v6": 0.4912838414311409, + "step": 137 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03182702702702703, + "grad_norm": 31.184953689575195, + "learning_rate": 6.9e-06, + "loss": 6.969273567199707, + "loss_alignment": 0.317626953125, + "loss_alignment_w": 0.1588134765625, + "loss_sft": 0.27679665945470333, + "loss_total_v6": 0.4355796203017235, + "step": 138 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03205765765765766, + "grad_norm": 29.05364990234375, + "learning_rate": 6.95e-06, + "loss": 7.678223609924316, + "loss_alignment": 0.32177734375, + "loss_alignment_w": 0.160888671875, + "loss_sft": 0.31938178837299347, + "loss_total_v6": 0.4798889793455601, + "step": 139 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03228828828828829, + "grad_norm": 32.77543258666992, + "learning_rate": 7e-06, + "loss": 7.397252559661865, + "loss_alignment": 0.316650390625, + "loss_alignment_w": 0.1583251953125, + "loss_sft": 0.3039420507848263, + "loss_total_v6": 0.4623282924294472, + "step": 140 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03251891891891892, + "grad_norm": 33.798545837402344, + "learning_rate": 7.05e-06, + "loss": 7.61773157119751, + "loss_alignment": 0.31591796875, + "loss_alignment_w": 0.157958984375, + "loss_sft": 0.31872908398509026, + "loss_total_v6": 0.47610823065042496, + "step": 141 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03274954954954955, + "grad_norm": 33.989444732666016, + "learning_rate": 7.100000000000001e-06, + "loss": 7.457625389099121, + "loss_alignment": 0.32275390625, + "loss_alignment_w": 0.161376953125, + "loss_sft": 0.30498402565717697, + "loss_total_v6": 0.46610157936811447, + "step": 142 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03298018018018018, + "grad_norm": 32.320037841796875, + "learning_rate": 7.15e-06, + "loss": 7.274205207824707, + "loss_alignment": 0.306396484375, + "loss_alignment_w": 0.1531982421875, + "loss_sft": 0.3015616685152054, + "loss_total_v6": 0.4546378441154957, + "step": 143 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03321081081081081, + "grad_norm": 34.55356216430664, + "learning_rate": 7.2000000000000005e-06, + "loss": 7.63621711730957, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.31719884648919106, + "loss_total_v6": 0.47726357355713844, + "step": 144 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03344144144144144, + "grad_norm": 28.578956604003906, + "learning_rate": 7.25e-06, + "loss": 7.305872917175293, + "loss_alignment": 0.316162109375, + "loss_alignment_w": 0.1580810546875, + "loss_sft": 0.2984444461762905, + "loss_total_v6": 0.4566170573234558, + "step": 145 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.033672072072072075, + "grad_norm": 26.3967227935791, + "learning_rate": 7.3e-06, + "loss": 7.857283592224121, + "loss_alignment": 0.31591796875, + "loss_alignment_w": 0.157958984375, + "loss_sft": 0.33359428122639656, + "loss_total_v6": 0.49108022823929787, + "step": 146 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0339027027027027, + "grad_norm": 30.663942337036133, + "learning_rate": 7.350000000000001e-06, + "loss": 7.162957668304443, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.287665918469429, + "loss_total_v6": 0.4476848542690277, + "step": 147 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.034133333333333335, + "grad_norm": 29.24234390258789, + "learning_rate": 7.4e-06, + "loss": 7.569999694824219, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.31524230912327766, + "loss_total_v6": 0.47312499210238457, + "step": 148 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03436396396396396, + "grad_norm": 29.2116756439209, + "learning_rate": 7.450000000000001e-06, + "loss": 7.354342460632324, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3014585245400667, + "loss_total_v6": 0.45964639261364937, + "step": 149 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.034594594594594595, + "grad_norm": 28.529033660888672, + "learning_rate": 7.500000000000001e-06, + "loss": 7.288630485534668, + "loss_alignment": 0.314208984375, + "loss_alignment_w": 0.1571044921875, + "loss_sft": 0.29872485250234604, + "loss_total_v6": 0.45553942397236824, + "step": 150 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03482522522522523, + "grad_norm": 26.045124053955078, + "learning_rate": 7.5500000000000006e-06, + "loss": 7.948911190032959, + "loss_alignment": 0.317138671875, + "loss_alignment_w": 0.1585693359375, + "loss_sft": 0.33813079819083214, + "loss_total_v6": 0.49680694565176964, + "step": 151 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.035055855855855855, + "grad_norm": 23.772342681884766, + "learning_rate": 7.600000000000001e-06, + "loss": 7.082908630371094, + "loss_alignment": 0.311767578125, + "loss_alignment_w": 0.1558837890625, + "loss_sft": 0.2867217157036066, + "loss_total_v6": 0.44268180429935455, + "step": 152 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03528648648648649, + "grad_norm": 30.03411865234375, + "learning_rate": 7.650000000000001e-06, + "loss": 7.308741092681885, + "loss_alignment": 0.31103515625, + "loss_alignment_w": 0.155517578125, + "loss_sft": 0.301187165081501, + "loss_total_v6": 0.4567963071167469, + "step": 153 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.035517117117117115, + "grad_norm": 33.46802520751953, + "learning_rate": 7.7e-06, + "loss": 7.365728855133057, + "loss_alignment": 0.31494140625, + "loss_alignment_w": 0.157470703125, + "loss_sft": 0.30258217826485634, + "loss_total_v6": 0.46035805344581604, + "step": 154 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03574774774774775, + "grad_norm": 29.788820266723633, + "learning_rate": 7.75e-06, + "loss": 7.638689994812012, + "loss_alignment": 0.318603515625, + "loss_alignment_w": 0.1593017578125, + "loss_sft": 0.3180706240236759, + "loss_total_v6": 0.4774181470274925, + "step": 155 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.035978378378378376, + "grad_norm": 30.30438804626465, + "learning_rate": 7.800000000000002e-06, + "loss": 7.308370590209961, + "loss_alignment": 0.31494140625, + "loss_alignment_w": 0.157470703125, + "loss_sft": 0.29937876760959625, + "loss_total_v6": 0.45677316933870316, + "step": 156 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03620900900900901, + "grad_norm": 30.54151725769043, + "learning_rate": 7.850000000000001e-06, + "loss": 7.041013717651367, + "loss_alignment": 0.309814453125, + "loss_alignment_w": 0.1549072265625, + "loss_sft": 0.2850035522133112, + "loss_total_v6": 0.44006337225437164, + "step": 157 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03643963963963964, + "grad_norm": 31.70608901977539, + "learning_rate": 7.9e-06, + "loss": 7.432353496551514, + "loss_alignment": 0.318603515625, + "loss_alignment_w": 0.1593017578125, + "loss_sft": 0.3054034598171711, + "loss_total_v6": 0.4645221009850502, + "step": 158 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03667027027027027, + "grad_norm": 31.376832962036133, + "learning_rate": 7.950000000000002e-06, + "loss": 7.421268463134766, + "loss_alignment": 0.314697265625, + "loss_alignment_w": 0.1573486328125, + "loss_sft": 0.30667901784181595, + "loss_total_v6": 0.46382927894592285, + "step": 159 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0369009009009009, + "grad_norm": 31.73944091796875, + "learning_rate": 8.000000000000001e-06, + "loss": 6.4541335105896, + "loss_alignment": 0.31005859375, + "loss_alignment_w": 0.155029296875, + "loss_sft": 0.24868973903357983, + "loss_total_v6": 0.4033833369612694, + "step": 160 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03713153153153153, + "grad_norm": 31.369504928588867, + "learning_rate": 8.050000000000001e-06, + "loss": 7.17879581451416, + "loss_alignment": 0.314697265625, + "loss_alignment_w": 0.1573486328125, + "loss_sft": 0.291432898491621, + "loss_total_v6": 0.4486747235059738, + "step": 161 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03736216216216216, + "grad_norm": 33.1863899230957, + "learning_rate": 8.1e-06, + "loss": 7.509721279144287, + "loss_alignment": 0.31982421875, + "loss_alignment_w": 0.159912109375, + "loss_sft": 0.30959805473685265, + "loss_total_v6": 0.46935758367180824, + "step": 162 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03759279279279279, + "grad_norm": 31.588104248046875, + "learning_rate": 8.15e-06, + "loss": 7.139142036437988, + "loss_alignment": 0.30615234375, + "loss_alignment_w": 0.153076171875, + "loss_sft": 0.2933643553406, + "loss_total_v6": 0.44619638472795486, + "step": 163 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03782342342342342, + "grad_norm": 30.754261016845703, + "learning_rate": 8.2e-06, + "loss": 7.538918495178223, + "loss_alignment": 0.31201171875, + "loss_alignment_w": 0.156005859375, + "loss_sft": 0.3152986355125904, + "loss_total_v6": 0.4711824283003807, + "step": 164 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03805405405405406, + "grad_norm": 32.414669036865234, + "learning_rate": 8.25e-06, + "loss": 6.913202285766602, + "loss_alignment": 0.314697265625, + "loss_alignment_w": 0.1573486328125, + "loss_sft": 0.27445184253156185, + "loss_total_v6": 0.4320751316845417, + "step": 165 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03828468468468468, + "grad_norm": 28.56916046142578, + "learning_rate": 8.3e-06, + "loss": 7.624694347381592, + "loss_alignment": 0.3134765625, + "loss_alignment_w": 0.15673828125, + "loss_sft": 0.3200187683105469, + "loss_total_v6": 0.4765434227883816, + "step": 166 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03851531531531532, + "grad_norm": 27.79007911682129, + "learning_rate": 8.35e-06, + "loss": 7.418807506561279, + "loss_alignment": 0.307861328125, + "loss_alignment_w": 0.1539306640625, + "loss_sft": 0.30956171825528145, + "loss_total_v6": 0.46367548778653145, + "step": 167 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.038745945945945943, + "grad_norm": 25.89882469177246, + "learning_rate": 8.400000000000001e-06, + "loss": 7.807316780090332, + "loss_alignment": 0.319091796875, + "loss_alignment_w": 0.1595458984375, + "loss_sft": 0.32862501963973045, + "loss_total_v6": 0.48795731365680695, + "step": 168 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03897657657657658, + "grad_norm": 29.86356544494629, + "learning_rate": 8.45e-06, + "loss": 7.65125846862793, + "loss_alignment": 0.310791015625, + "loss_alignment_w": 0.1553955078125, + "loss_sft": 0.3232049234211445, + "loss_total_v6": 0.4782036803662777, + "step": 169 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03920720720720721, + "grad_norm": 25.74285316467285, + "learning_rate": 8.5e-06, + "loss": 7.770871639251709, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.32946000806987286, + "loss_total_v6": 0.4856794811785221, + "step": 170 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03943783783783784, + "grad_norm": 28.03263282775879, + "learning_rate": 8.550000000000001e-06, + "loss": 7.621213912963867, + "loss_alignment": 0.30712890625, + "loss_alignment_w": 0.153564453125, + "loss_sft": 0.3228377178311348, + "loss_total_v6": 0.4763258770108223, + "step": 171 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.03966846846846847, + "grad_norm": 28.146636962890625, + "learning_rate": 8.6e-06, + "loss": 7.458940505981445, + "loss_alignment": 0.310302734375, + "loss_alignment_w": 0.1551513671875, + "loss_sft": 0.310879847034812, + "loss_total_v6": 0.4661837965250015, + "step": 172 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0398990990990991, + "grad_norm": 32.91970443725586, + "learning_rate": 8.65e-06, + "loss": 6.95399284362793, + "loss_alignment": 0.2979736328125, + "loss_alignment_w": 0.14898681640625, + "loss_sft": 0.28565300814807415, + "loss_total_v6": 0.4346245788037777, + "step": 173 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04012972972972973, + "grad_norm": 27.698162078857422, + "learning_rate": 8.700000000000001e-06, + "loss": 7.076379776000977, + "loss_alignment": 0.310546875, + "loss_alignment_w": 0.1552734375, + "loss_sft": 0.28713764250278473, + "loss_total_v6": 0.4422737620770931, + "step": 174 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04036036036036036, + "grad_norm": 26.95114517211914, + "learning_rate": 8.750000000000001e-06, + "loss": 7.470402717590332, + "loss_alignment": 0.31494140625, + "loss_alignment_w": 0.157470703125, + "loss_sft": 0.30950575321912766, + "loss_total_v6": 0.46690016612410545, + "step": 175 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04059099099099099, + "grad_norm": 30.659481048583984, + "learning_rate": 8.8e-06, + "loss": 7.333469390869141, + "loss_alignment": 0.313232421875, + "loss_alignment_w": 0.1566162109375, + "loss_sft": 0.30140516720712185, + "loss_total_v6": 0.4583418294787407, + "step": 176 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.040821621621621625, + "grad_norm": 33.22868728637695, + "learning_rate": 8.85e-06, + "loss": 7.178132057189941, + "loss_alignment": 0.314697265625, + "loss_alignment_w": 0.1573486328125, + "loss_sft": 0.2911778409034014, + "loss_total_v6": 0.44863327220082283, + "step": 177 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04105225225225225, + "grad_norm": 30.135486602783203, + "learning_rate": 8.900000000000001e-06, + "loss": 7.2712554931640625, + "loss_alignment": 0.309326171875, + "loss_alignment_w": 0.1546630859375, + "loss_sft": 0.299378402531147, + "loss_total_v6": 0.4544534645974636, + "step": 178 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.041282882882882885, + "grad_norm": 31.45252799987793, + "learning_rate": 8.95e-06, + "loss": 7.409850120544434, + "loss_alignment": 0.31005859375, + "loss_alignment_w": 0.155029296875, + "loss_sft": 0.3082694448530674, + "loss_total_v6": 0.4631156399846077, + "step": 179 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04151351351351351, + "grad_norm": 32.01137161254883, + "learning_rate": 9e-06, + "loss": 7.4261698722839355, + "loss_alignment": 0.31689453125, + "loss_alignment_w": 0.158447265625, + "loss_sft": 0.3057188745588064, + "loss_total_v6": 0.46413562074303627, + "step": 180 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.041744144144144145, + "grad_norm": 25.486154556274414, + "learning_rate": 9.050000000000001e-06, + "loss": 7.30325984954834, + "loss_alignment": 0.306640625, + "loss_alignment_w": 0.1533203125, + "loss_sft": 0.3027977403253317, + "loss_total_v6": 0.45645374059677124, + "step": 181 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04197477477477477, + "grad_norm": 27.693099975585938, + "learning_rate": 9.100000000000001e-06, + "loss": 7.325676918029785, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3034511059522629, + "loss_total_v6": 0.4578547850251198, + "step": 182 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.042205405405405405, + "grad_norm": 25.135229110717773, + "learning_rate": 9.15e-06, + "loss": 7.451719284057617, + "loss_alignment": 0.3037109375, + "loss_alignment_w": 0.15185546875, + "loss_sft": 0.31380072236061096, + "loss_total_v6": 0.46573247760534286, + "step": 183 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04243603603603604, + "grad_norm": 25.59937858581543, + "learning_rate": 9.200000000000002e-06, + "loss": 7.355603218078613, + "loss_alignment": 0.306396484375, + "loss_alignment_w": 0.1531982421875, + "loss_sft": 0.30623701959848404, + "loss_total_v6": 0.45972519367933273, + "step": 184 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.042666666666666665, + "grad_norm": 30.7474308013916, + "learning_rate": 9.250000000000001e-06, + "loss": 7.105137825012207, + "loss_alignment": 0.3076171875, + "loss_alignment_w": 0.15380859375, + "loss_sft": 0.29047616571187973, + "loss_total_v6": 0.44407112896442413, + "step": 185 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0428972972972973, + "grad_norm": 29.52057456970215, + "learning_rate": 9.3e-06, + "loss": 7.109696865081787, + "loss_alignment": 0.30615234375, + "loss_alignment_w": 0.153076171875, + "loss_sft": 0.29158506728708744, + "loss_total_v6": 0.4443560428917408, + "step": 186 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.043127927927927925, + "grad_norm": 32.030616760253906, + "learning_rate": 9.350000000000002e-06, + "loss": 7.281999588012695, + "loss_alignment": 0.302978515625, + "loss_alignment_w": 0.1514892578125, + "loss_sft": 0.3036357294768095, + "loss_total_v6": 0.45512499287724495, + "step": 187 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04335855855855856, + "grad_norm": 24.24785804748535, + "learning_rate": 9.4e-06, + "loss": 7.245304107666016, + "loss_alignment": 0.31103515625, + "loss_alignment_w": 0.155517578125, + "loss_sft": 0.29783271066844463, + "loss_total_v6": 0.4528314881026745, + "step": 188 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04358918918918919, + "grad_norm": 27.87604522705078, + "learning_rate": 9.450000000000001e-06, + "loss": 7.152040004730225, + "loss_alignment": 0.304931640625, + "loss_alignment_w": 0.1524658203125, + "loss_sft": 0.29459771513938904, + "loss_total_v6": 0.44700250774621964, + "step": 189 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04381981981981982, + "grad_norm": 28.993314743041992, + "learning_rate": 9.5e-06, + "loss": 7.393718719482422, + "loss_alignment": 0.303955078125, + "loss_alignment_w": 0.1519775390625, + "loss_sft": 0.3099467661231756, + "loss_total_v6": 0.4621074050664902, + "step": 190 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04405045045045045, + "grad_norm": 28.30815887451172, + "learning_rate": 9.55e-06, + "loss": 7.356443881988525, + "loss_alignment": 0.310302734375, + "loss_alignment_w": 0.1551513671875, + "loss_sft": 0.3046416398137808, + "loss_total_v6": 0.45977773889899254, + "step": 191 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04428108108108108, + "grad_norm": 30.044450759887695, + "learning_rate": 9.600000000000001e-06, + "loss": 7.347110748291016, + "loss_alignment": 0.298828125, + "loss_alignment_w": 0.1494140625, + "loss_sft": 0.31011606380343437, + "loss_total_v6": 0.45919443294405937, + "step": 192 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04451171171171171, + "grad_norm": 27.70938491821289, + "learning_rate": 9.65e-06, + "loss": 7.116272449493408, + "loss_alignment": 0.307373046875, + "loss_alignment_w": 0.1536865234375, + "loss_sft": 0.2907753437757492, + "loss_total_v6": 0.4447670355439186, + "step": 193 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04474234234234234, + "grad_norm": 27.57965850830078, + "learning_rate": 9.7e-06, + "loss": 7.146503448486328, + "loss_alignment": 0.305908203125, + "loss_alignment_w": 0.1529541015625, + "loss_sft": 0.2938702292740345, + "loss_total_v6": 0.4466564655303955, + "step": 194 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04497297297297297, + "grad_norm": 27.337099075317383, + "learning_rate": 9.75e-06, + "loss": 7.476436614990234, + "loss_alignment": 0.299072265625, + "loss_alignment_w": 0.1495361328125, + "loss_sft": 0.31790900602936745, + "loss_total_v6": 0.46727727726101875, + "step": 195 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.045203603603603607, + "grad_norm": 29.834226608276367, + "learning_rate": 9.800000000000001e-06, + "loss": 7.650789737701416, + "loss_alignment": 0.30322265625, + "loss_alignment_w": 0.151611328125, + "loss_sft": 0.3262426182627678, + "loss_total_v6": 0.47817437723279, + "step": 196 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04543423423423423, + "grad_norm": 29.195938110351562, + "learning_rate": 9.85e-06, + "loss": 7.414488792419434, + "loss_alignment": 0.315185546875, + "loss_alignment_w": 0.1575927734375, + "loss_sft": 0.30596533231437206, + "loss_total_v6": 0.4634055271744728, + "step": 197 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04566486486486487, + "grad_norm": 28.569303512573242, + "learning_rate": 9.9e-06, + "loss": 7.08899450302124, + "loss_alignment": 0.30126953125, + "loss_alignment_w": 0.150634765625, + "loss_sft": 0.2926715426146984, + "loss_total_v6": 0.4430621527135372, + "step": 198 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04589549549549549, + "grad_norm": 31.132104873657227, + "learning_rate": 9.950000000000001e-06, + "loss": 7.319270133972168, + "loss_alignment": 0.308837890625, + "loss_alignment_w": 0.1544189453125, + "loss_sft": 0.3031117469072342, + "loss_total_v6": 0.4574543945491314, + "step": 199 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04612612612612613, + "grad_norm": 28.55329132080078, + "learning_rate": 1e-05, + "loss": 7.9708251953125, + "loss_alignment": 0.30224609375, + "loss_alignment_w": 0.151123046875, + "loss_sft": 0.34685521200299263, + "loss_total_v6": 0.49817660823464394, + "step": 200 + }, + { + "epoch": 0.04612612612612613, + "eval_loss": 0.5997616052627563, + "eval_loss_alignment": 0.2759591716609589, + "eval_loss_alignment_w": 0.13797958583047945, + "eval_loss_sft": 0.4617820562003818, + "eval_loss_total_v6": 0.5997616418522514, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.206, + "eval_samples_per_second": 19.133, + "eval_steps_per_second": 2.401, + "step": 200 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04635675675675675, + "grad_norm": 34.7057991027832, + "learning_rate": 9.999998557623385e-06, + "loss": 7.7089972496032715, + "loss_alignment": 0.30810546875, + "loss_alignment_w": 0.154052734375, + "loss_sft": 0.32812582328915596, + "loss_total_v6": 0.48181233182549477, + "step": 201 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04658738738738739, + "grad_norm": 29.15117835998535, + "learning_rate": 9.999994230494368e-06, + "loss": 7.405484676361084, + "loss_alignment": 0.300048828125, + "loss_alignment_w": 0.1500244140625, + "loss_sft": 0.31257423385977745, + "loss_total_v6": 0.46284278109669685, + "step": 202 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04681801801801802, + "grad_norm": 27.981767654418945, + "learning_rate": 9.999987018615447e-06, + "loss": 7.69442081451416, + "loss_alignment": 0.305908203125, + "loss_alignment_w": 0.1529541015625, + "loss_sft": 0.3283286914229393, + "loss_total_v6": 0.4809013120830059, + "step": 203 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04704864864864865, + "grad_norm": 27.707916259765625, + "learning_rate": 9.999976921990784e-06, + "loss": 7.047355651855469, + "loss_alignment": 0.30810546875, + "loss_alignment_w": 0.154052734375, + "loss_sft": 0.2864833027124405, + "loss_total_v6": 0.4404597394168377, + "step": 204 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04727927927927928, + "grad_norm": 29.175647735595703, + "learning_rate": 9.999963940626203e-06, + "loss": 8.027087211608887, + "loss_alignment": 0.307861328125, + "loss_alignment_w": 0.1539306640625, + "loss_sft": 0.3473197966814041, + "loss_total_v6": 0.5016929768025875, + "step": 205 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04750990990990991, + "grad_norm": 26.932531356811523, + "learning_rate": 9.999948074529194e-06, + "loss": 7.274520397186279, + "loss_alignment": 0.303466796875, + "loss_alignment_w": 0.1517333984375, + "loss_sft": 0.3028478380292654, + "loss_total_v6": 0.45465752109885216, + "step": 206 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04774054054054054, + "grad_norm": 26.39837646484375, + "learning_rate": 9.999929323708913e-06, + "loss": 7.091627597808838, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2930344343185425, + "loss_total_v6": 0.4432267136871815, + "step": 207 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.047971171171171174, + "grad_norm": 32.25065231323242, + "learning_rate": 9.999907688176173e-06, + "loss": 7.509570121765137, + "loss_alignment": 0.300048828125, + "loss_alignment_w": 0.1500244140625, + "loss_sft": 0.31935423985123634, + "loss_total_v6": 0.46934812143445015, + "step": 208 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0482018018018018, + "grad_norm": 30.73077964782715, + "learning_rate": 9.999883167943461e-06, + "loss": 7.701542854309082, + "loss_alignment": 0.3017578125, + "loss_alignment_w": 0.15087890625, + "loss_sft": 0.33060482889413834, + "loss_total_v6": 0.48134641349315643, + "step": 209 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.048432432432432435, + "grad_norm": 30.035545349121094, + "learning_rate": 9.999855763024923e-06, + "loss": 7.112581253051758, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2940083686262369, + "loss_total_v6": 0.44453631341457367, + "step": 210 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04866306306306306, + "grad_norm": 29.198650360107422, + "learning_rate": 9.99982547343637e-06, + "loss": 7.265188694000244, + "loss_alignment": 0.300048828125, + "loss_alignment_w": 0.1500244140625, + "loss_sft": 0.30395832285284996, + "loss_total_v6": 0.45407429337501526, + "step": 211 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.048893693693693695, + "grad_norm": 30.438865661621094, + "learning_rate": 9.999792299195278e-06, + "loss": 7.791704177856445, + "loss_alignment": 0.30029296875, + "loss_alignment_w": 0.150146484375, + "loss_sft": 0.33662140369415283, + "loss_total_v6": 0.48698149621486664, + "step": 212 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04912432432432432, + "grad_norm": 27.66889190673828, + "learning_rate": 9.999756240320786e-06, + "loss": 7.20981502532959, + "loss_alignment": 0.304931640625, + "loss_alignment_w": 0.1524658203125, + "loss_sft": 0.2984680403023958, + "loss_total_v6": 0.45061342790722847, + "step": 213 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.049354954954954955, + "grad_norm": 42.15926742553711, + "learning_rate": 9.9997172968337e-06, + "loss": 7.20876407623291, + "loss_alignment": 0.301025390625, + "loss_alignment_w": 0.1505126953125, + "loss_sft": 0.3001876473426819, + "loss_total_v6": 0.4505477547645569, + "step": 214 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.04958558558558559, + "grad_norm": 30.912364959716797, + "learning_rate": 9.999675468756485e-06, + "loss": 7.277464866638184, + "loss_alignment": 0.302734375, + "loss_alignment_w": 0.1513671875, + "loss_sft": 0.30321498587727547, + "loss_total_v6": 0.45484157651662827, + "step": 215 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.049816216216216215, + "grad_norm": 29.352874755859375, + "learning_rate": 9.999630756113278e-06, + "loss": 7.55600643157959, + "loss_alignment": 0.304931640625, + "loss_alignment_w": 0.1524658203125, + "loss_sft": 0.32007449120283127, + "loss_total_v6": 0.47225040197372437, + "step": 216 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05004684684684685, + "grad_norm": 27.174158096313477, + "learning_rate": 9.999583158929873e-06, + "loss": 7.073088645935059, + "loss_alignment": 0.30419921875, + "loss_alignment_w": 0.152099609375, + "loss_sft": 0.29048724845051765, + "loss_total_v6": 0.44206804782152176, + "step": 217 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.050277477477477475, + "grad_norm": 34.13699722290039, + "learning_rate": 9.999532677233733e-06, + "loss": 7.177524089813232, + "loss_alignment": 0.29638671875, + "loss_alignment_w": 0.148193359375, + "loss_sft": 0.30046289414167404, + "loss_total_v6": 0.44859524071216583, + "step": 218 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05050810810810811, + "grad_norm": 37.24513244628906, + "learning_rate": 9.999479311053982e-06, + "loss": 7.259710788726807, + "loss_alignment": 0.306884765625, + "loss_alignment_w": 0.1534423828125, + "loss_sft": 0.2998928092420101, + "loss_total_v6": 0.4537319242954254, + "step": 219 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.050738738738738735, + "grad_norm": 30.127105712890625, + "learning_rate": 9.999423060421412e-06, + "loss": 7.28248929977417, + "loss_alignment": 0.294189453125, + "loss_alignment_w": 0.1470947265625, + "loss_sft": 0.3073894567787647, + "loss_total_v6": 0.4551555812358856, + "step": 220 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05096936936936937, + "grad_norm": 31.26276397705078, + "learning_rate": 9.999363925368472e-06, + "loss": 6.829745292663574, + "loss_alignment": 0.299072265625, + "loss_alignment_w": 0.1495361328125, + "loss_sft": 0.27742975763976574, + "loss_total_v6": 0.4268590807914734, + "step": 221 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0512, + "grad_norm": 35.68559646606445, + "learning_rate": 9.999301905929286e-06, + "loss": 7.34316349029541, + "loss_alignment": 0.298095703125, + "loss_alignment_w": 0.1490478515625, + "loss_sft": 0.3101287502795458, + "loss_total_v6": 0.45894771441817284, + "step": 222 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05143063063063063, + "grad_norm": 29.03331756591797, + "learning_rate": 9.999237002139633e-06, + "loss": 7.81610107421875, + "loss_alignment": 0.294921875, + "loss_alignment_w": 0.1474609375, + "loss_sft": 0.340938538312912, + "loss_total_v6": 0.4885062985122204, + "step": 223 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05166126126126126, + "grad_norm": 37.439517974853516, + "learning_rate": 9.999169214036958e-06, + "loss": 7.302586555480957, + "loss_alignment": 0.2958984375, + "loss_alignment_w": 0.14794921875, + "loss_sft": 0.3084013760089874, + "loss_total_v6": 0.456411637365818, + "step": 224 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05189189189189189, + "grad_norm": 33.84893035888672, + "learning_rate": 9.999098541660375e-06, + "loss": 7.319587707519531, + "loss_alignment": 0.2939453125, + "loss_alignment_w": 0.14697265625, + "loss_sft": 0.3106388971209526, + "loss_total_v6": 0.4574742279946804, + "step": 225 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05212252252252252, + "grad_norm": 33.2064208984375, + "learning_rate": 9.999024985050653e-06, + "loss": 7.405197620391846, + "loss_alignment": 0.296142578125, + "loss_alignment_w": 0.1480712890625, + "loss_sft": 0.31490612775087357, + "loss_total_v6": 0.46282484754920006, + "step": 226 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.052353153153153156, + "grad_norm": 31.013126373291016, + "learning_rate": 9.998948544250237e-06, + "loss": 7.254538059234619, + "loss_alignment": 0.294921875, + "loss_alignment_w": 0.1474609375, + "loss_sft": 0.30591716431081295, + "loss_total_v6": 0.4534086249768734, + "step": 227 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05258378378378378, + "grad_norm": 29.94978141784668, + "learning_rate": 9.998869219303227e-06, + "loss": 7.550850868225098, + "loss_alignment": 0.299072265625, + "loss_alignment_w": 0.1495361328125, + "loss_sft": 0.3225446231663227, + "loss_total_v6": 0.4719281606376171, + "step": 228 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.052814414414414416, + "grad_norm": 38.563812255859375, + "learning_rate": 9.998787010255388e-06, + "loss": 7.11374044418335, + "loss_alignment": 0.296630859375, + "loss_alignment_w": 0.1483154296875, + "loss_sft": 0.29629334434866905, + "loss_total_v6": 0.44460878148674965, + "step": 229 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05304504504504504, + "grad_norm": 25.73419189453125, + "learning_rate": 9.998701917154152e-06, + "loss": 6.705093860626221, + "loss_alignment": 0.298583984375, + "loss_alignment_w": 0.1492919921875, + "loss_sft": 0.26985269226133823, + "loss_total_v6": 0.4190683811903, + "step": 230 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05327567567567568, + "grad_norm": 31.546878814697266, + "learning_rate": 9.998613940048613e-06, + "loss": 7.249794006347656, + "loss_alignment": 0.2978515625, + "loss_alignment_w": 0.14892578125, + "loss_sft": 0.3038811609148979, + "loss_total_v6": 0.4531121291220188, + "step": 231 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0535063063063063, + "grad_norm": 28.254640579223633, + "learning_rate": 9.99852307898953e-06, + "loss": 7.053282260894775, + "loss_alignment": 0.297119140625, + "loss_alignment_w": 0.1485595703125, + "loss_sft": 0.2927893605083227, + "loss_total_v6": 0.44083013758063316, + "step": 232 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05373693693693694, + "grad_norm": 28.638641357421875, + "learning_rate": 9.998429334029323e-06, + "loss": 7.536474704742432, + "loss_alignment": 0.292236328125, + "loss_alignment_w": 0.1461181640625, + "loss_sft": 0.3247894421219826, + "loss_total_v6": 0.471029669046402, + "step": 233 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05396756756756757, + "grad_norm": 31.75786781311035, + "learning_rate": 9.998332705222083e-06, + "loss": 7.230053901672363, + "loss_alignment": 0.294189453125, + "loss_alignment_w": 0.1470947265625, + "loss_sft": 0.3043563812971115, + "loss_total_v6": 0.4518783688545227, + "step": 234 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0541981981981982, + "grad_norm": 30.58510971069336, + "learning_rate": 9.998233192623556e-06, + "loss": 7.5806708335876465, + "loss_alignment": 0.298095703125, + "loss_alignment_w": 0.1490478515625, + "loss_sft": 0.3248661458492279, + "loss_total_v6": 0.4737919270992279, + "step": 235 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05442882882882883, + "grad_norm": 32.434776306152344, + "learning_rate": 9.998130796291156e-06, + "loss": 7.386203765869141, + "loss_alignment": 0.297607421875, + "loss_alignment_w": 0.1488037109375, + "loss_sft": 0.3126509413123131, + "loss_total_v6": 0.4616377465426922, + "step": 236 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05465945945945946, + "grad_norm": 26.846961975097656, + "learning_rate": 9.998025516283964e-06, + "loss": 7.448126316070557, + "loss_alignment": 0.294677734375, + "loss_alignment_w": 0.1473388671875, + "loss_sft": 0.318352147936821, + "loss_total_v6": 0.4655078910291195, + "step": 237 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05489009009009009, + "grad_norm": 27.01883888244629, + "learning_rate": 9.997917352662718e-06, + "loss": 7.350298881530762, + "loss_alignment": 0.29248046875, + "loss_alignment_w": 0.146240234375, + "loss_sft": 0.3133517876267433, + "loss_total_v6": 0.4593936614692211, + "step": 238 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05512072072072072, + "grad_norm": 29.169675827026367, + "learning_rate": 9.997806305489826e-06, + "loss": 7.267745018005371, + "loss_alignment": 0.296142578125, + "loss_alignment_w": 0.1480712890625, + "loss_sft": 0.305949155241251, + "loss_total_v6": 0.4542340561747551, + "step": 239 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05535135135135135, + "grad_norm": 28.093467712402344, + "learning_rate": 9.997692374829352e-06, + "loss": 7.360121250152588, + "loss_alignment": 0.2978515625, + "loss_alignment_w": 0.14892578125, + "loss_sft": 0.31088341772556305, + "loss_total_v6": 0.46000758185982704, + "step": 240 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.055581981981981984, + "grad_norm": 30.678316116333008, + "learning_rate": 9.997575560747033e-06, + "loss": 7.616323947906494, + "loss_alignment": 0.2978515625, + "loss_alignment_w": 0.14892578125, + "loss_sft": 0.3268350660800934, + "loss_total_v6": 0.4760202504694462, + "step": 241 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05581261261261261, + "grad_norm": 32.65866470336914, + "learning_rate": 9.997455863310263e-06, + "loss": 7.3746337890625, + "loss_alignment": 0.29736328125, + "loss_alignment_w": 0.148681640625, + "loss_sft": 0.31238557770848274, + "loss_total_v6": 0.46091462671756744, + "step": 242 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.056043243243243245, + "grad_norm": 28.576406478881836, + "learning_rate": 9.997333282588103e-06, + "loss": 7.572992324829102, + "loss_alignment": 0.302978515625, + "loss_alignment_w": 0.1514892578125, + "loss_sft": 0.32167018204927444, + "loss_total_v6": 0.47331203892827034, + "step": 243 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05627387387387387, + "grad_norm": 29.55866050720215, + "learning_rate": 9.997207818651273e-06, + "loss": 6.986050605773926, + "loss_alignment": 0.29150390625, + "loss_alignment_w": 0.145751953125, + "loss_sft": 0.2907846812158823, + "loss_total_v6": 0.43662818893790245, + "step": 244 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.056504504504504505, + "grad_norm": 32.088497161865234, + "learning_rate": 9.997079471572163e-06, + "loss": 7.192256927490234, + "loss_alignment": 0.291748046875, + "loss_alignment_w": 0.1458740234375, + "loss_sft": 0.30384039133787155, + "loss_total_v6": 0.44951606169342995, + "step": 245 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05673513513513514, + "grad_norm": 30.75102996826172, + "learning_rate": 9.996948241424819e-06, + "loss": 7.082252025604248, + "loss_alignment": 0.290283203125, + "loss_alignment_w": 0.1451416015625, + "loss_sft": 0.2972550094127655, + "loss_total_v6": 0.4426407478749752, + "step": 246 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.056965765765765765, + "grad_norm": 28.661344528198242, + "learning_rate": 9.99681412828496e-06, + "loss": 7.459188461303711, + "loss_alignment": 0.29150390625, + "loss_alignment_w": 0.145751953125, + "loss_sft": 0.32052363082766533, + "loss_total_v6": 0.4661992937326431, + "step": 247 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0571963963963964, + "grad_norm": 28.96517562866211, + "learning_rate": 9.996677132229957e-06, + "loss": 7.686556339263916, + "loss_alignment": 0.2978515625, + "loss_alignment_w": 0.14892578125, + "loss_sft": 0.33146873116493225, + "loss_total_v6": 0.48040977120399475, + "step": 248 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.057427027027027025, + "grad_norm": 28.188745498657227, + "learning_rate": 9.996537253338852e-06, + "loss": 7.041421890258789, + "loss_alignment": 0.297607421875, + "loss_alignment_w": 0.1488037109375, + "loss_sft": 0.2911020256578922, + "loss_total_v6": 0.44008883833885193, + "step": 249 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05765765765765766, + "grad_norm": 25.169702529907227, + "learning_rate": 9.99639449169235e-06, + "loss": 7.214837551116943, + "loss_alignment": 0.29541015625, + "loss_alignment_w": 0.147705078125, + "loss_sft": 0.30302388593554497, + "loss_total_v6": 0.45092733576893806, + "step": 250 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.057888288288288285, + "grad_norm": 29.676700592041016, + "learning_rate": 9.996248847372813e-06, + "loss": 6.876317977905273, + "loss_alignment": 0.291015625, + "loss_alignment_w": 0.1455078125, + "loss_sft": 0.28398742340505123, + "loss_total_v6": 0.4297698959708214, + "step": 251 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05811891891891892, + "grad_norm": 26.95092010498047, + "learning_rate": 9.996100320464274e-06, + "loss": 7.07687520980835, + "loss_alignment": 0.2939453125, + "loss_alignment_w": 0.14697265625, + "loss_sft": 0.29549986869096756, + "loss_total_v6": 0.44230469688773155, + "step": 252 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05834954954954955, + "grad_norm": 28.532100677490234, + "learning_rate": 9.995948911052427e-06, + "loss": 7.496486663818359, + "loss_alignment": 0.29833984375, + "loss_alignment_w": 0.149169921875, + "loss_sft": 0.31940625607967377, + "loss_total_v6": 0.46853040903806686, + "step": 253 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05858018018018018, + "grad_norm": 27.71693229675293, + "learning_rate": 9.995794619224626e-06, + "loss": 7.7351298332214355, + "loss_alignment": 0.289794921875, + "loss_alignment_w": 0.1448974609375, + "loss_sft": 0.3381056897342205, + "loss_total_v6": 0.4834456332027912, + "step": 254 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05881081081081081, + "grad_norm": 32.071895599365234, + "learning_rate": 9.995637445069889e-06, + "loss": 7.267612457275391, + "loss_alignment": 0.291259765625, + "loss_alignment_w": 0.1456298828125, + "loss_sft": 0.30835171788930893, + "loss_total_v6": 0.4542257525026798, + "step": 255 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05904144144144144, + "grad_norm": 31.859342575073242, + "learning_rate": 9.995477388678898e-06, + "loss": 7.724972724914551, + "loss_alignment": 0.29931640625, + "loss_alignment_w": 0.149658203125, + "loss_sft": 0.333366222679615, + "loss_total_v6": 0.4828107990324497, + "step": 256 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05927207207207207, + "grad_norm": 29.657617568969727, + "learning_rate": 9.995314450143999e-06, + "loss": 7.063473224639893, + "loss_alignment": 0.29150390625, + "loss_alignment_w": 0.145751953125, + "loss_sft": 0.2956846058368683, + "loss_total_v6": 0.4414670690894127, + "step": 257 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.059502702702702706, + "grad_norm": 26.625463485717773, + "learning_rate": 9.995148629559196e-06, + "loss": 7.739208698272705, + "loss_alignment": 0.295654296875, + "loss_alignment_w": 0.1478271484375, + "loss_sft": 0.33639219030737877, + "loss_total_v6": 0.48370053619146347, + "step": 258 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.05973333333333333, + "grad_norm": 25.147598266601562, + "learning_rate": 9.994979927020163e-06, + "loss": 7.435618877410889, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.32002707198262215, + "loss_total_v6": 0.46472617983818054, + "step": 259 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.059963963963963966, + "grad_norm": 27.4565372467041, + "learning_rate": 9.994808342624234e-06, + "loss": 7.12758731842041, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.30339961871504784, + "loss_total_v6": 0.44547421112656593, + "step": 260 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06019459459459459, + "grad_norm": 25.826398849487305, + "learning_rate": 9.9946338764704e-06, + "loss": 7.41049861907959, + "loss_alignment": 0.294677734375, + "loss_alignment_w": 0.1473388671875, + "loss_sft": 0.3156952075660229, + "loss_total_v6": 0.46315614879131317, + "step": 261 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.060425225225225226, + "grad_norm": 26.988557815551758, + "learning_rate": 9.994456528659322e-06, + "loss": 7.395825386047363, + "loss_alignment": 0.288818359375, + "loss_alignment_w": 0.1444091796875, + "loss_sft": 0.31781468354165554, + "loss_total_v6": 0.4622391127049923, + "step": 262 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06065585585585585, + "grad_norm": 32.01319885253906, + "learning_rate": 9.994276299293321e-06, + "loss": 6.937900066375732, + "loss_alignment": 0.29052734375, + "loss_alignment_w": 0.145263671875, + "loss_sft": 0.28823300264775753, + "loss_total_v6": 0.433618750423193, + "step": 263 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06088648648648649, + "grad_norm": 29.930259704589844, + "learning_rate": 9.994093188476383e-06, + "loss": 7.786201000213623, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.33623168990015984, + "loss_total_v6": 0.48663756251335144, + "step": 264 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06111711711711712, + "grad_norm": 29.953153610229492, + "learning_rate": 9.993907196314148e-06, + "loss": 7.270598411560059, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.30977435782551765, + "loss_total_v6": 0.45441241562366486, + "step": 265 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06134774774774775, + "grad_norm": 26.516769409179688, + "learning_rate": 9.99371832291393e-06, + "loss": 7.317441940307617, + "loss_alignment": 0.2919921875, + "loss_alignment_w": 0.14599609375, + "loss_sft": 0.3111151345074177, + "loss_total_v6": 0.4573401063680649, + "step": 266 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06157837837837838, + "grad_norm": 27.30320167541504, + "learning_rate": 9.993526568384694e-06, + "loss": 7.311214447021484, + "loss_alignment": 0.291748046875, + "loss_alignment_w": 0.1458740234375, + "loss_sft": 0.31061914563179016, + "loss_total_v6": 0.45695091038942337, + "step": 267 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06180900900900901, + "grad_norm": 25.3011531829834, + "learning_rate": 9.99333193283708e-06, + "loss": 7.388277053833008, + "loss_alignment": 0.29052734375, + "loss_alignment_w": 0.145263671875, + "loss_sft": 0.3168088234961033, + "loss_total_v6": 0.4617673121392727, + "step": 268 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06203963963963964, + "grad_norm": 36.227962493896484, + "learning_rate": 9.993134416383376e-06, + "loss": 6.939764022827148, + "loss_alignment": 0.283935546875, + "loss_alignment_w": 0.1419677734375, + "loss_sft": 0.2915385812520981, + "loss_total_v6": 0.433735229074955, + "step": 269 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06227027027027027, + "grad_norm": 28.06507682800293, + "learning_rate": 9.992934019137544e-06, + "loss": 6.742907524108887, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.27967756800353527, + "loss_total_v6": 0.4214317090809345, + "step": 270 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06250090090090091, + "grad_norm": 30.188220977783203, + "learning_rate": 9.992730741215202e-06, + "loss": 7.476947784423828, + "loss_alignment": 0.29150390625, + "loss_alignment_w": 0.145751953125, + "loss_sft": 0.32169463112950325, + "loss_total_v6": 0.46730923280119896, + "step": 271 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06273153153153153, + "grad_norm": 26.76581573486328, + "learning_rate": 9.99252458273363e-06, + "loss": 7.18979549407959, + "loss_alignment": 0.29736328125, + "loss_alignment_w": 0.148681640625, + "loss_sft": 0.30068057402968407, + "loss_total_v6": 0.44936221465468407, + "step": 272 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06296216216216216, + "grad_norm": 25.703310012817383, + "learning_rate": 9.992315543811773e-06, + "loss": 7.145261287689209, + "loss_alignment": 0.291259765625, + "loss_alignment_w": 0.1456298828125, + "loss_sft": 0.30093368142843246, + "loss_total_v6": 0.44657883420586586, + "step": 273 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06319279279279279, + "grad_norm": 26.21283721923828, + "learning_rate": 9.992103624570234e-06, + "loss": 7.391260147094727, + "loss_alignment": 0.29541015625, + "loss_alignment_w": 0.147705078125, + "loss_sft": 0.3144012466073036, + "loss_total_v6": 0.4619537405669689, + "step": 274 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06342342342342343, + "grad_norm": 27.401752471923828, + "learning_rate": 9.991888825131283e-06, + "loss": 7.419635772705078, + "loss_alignment": 0.287841796875, + "loss_alignment_w": 0.1439208984375, + "loss_sft": 0.3197910860180855, + "loss_total_v6": 0.4637272357940674, + "step": 275 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06365405405405405, + "grad_norm": 25.876880645751953, + "learning_rate": 9.991671145618847e-06, + "loss": 7.3626790046691895, + "loss_alignment": 0.289306640625, + "loss_alignment_w": 0.1446533203125, + "loss_sft": 0.31578876450657845, + "loss_total_v6": 0.46016743406653404, + "step": 276 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06388468468468468, + "grad_norm": 29.063879013061523, + "learning_rate": 9.991450586158515e-06, + "loss": 7.421925067901611, + "loss_alignment": 0.29541015625, + "loss_alignment_w": 0.147705078125, + "loss_sft": 0.315905824303627, + "loss_total_v6": 0.4638703092932701, + "step": 277 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06411531531531532, + "grad_norm": 26.62662696838379, + "learning_rate": 9.991227146877542e-06, + "loss": 7.198575496673584, + "loss_alignment": 0.28564453125, + "loss_alignment_w": 0.142822265625, + "loss_sft": 0.3069666363298893, + "loss_total_v6": 0.4499109722673893, + "step": 278 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06434594594594595, + "grad_norm": 27.592899322509766, + "learning_rate": 9.991000827904839e-06, + "loss": 7.4305009841918945, + "loss_alignment": 0.28564453125, + "loss_alignment_w": 0.142822265625, + "loss_sft": 0.3215229883790016, + "loss_total_v6": 0.4644063077867031, + "step": 279 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06457657657657657, + "grad_norm": 25.987462997436523, + "learning_rate": 9.99077162937098e-06, + "loss": 7.476068496704102, + "loss_alignment": 0.28857421875, + "loss_alignment_w": 0.144287109375, + "loss_sft": 0.32270780205726624, + "loss_total_v6": 0.46725429594516754, + "step": 280 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0648072072072072, + "grad_norm": 26.38298988342285, + "learning_rate": 9.990539551408207e-06, + "loss": 7.418963432312012, + "loss_alignment": 0.289306640625, + "loss_alignment_w": 0.1446533203125, + "loss_sft": 0.319153955206275, + "loss_total_v6": 0.46368519961833954, + "step": 281 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06503783783783784, + "grad_norm": 29.988853454589844, + "learning_rate": 9.990304594150411e-06, + "loss": 7.317194938659668, + "loss_alignment": 0.290771484375, + "loss_alignment_w": 0.1453857421875, + "loss_sft": 0.31190845370292664, + "loss_total_v6": 0.45732469111680984, + "step": 282 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06526846846846847, + "grad_norm": 26.786039352416992, + "learning_rate": 9.990066757733152e-06, + "loss": 7.522168159484863, + "loss_alignment": 0.289794921875, + "loss_alignment_w": 0.1448974609375, + "loss_sft": 0.3253753613680601, + "loss_total_v6": 0.47013548761606216, + "step": 283 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0654990990990991, + "grad_norm": 25.97243881225586, + "learning_rate": 9.989826042293653e-06, + "loss": 7.4283599853515625, + "loss_alignment": 0.294189453125, + "loss_alignment_w": 0.1470947265625, + "loss_sft": 0.31748295202851295, + "loss_total_v6": 0.46427249535918236, + "step": 284 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06572972972972974, + "grad_norm": 26.67731475830078, + "learning_rate": 9.989582447970792e-06, + "loss": 7.506103515625, + "loss_alignment": 0.2900390625, + "loss_alignment_w": 0.14501953125, + "loss_sft": 0.3240051306784153, + "loss_total_v6": 0.4691314660012722, + "step": 285 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06596036036036036, + "grad_norm": 27.75828742980957, + "learning_rate": 9.989335974905112e-06, + "loss": 7.565095901489258, + "loss_alignment": 0.294677734375, + "loss_alignment_w": 0.1473388671875, + "loss_sft": 0.3257390037178993, + "loss_total_v6": 0.4728184677660465, + "step": 286 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06619099099099099, + "grad_norm": 25.088685989379883, + "learning_rate": 9.989086623238815e-06, + "loss": 6.983644962310791, + "loss_alignment": 0.29248046875, + "loss_alignment_w": 0.146240234375, + "loss_sft": 0.2904512193053961, + "loss_total_v6": 0.43647781759500504, + "step": 287 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06642162162162162, + "grad_norm": 27.57876205444336, + "learning_rate": 9.988834393115768e-06, + "loss": 7.122591972351074, + "loss_alignment": 0.287353515625, + "loss_alignment_w": 0.1436767578125, + "loss_sft": 0.3015157449990511, + "loss_total_v6": 0.44516198709607124, + "step": 288 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06665225225225226, + "grad_norm": 28.369352340698242, + "learning_rate": 9.98857928468149e-06, + "loss": 6.93820858001709, + "loss_alignment": 0.290771484375, + "loss_alignment_w": 0.1453857421875, + "loss_sft": 0.2883285768330097, + "loss_total_v6": 0.4336380362510681, + "step": 289 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06688288288288288, + "grad_norm": 25.885549545288086, + "learning_rate": 9.988321298083167e-06, + "loss": 7.512146472930908, + "loss_alignment": 0.2939453125, + "loss_alignment_w": 0.14697265625, + "loss_sft": 0.32284168526530266, + "loss_total_v6": 0.46950916200876236, + "step": 290 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06711351351351351, + "grad_norm": 25.600906372070312, + "learning_rate": 9.98806043346965e-06, + "loss": 7.279273986816406, + "loss_alignment": 0.289306640625, + "loss_alignment_w": 0.1446533203125, + "loss_sft": 0.3103623278439045, + "loss_total_v6": 0.4549546130001545, + "step": 291 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06734414414414415, + "grad_norm": 27.10985565185547, + "learning_rate": 9.987796690991438e-06, + "loss": 7.6518988609313965, + "loss_alignment": 0.291015625, + "loss_alignment_w": 0.1455078125, + "loss_sft": 0.3326748386025429, + "loss_total_v6": 0.4782436825335026, + "step": 292 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06757477477477478, + "grad_norm": 25.509281158447266, + "learning_rate": 9.987530070800703e-06, + "loss": 7.400374889373779, + "loss_alignment": 0.28564453125, + "loss_alignment_w": 0.142822265625, + "loss_sft": 0.3198537640273571, + "loss_total_v6": 0.4625234268605709, + "step": 293 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0678054054054054, + "grad_norm": 27.710512161254883, + "learning_rate": 9.987260573051268e-06, + "loss": 7.082556247711182, + "loss_alignment": 0.28759765625, + "loss_alignment_w": 0.143798828125, + "loss_sft": 0.29857105016708374, + "loss_total_v6": 0.44265977293252945, + "step": 294 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06803603603603603, + "grad_norm": 29.639846801757812, + "learning_rate": 9.986988197898621e-06, + "loss": 7.568987846374512, + "loss_alignment": 0.29345703125, + "loss_alignment_w": 0.146728515625, + "loss_sft": 0.3263637199997902, + "loss_total_v6": 0.473061740398407, + "step": 295 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06826666666666667, + "grad_norm": 27.25076675415039, + "learning_rate": 9.98671294549991e-06, + "loss": 7.063677787780762, + "loss_alignment": 0.28564453125, + "loss_alignment_w": 0.142822265625, + "loss_sft": 0.2980625070631504, + "loss_total_v6": 0.4414798580110073, + "step": 296 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0684972972972973, + "grad_norm": 26.683088302612305, + "learning_rate": 9.986434816013941e-06, + "loss": 7.40207576751709, + "loss_alignment": 0.288818359375, + "loss_alignment_w": 0.1444091796875, + "loss_sft": 0.3181747794151306, + "loss_total_v6": 0.4626297205686569, + "step": 297 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06872792792792792, + "grad_norm": 26.332963943481445, + "learning_rate": 9.98615380960118e-06, + "loss": 7.296384811401367, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.31288131326436996, + "loss_total_v6": 0.45602403208613396, + "step": 298 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06895855855855856, + "grad_norm": 29.146087646484375, + "learning_rate": 9.985869926423757e-06, + "loss": 7.211080551147461, + "loss_alignment": 0.2861328125, + "loss_alignment_w": 0.14306640625, + "loss_sft": 0.30762615613639355, + "loss_total_v6": 0.4506925456225872, + "step": 299 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06918918918918919, + "grad_norm": 27.767061233520508, + "learning_rate": 9.985583166645455e-06, + "loss": 7.883679389953613, + "loss_alignment": 0.2880859375, + "loss_alignment_w": 0.14404296875, + "loss_sft": 0.34848862886428833, + "loss_total_v6": 0.49272996187210083, + "step": 300 + }, + { + "epoch": 0.06918918918918919, + "eval_loss": 0.6006011366844177, + "eval_loss_alignment": 0.2606463149257991, + "eval_loss_alignment_w": 0.13032315746289955, + "eval_loss_sft": 0.47027798860278575, + "eval_loss_total_v6": 0.6006011462442951, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 92.0643, + "eval_samples_per_second": 18.954, + "eval_steps_per_second": 2.379, + "step": 300 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06941981981981982, + "grad_norm": 31.661958694458008, + "learning_rate": 9.985293530431722e-06, + "loss": 7.014008045196533, + "loss_alignment": 0.28857421875, + "loss_alignment_w": 0.144287109375, + "loss_sft": 0.2942562401294708, + "loss_total_v6": 0.4383755102753639, + "step": 301 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06965045045045046, + "grad_norm": 25.297016143798828, + "learning_rate": 9.985001017949664e-06, + "loss": 7.175520896911621, + "loss_alignment": 0.288330078125, + "loss_alignment_w": 0.1441650390625, + "loss_sft": 0.3046559728682041, + "loss_total_v6": 0.4484700672328472, + "step": 302 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.06988108108108108, + "grad_norm": 26.726703643798828, + "learning_rate": 9.984705629368046e-06, + "loss": 7.251530647277832, + "loss_alignment": 0.283935546875, + "loss_alignment_w": 0.1419677734375, + "loss_sft": 0.31128340028226376, + "loss_total_v6": 0.4532206505537033, + "step": 303 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07011171171171171, + "grad_norm": 26.746166229248047, + "learning_rate": 9.984407364857292e-06, + "loss": 6.988336086273193, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.2949252836406231, + "loss_total_v6": 0.4367710016667843, + "step": 304 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07034234234234234, + "grad_norm": 26.74742317199707, + "learning_rate": 9.984106224589488e-06, + "loss": 7.422445774078369, + "loss_alignment": 0.287109375, + "loss_alignment_w": 0.1435546875, + "loss_sft": 0.32074491679668427, + "loss_total_v6": 0.46390287578105927, + "step": 305 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07057297297297298, + "grad_norm": 28.40764617919922, + "learning_rate": 9.983802208738373e-06, + "loss": 7.316166877746582, + "loss_alignment": 0.28857421875, + "loss_alignment_w": 0.144287109375, + "loss_sft": 0.3126071207225323, + "loss_total_v6": 0.4572604261338711, + "step": 306 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0708036036036036, + "grad_norm": 31.763179779052734, + "learning_rate": 9.983495317479353e-06, + "loss": 6.923244953155518, + "loss_alignment": 0.29052734375, + "loss_alignment_w": 0.145263671875, + "loss_sft": 0.28724078834056854, + "loss_total_v6": 0.43270280957221985, + "step": 307 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07103423423423423, + "grad_norm": 27.982694625854492, + "learning_rate": 9.983185550989488e-06, + "loss": 7.471534252166748, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.32527776807546616, + "loss_total_v6": 0.46697087958455086, + "step": 308 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07126486486486487, + "grad_norm": 28.73655128479004, + "learning_rate": 9.982872909447496e-06, + "loss": 7.1918840408325195, + "loss_alignment": 0.285400390625, + "loss_alignment_w": 0.1427001953125, + "loss_sft": 0.30677729472517967, + "loss_total_v6": 0.4494927413761616, + "step": 309 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0714954954954955, + "grad_norm": 25.615215301513672, + "learning_rate": 9.982557393033758e-06, + "loss": 7.320422649383545, + "loss_alignment": 0.2919921875, + "loss_alignment_w": 0.14599609375, + "loss_sft": 0.31162188202142715, + "loss_total_v6": 0.45752642676234245, + "step": 310 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07172612612612612, + "grad_norm": 24.62273406982422, + "learning_rate": 9.982239001930311e-06, + "loss": 7.766911029815674, + "loss_alignment": 0.298828125, + "loss_alignment_w": 0.1494140625, + "loss_sft": 0.3358195163309574, + "loss_total_v6": 0.4854319430887699, + "step": 311 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07195675675675675, + "grad_norm": 24.81925392150879, + "learning_rate": 9.981917736320852e-06, + "loss": 7.378236293792725, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.317905530333519, + "loss_total_v6": 0.4611397683620453, + "step": 312 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07218738738738739, + "grad_norm": 27.97311782836914, + "learning_rate": 9.981593596390733e-06, + "loss": 7.056636810302734, + "loss_alignment": 0.2919921875, + "loss_alignment_w": 0.14599609375, + "loss_sft": 0.29539466463029385, + "loss_total_v6": 0.4410397969186306, + "step": 313 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07241801801801802, + "grad_norm": 29.30668830871582, + "learning_rate": 9.98126658232697e-06, + "loss": 7.503156661987305, + "loss_alignment": 0.285888671875, + "loss_alignment_w": 0.1429443359375, + "loss_sft": 0.32636917755007744, + "loss_total_v6": 0.46894730255007744, + "step": 314 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07264864864864864, + "grad_norm": 28.011198043823242, + "learning_rate": 9.980936694318231e-06, + "loss": 7.273270130157471, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.3142900615930557, + "loss_total_v6": 0.4545793756842613, + "step": 315 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07287927927927929, + "grad_norm": 25.198633193969727, + "learning_rate": 9.980603932554844e-06, + "loss": 7.423611640930176, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.3227098770439625, + "loss_total_v6": 0.4639757350087166, + "step": 316 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07310990990990991, + "grad_norm": 30.75092315673828, + "learning_rate": 9.9802682972288e-06, + "loss": 7.32981014251709, + "loss_alignment": 0.291259765625, + "loss_alignment_w": 0.1456298828125, + "loss_sft": 0.3120407462120056, + "loss_total_v6": 0.4581131562590599, + "step": 317 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07334054054054054, + "grad_norm": 27.037919998168945, + "learning_rate": 9.979929788533742e-06, + "loss": 7.330204486846924, + "loss_alignment": 0.291748046875, + "loss_alignment_w": 0.1458740234375, + "loss_sft": 0.31205013394355774, + "loss_total_v6": 0.45813777297735214, + "step": 318 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07357117117117117, + "grad_norm": 25.489566802978516, + "learning_rate": 9.979588406664972e-06, + "loss": 6.8493146896362305, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.28646534867584705, + "loss_total_v6": 0.4280821681022644, + "step": 319 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0738018018018018, + "grad_norm": 27.547983169555664, + "learning_rate": 9.979244151819454e-06, + "loss": 7.568611145019531, + "loss_alignment": 0.2919921875, + "loss_alignment_w": 0.14599609375, + "loss_sft": 0.3270573541522026, + "loss_total_v6": 0.4730381965637207, + "step": 320 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07403243243243243, + "grad_norm": 26.150300979614258, + "learning_rate": 9.978897024195801e-06, + "loss": 7.416661739349365, + "loss_alignment": 0.2861328125, + "loss_alignment_w": 0.14306640625, + "loss_sft": 0.32061226293444633, + "loss_total_v6": 0.4635413624346256, + "step": 321 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07426306306306306, + "grad_norm": 27.392948150634766, + "learning_rate": 9.978547023994292e-06, + "loss": 7.321564674377441, + "loss_alignment": 0.2841796875, + "loss_alignment_w": 0.14208984375, + "loss_sft": 0.3154926784336567, + "loss_total_v6": 0.4575977921485901, + "step": 322 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0744936936936937, + "grad_norm": 26.5132999420166, + "learning_rate": 9.978194151416857e-06, + "loss": 7.000199794769287, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.29586514830589294, + "loss_total_v6": 0.43751249089837074, + "step": 323 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07472432432432433, + "grad_norm": 28.732629776000977, + "learning_rate": 9.977838406667091e-06, + "loss": 7.094005584716797, + "loss_alignment": 0.287353515625, + "loss_alignment_w": 0.1436767578125, + "loss_sft": 0.30009531415998936, + "loss_total_v6": 0.4433753564953804, + "step": 324 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07495495495495495, + "grad_norm": 24.38580322265625, + "learning_rate": 9.977479789950238e-06, + "loss": 7.420315265655518, + "loss_alignment": 0.293212890625, + "loss_alignment_w": 0.1466064453125, + "loss_sft": 0.3171022403985262, + "loss_total_v6": 0.46376971527934074, + "step": 325 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07518558558558558, + "grad_norm": 27.804033279418945, + "learning_rate": 9.977118301473199e-06, + "loss": 7.223193168640137, + "loss_alignment": 0.29248046875, + "loss_alignment_w": 0.146240234375, + "loss_sft": 0.30519408732652664, + "loss_total_v6": 0.45144957676529884, + "step": 326 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07541621621621622, + "grad_norm": 28.573699951171875, + "learning_rate": 9.976753941444541e-06, + "loss": 6.850793838500977, + "loss_alignment": 0.286865234375, + "loss_alignment_w": 0.1434326171875, + "loss_sft": 0.2847267556935549, + "loss_total_v6": 0.4281746372580528, + "step": 327 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07564684684684685, + "grad_norm": 26.688669204711914, + "learning_rate": 9.976386710074479e-06, + "loss": 7.516448974609375, + "loss_alignment": 0.292236328125, + "loss_alignment_w": 0.1461181640625, + "loss_sft": 0.32365988940000534, + "loss_total_v6": 0.46977806091308594, + "step": 328 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07587747747747747, + "grad_norm": 27.960527420043945, + "learning_rate": 9.976016607574886e-06, + "loss": 6.887320041656494, + "loss_alignment": 0.28466796875, + "loss_alignment_w": 0.142333984375, + "loss_sft": 0.28783358447253704, + "loss_total_v6": 0.43045749142766, + "step": 329 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07610810810810811, + "grad_norm": 26.159955978393555, + "learning_rate": 9.975643634159294e-06, + "loss": 7.013606071472168, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.29880874790251255, + "loss_total_v6": 0.4383503794670105, + "step": 330 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07633873873873874, + "grad_norm": 25.789901733398438, + "learning_rate": 9.975267790042892e-06, + "loss": 7.1801066398620605, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.3077044356614351, + "loss_total_v6": 0.4487566836178303, + "step": 331 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07656936936936937, + "grad_norm": 27.627466201782227, + "learning_rate": 9.97488907544252e-06, + "loss": 7.132245063781738, + "loss_alignment": 0.288330078125, + "loss_alignment_w": 0.1441650390625, + "loss_sft": 0.30173763260245323, + "loss_total_v6": 0.44576533511281013, + "step": 332 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0768, + "grad_norm": 26.792104721069336, + "learning_rate": 9.974507490576682e-06, + "loss": 6.888384819030762, + "loss_alignment": 0.28662109375, + "loss_alignment_w": 0.143310546875, + "loss_sft": 0.2871982604265213, + "loss_total_v6": 0.4305240660905838, + "step": 333 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07703063063063063, + "grad_norm": 24.836328506469727, + "learning_rate": 9.974123035665531e-06, + "loss": 7.100693702697754, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.3008947931230068, + "loss_total_v6": 0.4437933675944805, + "step": 334 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07726126126126126, + "grad_norm": 29.768312454223633, + "learning_rate": 9.973735710930878e-06, + "loss": 7.216776371002197, + "loss_alignment": 0.285400390625, + "loss_alignment_w": 0.1427001953125, + "loss_sft": 0.30843988060951233, + "loss_total_v6": 0.45104851946234703, + "step": 335 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07749189189189189, + "grad_norm": 28.10756492614746, + "learning_rate": 9.973345516596192e-06, + "loss": 7.235283374786377, + "loss_alignment": 0.284423828125, + "loss_alignment_w": 0.1422119140625, + "loss_sft": 0.31034425273537636, + "loss_total_v6": 0.45220520719885826, + "step": 336 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07772252252252253, + "grad_norm": 28.529661178588867, + "learning_rate": 9.972952452886594e-06, + "loss": 7.3849406242370605, + "loss_alignment": 0.28955078125, + "loss_alignment_w": 0.144775390625, + "loss_sft": 0.3167376220226288, + "loss_total_v6": 0.4615588001906872, + "step": 337 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07795315315315315, + "grad_norm": 26.256942749023438, + "learning_rate": 9.972556520028863e-06, + "loss": 7.258599281311035, + "loss_alignment": 0.287353515625, + "loss_alignment_w": 0.1436767578125, + "loss_sft": 0.3098788745701313, + "loss_total_v6": 0.4536624513566494, + "step": 338 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07818378378378378, + "grad_norm": 24.54588508605957, + "learning_rate": 9.972157718251433e-06, + "loss": 7.287245750427246, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.3120660465210676, + "loss_total_v6": 0.45545288175344467, + "step": 339 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07841441441441442, + "grad_norm": 27.85732078552246, + "learning_rate": 9.971756047784393e-06, + "loss": 7.145552635192871, + "loss_alignment": 0.288818359375, + "loss_alignment_w": 0.1444091796875, + "loss_sft": 0.30266087874770164, + "loss_total_v6": 0.44659703597426414, + "step": 340 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07864504504504505, + "grad_norm": 24.613122940063477, + "learning_rate": 9.971351508859488e-06, + "loss": 7.030911445617676, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.29487021267414093, + "loss_total_v6": 0.43943197652697563, + "step": 341 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07887567567567567, + "grad_norm": 23.786380767822266, + "learning_rate": 9.970944101710115e-06, + "loss": 7.174158096313477, + "loss_alignment": 0.287109375, + "loss_alignment_w": 0.1435546875, + "loss_sft": 0.3047691695392132, + "loss_total_v6": 0.4483848810195923, + "step": 342 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0791063063063063, + "grad_norm": 26.271366119384766, + "learning_rate": 9.970533826571329e-06, + "loss": 7.224339008331299, + "loss_alignment": 0.28759765625, + "loss_alignment_w": 0.143798828125, + "loss_sft": 0.30796649120748043, + "loss_total_v6": 0.4515211954712868, + "step": 343 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07933693693693694, + "grad_norm": 25.318002700805664, + "learning_rate": 9.970120683679837e-06, + "loss": 7.224977970123291, + "loss_alignment": 0.289306640625, + "loss_alignment_w": 0.1446533203125, + "loss_sft": 0.30645003728568554, + "loss_total_v6": 0.4515611305832863, + "step": 344 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.07956756756756757, + "grad_norm": 29.001176834106445, + "learning_rate": 9.969704673274006e-06, + "loss": 7.3386664390563965, + "loss_alignment": 0.286865234375, + "loss_alignment_w": 0.1434326171875, + "loss_sft": 0.315600261092186, + "loss_total_v6": 0.4586666524410248, + "step": 345 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0797981981981982, + "grad_norm": 29.746931076049805, + "learning_rate": 9.969285795593851e-06, + "loss": 7.435697555541992, + "loss_alignment": 0.291015625, + "loss_alignment_w": 0.1455078125, + "loss_sft": 0.3192538246512413, + "loss_total_v6": 0.4647311232984066, + "step": 346 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08002882882882884, + "grad_norm": 25.390949249267578, + "learning_rate": 9.96886405088104e-06, + "loss": 7.448290824890137, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.32428284734487534, + "loss_total_v6": 0.46551819518208504, + "step": 347 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08025945945945946, + "grad_norm": 27.892349243164062, + "learning_rate": 9.968439439378905e-06, + "loss": 7.504497051239014, + "loss_alignment": 0.292236328125, + "loss_alignment_w": 0.1461181640625, + "loss_sft": 0.32279082015156746, + "loss_total_v6": 0.46903105825185776, + "step": 348 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08049009009009009, + "grad_norm": 26.42626190185547, + "learning_rate": 9.968011961332424e-06, + "loss": 7.268771648406982, + "loss_alignment": 0.2861328125, + "loss_alignment_w": 0.14306640625, + "loss_sft": 0.3111097477376461, + "loss_total_v6": 0.4542982280254364, + "step": 349 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08072072072072072, + "grad_norm": 26.518856048583984, + "learning_rate": 9.967581616988229e-06, + "loss": 7.423744201660156, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.32299280911684036, + "loss_total_v6": 0.46398401260375977, + "step": 350 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08095135135135136, + "grad_norm": 23.02914047241211, + "learning_rate": 9.967148406594608e-06, + "loss": 7.603463172912598, + "loss_alignment": 0.28759765625, + "loss_alignment_w": 0.143798828125, + "loss_sft": 0.33190590888261795, + "loss_total_v6": 0.47521645948290825, + "step": 351 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08118198198198198, + "grad_norm": 24.085098266601562, + "learning_rate": 9.966712330401503e-06, + "loss": 7.184819221496582, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.3044436573982239, + "loss_total_v6": 0.4490512013435364, + "step": 352 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08141261261261261, + "grad_norm": 23.692100524902344, + "learning_rate": 9.966273388660508e-06, + "loss": 7.5667266845703125, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.33076953142881393, + "loss_total_v6": 0.47292041033506393, + "step": 353 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08164324324324325, + "grad_norm": 22.429271697998047, + "learning_rate": 9.965831581624872e-06, + "loss": 7.007347583770752, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.2974410764873028, + "loss_total_v6": 0.4379592463374138, + "step": 354 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08187387387387388, + "grad_norm": 24.041034698486328, + "learning_rate": 9.965386909549492e-06, + "loss": 7.107345104217529, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.3027906157076359, + "loss_total_v6": 0.4442090578377247, + "step": 355 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0821045045045045, + "grad_norm": 27.506389617919922, + "learning_rate": 9.964939372690927e-06, + "loss": 7.509897232055664, + "loss_alignment": 0.2919921875, + "loss_alignment_w": 0.14599609375, + "loss_sft": 0.3236013874411583, + "loss_total_v6": 0.4693685993552208, + "step": 356 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08233513513513513, + "grad_norm": 29.867103576660156, + "learning_rate": 9.96448897130738e-06, + "loss": 7.106472015380859, + "loss_alignment": 0.28662109375, + "loss_alignment_w": 0.143310546875, + "loss_sft": 0.3007218725979328, + "loss_total_v6": 0.4441544823348522, + "step": 357 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08256576576576577, + "grad_norm": 25.66002082824707, + "learning_rate": 9.96403570565871e-06, + "loss": 7.247155666351318, + "loss_alignment": 0.28759765625, + "loss_alignment_w": 0.143798828125, + "loss_sft": 0.3092552088201046, + "loss_total_v6": 0.452947236597538, + "step": 358 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0827963963963964, + "grad_norm": 26.665552139282227, + "learning_rate": 9.963579576006433e-06, + "loss": 6.853336334228516, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.28914283961057663, + "loss_total_v6": 0.4283335246145725, + "step": 359 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08302702702702702, + "grad_norm": 24.54277229309082, + "learning_rate": 9.96312058261371e-06, + "loss": 6.945476055145264, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.2933146394789219, + "loss_total_v6": 0.4340922310948372, + "step": 360 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08325765765765766, + "grad_norm": 26.339298248291016, + "learning_rate": 9.962658725745358e-06, + "loss": 7.483331203460693, + "loss_alignment": 0.285400390625, + "loss_alignment_w": 0.1427001953125, + "loss_sft": 0.32532844692468643, + "loss_total_v6": 0.46770820766687393, + "step": 361 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08348828828828829, + "grad_norm": 24.5319766998291, + "learning_rate": 9.962194005667844e-06, + "loss": 6.968182563781738, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.2956340927630663, + "loss_total_v6": 0.43551141396164894, + "step": 362 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08371891891891892, + "grad_norm": 26.36789321899414, + "learning_rate": 9.96172642264929e-06, + "loss": 7.847744464874268, + "loss_alignment": 0.2880859375, + "loss_alignment_w": 0.14404296875, + "loss_sft": 0.3464868329465389, + "loss_total_v6": 0.4904840290546417, + "step": 363 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08394954954954954, + "grad_norm": 24.350671768188477, + "learning_rate": 9.961255976959473e-06, + "loss": 7.733414649963379, + "loss_alignment": 0.287109375, + "loss_alignment_w": 0.1435546875, + "loss_sft": 0.3401499465107918, + "loss_total_v6": 0.4833384156227112, + "step": 364 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08418018018018018, + "grad_norm": 28.35759162902832, + "learning_rate": 9.960782668869811e-06, + "loss": 7.293044090270996, + "loss_alignment": 0.2841796875, + "loss_alignment_w": 0.14208984375, + "loss_sft": 0.31355755403637886, + "loss_total_v6": 0.45581523701548576, + "step": 365 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08441081081081081, + "grad_norm": 27.850078582763672, + "learning_rate": 9.96030649865338e-06, + "loss": 6.994415283203125, + "loss_alignment": 0.28662109375, + "loss_alignment_w": 0.143310546875, + "loss_sft": 0.29356577061116695, + "loss_total_v6": 0.4371509738266468, + "step": 366 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08464144144144144, + "grad_norm": 23.91880989074707, + "learning_rate": 9.959827466584907e-06, + "loss": 7.582955360412598, + "loss_alignment": 0.291748046875, + "loss_alignment_w": 0.1458740234375, + "loss_sft": 0.32856418564915657, + "loss_total_v6": 0.47393468394875526, + "step": 367 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08487207207207208, + "grad_norm": 24.722064971923828, + "learning_rate": 9.959345572940771e-06, + "loss": 7.041889190673828, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.29824186116456985, + "loss_total_v6": 0.44011808186769485, + "step": 368 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0851027027027027, + "grad_norm": 26.78715705871582, + "learning_rate": 9.958860817999002e-06, + "loss": 7.415759086608887, + "loss_alignment": 0.2939453125, + "loss_alignment_w": 0.14697265625, + "loss_sft": 0.31666488759219646, + "loss_total_v6": 0.4634849615395069, + "step": 369 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08533333333333333, + "grad_norm": 30.382991790771484, + "learning_rate": 9.958373202039278e-06, + "loss": 7.746700286865234, + "loss_alignment": 0.29248046875, + "loss_alignment_w": 0.146240234375, + "loss_sft": 0.33786752074956894, + "loss_total_v6": 0.48416880145668983, + "step": 370 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08556396396396396, + "grad_norm": 25.658363342285156, + "learning_rate": 9.957882725342926e-06, + "loss": 7.09286642074585, + "loss_alignment": 0.283203125, + "loss_alignment_w": 0.1416015625, + "loss_sft": 0.3018704690039158, + "loss_total_v6": 0.4433041624724865, + "step": 371 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0857945945945946, + "grad_norm": 24.477251052856445, + "learning_rate": 9.957389388192935e-06, + "loss": 7.429023742675781, + "loss_alignment": 0.28955078125, + "loss_alignment_w": 0.144775390625, + "loss_sft": 0.31934021413326263, + "loss_total_v6": 0.46431396529078484, + "step": 372 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08602522522522522, + "grad_norm": 25.73646354675293, + "learning_rate": 9.956893190873928e-06, + "loss": 6.921986103057861, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.29363181814551353, + "loss_total_v6": 0.43262413889169693, + "step": 373 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08625585585585585, + "grad_norm": 26.614490509033203, + "learning_rate": 9.956394133672193e-06, + "loss": 7.058215141296387, + "loss_alignment": 0.284423828125, + "loss_alignment_w": 0.1422119140625, + "loss_sft": 0.29904860258102417, + "loss_total_v6": 0.44113844633102417, + "step": 374 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08648648648648649, + "grad_norm": 26.31705093383789, + "learning_rate": 9.955892216875656e-06, + "loss": 7.783310413360596, + "loss_alignment": 0.286865234375, + "loss_alignment_w": 0.1434326171875, + "loss_sft": 0.34288695827126503, + "loss_total_v6": 0.48645690456032753, + "step": 375 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08671711711711712, + "grad_norm": 24.050695419311523, + "learning_rate": 9.955387440773902e-06, + "loss": 7.41633415222168, + "loss_alignment": 0.291748046875, + "loss_alignment_w": 0.1458740234375, + "loss_sft": 0.3178604692220688, + "loss_total_v6": 0.4635208658874035, + "step": 376 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08694774774774774, + "grad_norm": 23.725296020507812, + "learning_rate": 9.95487980565816e-06, + "loss": 7.5460638999938965, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.32999693416059017, + "loss_total_v6": 0.47162899747490883, + "step": 377 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08717837837837838, + "grad_norm": 25.289506912231445, + "learning_rate": 9.95436931182131e-06, + "loss": 7.228841781616211, + "loss_alignment": 0.290771484375, + "loss_alignment_w": 0.1453857421875, + "loss_sft": 0.3065847307443619, + "loss_total_v6": 0.4518026225268841, + "step": 378 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08740900900900901, + "grad_norm": 24.170259475708008, + "learning_rate": 9.953855959557883e-06, + "loss": 7.1944756507873535, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.3067866936326027, + "loss_total_v6": 0.4496547244489193, + "step": 379 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08763963963963964, + "grad_norm": 27.013938903808594, + "learning_rate": 9.953339749164057e-06, + "loss": 6.988319396972656, + "loss_alignment": 0.284912109375, + "loss_alignment_w": 0.1424560546875, + "loss_sft": 0.2942528575658798, + "loss_total_v6": 0.4367699511349201, + "step": 380 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08787027027027026, + "grad_norm": 25.261043548583984, + "learning_rate": 9.95282068093766e-06, + "loss": 7.6874237060546875, + "loss_alignment": 0.29150390625, + "loss_alignment_w": 0.145751953125, + "loss_sft": 0.33486461639404297, + "loss_total_v6": 0.48046400025486946, + "step": 381 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0881009009009009, + "grad_norm": 26.793012619018555, + "learning_rate": 9.952298755178171e-06, + "loss": 7.456446170806885, + "loss_alignment": 0.287109375, + "loss_alignment_w": 0.1435546875, + "loss_sft": 0.3224732019007206, + "loss_total_v6": 0.4660278931260109, + "step": 382 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08833153153153153, + "grad_norm": 26.03934097290039, + "learning_rate": 9.951773972186712e-06, + "loss": 7.4471845626831055, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.3244578316807747, + "loss_total_v6": 0.4654490277171135, + "step": 383 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08856216216216216, + "grad_norm": 26.65264892578125, + "learning_rate": 9.951246332266058e-06, + "loss": 7.084659576416016, + "loss_alignment": 0.2880859375, + "loss_alignment_w": 0.14404296875, + "loss_sft": 0.29917551949620247, + "loss_total_v6": 0.44279123470187187, + "step": 384 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0887927927927928, + "grad_norm": 29.305906295776367, + "learning_rate": 9.950715835720633e-06, + "loss": 7.155921936035156, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.30262231826782227, + "loss_total_v6": 0.44724512472748756, + "step": 385 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08902342342342343, + "grad_norm": 24.153602600097656, + "learning_rate": 9.950182482856506e-06, + "loss": 7.25124454498291, + "loss_alignment": 0.28759765625, + "loss_alignment_w": 0.143798828125, + "loss_sft": 0.3093429319560528, + "loss_total_v6": 0.4532027915120125, + "step": 386 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08925405405405405, + "grad_norm": 27.498687744140625, + "learning_rate": 9.949646273981394e-06, + "loss": 7.241844177246094, + "loss_alignment": 0.291015625, + "loss_alignment_w": 0.1455078125, + "loss_sft": 0.3071226868778467, + "loss_total_v6": 0.45261523872613907, + "step": 387 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08948468468468468, + "grad_norm": 28.776168823242188, + "learning_rate": 9.949107209404664e-06, + "loss": 7.59818172454834, + "loss_alignment": 0.288330078125, + "loss_alignment_w": 0.1441650390625, + "loss_sft": 0.33052292466163635, + "loss_total_v6": 0.47488634288311005, + "step": 388 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08971531531531532, + "grad_norm": 27.50876808166504, + "learning_rate": 9.94856528943733e-06, + "loss": 7.428543567657471, + "loss_alignment": 0.2880859375, + "loss_alignment_w": 0.14404296875, + "loss_sft": 0.3203020356595516, + "loss_total_v6": 0.4642839692533016, + "step": 389 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.08994594594594595, + "grad_norm": 27.90569496154785, + "learning_rate": 9.948020514392055e-06, + "loss": 7.59851598739624, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.3344806134700775, + "loss_total_v6": 0.4749072529375553, + "step": 390 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09017657657657657, + "grad_norm": 24.870254516601562, + "learning_rate": 9.947472884583145e-06, + "loss": 7.360721588134766, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.31560540199279785, + "loss_total_v6": 0.46004508808255196, + "step": 391 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09040720720720721, + "grad_norm": 25.85468292236328, + "learning_rate": 9.946922400326554e-06, + "loss": 7.44698429107666, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.3240638040006161, + "loss_total_v6": 0.46543649584054947, + "step": 392 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09063783783783784, + "grad_norm": 25.600770950317383, + "learning_rate": 9.946369061939886e-06, + "loss": 7.664048194885254, + "loss_alignment": 0.284423828125, + "loss_alignment_w": 0.1422119140625, + "loss_sft": 0.33750828728079796, + "loss_total_v6": 0.47900303080677986, + "step": 393 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09086846846846847, + "grad_norm": 27.072858810424805, + "learning_rate": 9.94581286974239e-06, + "loss": 7.334174633026123, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.3179135024547577, + "loss_total_v6": 0.4583859108388424, + "step": 394 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09109909909909909, + "grad_norm": 23.649900436401367, + "learning_rate": 9.945253824054964e-06, + "loss": 7.146882057189941, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.3074131906032562, + "loss_total_v6": 0.44668014720082283, + "step": 395 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09132972972972973, + "grad_norm": 26.70948600769043, + "learning_rate": 9.944691925200145e-06, + "loss": 7.023816108703613, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.29775315150618553, + "loss_total_v6": 0.43898850306868553, + "step": 396 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09156036036036036, + "grad_norm": 26.218563079833984, + "learning_rate": 9.944127173502123e-06, + "loss": 7.380528450012207, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.31698065623641014, + "loss_total_v6": 0.46128301694989204, + "step": 397 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09179099099099099, + "grad_norm": 24.913524627685547, + "learning_rate": 9.943559569286731e-06, + "loss": 7.37502384185791, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3203750476241112, + "loss_total_v6": 0.4609390087425709, + "step": 398 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09202162162162163, + "grad_norm": 25.069080352783203, + "learning_rate": 9.942989112881451e-06, + "loss": 7.425981044769287, + "loss_alignment": 0.28662109375, + "loss_alignment_w": 0.143310546875, + "loss_sft": 0.32075220346450806, + "loss_total_v6": 0.46412380412220955, + "step": 399 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09225225225225225, + "grad_norm": 24.785640716552734, + "learning_rate": 9.942415804615407e-06, + "loss": 7.093409538269043, + "loss_alignment": 0.285888671875, + "loss_alignment_w": 0.1429443359375, + "loss_sft": 0.3007446900010109, + "loss_total_v6": 0.4433380886912346, + "step": 400 + }, + { + "epoch": 0.09225225225225225, + "eval_loss": 0.6003463864326477, + "eval_loss_alignment": 0.2561915846175799, + "eval_loss_alignment_w": 0.12809579230878995, + "eval_loss_sft": 0.4722505659460341, + "eval_loss_total_v6": 0.6003463583653921, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.962, + "eval_samples_per_second": 18.975, + "eval_steps_per_second": 2.381, + "step": 400 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09248288288288288, + "grad_norm": 23.95174217224121, + "learning_rate": 9.941839644819366e-06, + "loss": 6.796367645263672, + "loss_alignment": 0.285400390625, + "loss_alignment_w": 0.1427001953125, + "loss_sft": 0.2817523591220379, + "loss_total_v6": 0.4247729852795601, + "step": 401 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0927135135135135, + "grad_norm": 25.516630172729492, + "learning_rate": 9.941260633825752e-06, + "loss": 7.472585678100586, + "loss_alignment": 0.285400390625, + "loss_alignment_w": 0.1427001953125, + "loss_sft": 0.3239854760468006, + "loss_total_v6": 0.4670366197824478, + "step": 402 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09294414414414415, + "grad_norm": 25.75156021118164, + "learning_rate": 9.940678771968617e-06, + "loss": 7.113954067230225, + "loss_alignment": 0.288330078125, + "loss_alignment_w": 0.1441650390625, + "loss_sft": 0.30027398839592934, + "loss_total_v6": 0.44462213665246964, + "step": 403 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09317477477477477, + "grad_norm": 26.101158142089844, + "learning_rate": 9.940094059583672e-06, + "loss": 7.2751007080078125, + "loss_alignment": 0.289794921875, + "loss_alignment_w": 0.1448974609375, + "loss_sft": 0.310116745531559, + "loss_total_v6": 0.4546937867999077, + "step": 404 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0934054054054054, + "grad_norm": 24.337228775024414, + "learning_rate": 9.939506497008264e-06, + "loss": 7.014217376708984, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.29747365415096283, + "loss_total_v6": 0.43838857114315033, + "step": 405 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09363603603603604, + "grad_norm": 26.594327926635742, + "learning_rate": 9.938916084581391e-06, + "loss": 7.609006404876709, + "loss_alignment": 0.288330078125, + "loss_alignment_w": 0.1441650390625, + "loss_sft": 0.3314436487853527, + "loss_total_v6": 0.4755629003047943, + "step": 406 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09386666666666667, + "grad_norm": 25.003795623779297, + "learning_rate": 9.938322822643688e-06, + "loss": 6.886224269866943, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.28799400106072426, + "loss_total_v6": 0.43038900941610336, + "step": 407 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0940972972972973, + "grad_norm": 25.401756286621094, + "learning_rate": 9.937726711537441e-06, + "loss": 7.29933500289917, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.31500358507037163, + "loss_total_v6": 0.4562084339559078, + "step": 408 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09432792792792793, + "grad_norm": 27.8015079498291, + "learning_rate": 9.937127751606577e-06, + "loss": 6.786284446716309, + "loss_alignment": 0.284423828125, + "loss_alignment_w": 0.1422119140625, + "loss_sft": 0.282007172703743, + "loss_total_v6": 0.4241428077220917, + "step": 409 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09455855855855856, + "grad_norm": 25.717859268188477, + "learning_rate": 9.936525943196663e-06, + "loss": 7.124460220336914, + "loss_alignment": 0.28857421875, + "loss_alignment_w": 0.144287109375, + "loss_sft": 0.3012663070112467, + "loss_total_v6": 0.44527876004576683, + "step": 410 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09478918918918919, + "grad_norm": 26.309680938720703, + "learning_rate": 9.935921286654914e-06, + "loss": 7.438928604125977, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.32475055009126663, + "loss_total_v6": 0.464933056384325, + "step": 411 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09501981981981981, + "grad_norm": 23.80106544494629, + "learning_rate": 9.93531378233019e-06, + "loss": 7.245623588562012, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.30970876663923264, + "loss_total_v6": 0.45285147055983543, + "step": 412 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09525045045045046, + "grad_norm": 23.860260009765625, + "learning_rate": 9.934703430572988e-06, + "loss": 7.273680686950684, + "loss_alignment": 0.2861328125, + "loss_alignment_w": 0.14306640625, + "loss_sft": 0.31123347394168377, + "loss_total_v6": 0.4546050615608692, + "step": 413 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09548108108108108, + "grad_norm": 23.841468811035156, + "learning_rate": 9.93409023173545e-06, + "loss": 7.247809410095215, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.31234786100685596, + "loss_total_v6": 0.4529880955815315, + "step": 414 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09571171171171171, + "grad_norm": 31.975452423095703, + "learning_rate": 9.933474186171363e-06, + "loss": 6.976733207702637, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.29561918042600155, + "loss_total_v6": 0.4360458068549633, + "step": 415 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09594234234234235, + "grad_norm": 26.765989303588867, + "learning_rate": 9.932855294236154e-06, + "loss": 7.723121643066406, + "loss_alignment": 0.28759765625, + "loss_alignment_w": 0.143798828125, + "loss_sft": 0.3388199768960476, + "loss_total_v6": 0.482695110142231, + "step": 416 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09617297297297298, + "grad_norm": 28.0001277923584, + "learning_rate": 9.932233556286897e-06, + "loss": 7.407644748687744, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.3231462463736534, + "loss_total_v6": 0.4629777893424034, + "step": 417 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0964036036036036, + "grad_norm": 25.205341339111328, + "learning_rate": 9.931608972682298e-06, + "loss": 7.585402488708496, + "loss_alignment": 0.283203125, + "loss_alignment_w": 0.1416015625, + "loss_sft": 0.3326081708073616, + "loss_total_v6": 0.474087655544281, + "step": 418 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09663423423423423, + "grad_norm": 26.078617095947266, + "learning_rate": 9.930981543782716e-06, + "loss": 7.772979736328125, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.3429889753460884, + "loss_total_v6": 0.4858112335205078, + "step": 419 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09686486486486487, + "grad_norm": 29.503498077392578, + "learning_rate": 9.930351269950144e-06, + "loss": 7.664107322692871, + "loss_alignment": 0.293212890625, + "loss_alignment_w": 0.1466064453125, + "loss_sft": 0.33224768191576004, + "loss_total_v6": 0.47900672256946564, + "step": 420 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0970954954954955, + "grad_norm": 25.30586051940918, + "learning_rate": 9.929718151548218e-06, + "loss": 7.257260322570801, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.31237392500042915, + "loss_total_v6": 0.45357875525951385, + "step": 421 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09732612612612612, + "grad_norm": 28.477359771728516, + "learning_rate": 9.929082188942219e-06, + "loss": 7.2697062492370605, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.312800869345665, + "loss_total_v6": 0.4543566480278969, + "step": 422 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09755675675675676, + "grad_norm": 23.070049285888672, + "learning_rate": 9.928443382499063e-06, + "loss": 7.204568386077881, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.30923328548669815, + "loss_total_v6": 0.45028552040457726, + "step": 423 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09778738738738739, + "grad_norm": 105.86271667480469, + "learning_rate": 9.927801732587311e-06, + "loss": 6.993407249450684, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.2974090166389942, + "loss_total_v6": 0.4370879679918289, + "step": 424 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09801801801801802, + "grad_norm": 24.655336380004883, + "learning_rate": 9.927157239577165e-06, + "loss": 7.572198390960693, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.33173712715506554, + "loss_total_v6": 0.47326240316033363, + "step": 425 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09824864864864864, + "grad_norm": 25.073116302490234, + "learning_rate": 9.926509903840464e-06, + "loss": 7.802349090576172, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.345587532967329, + "loss_total_v6": 0.48764683306217194, + "step": 426 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09847927927927928, + "grad_norm": 25.254688262939453, + "learning_rate": 9.92585972575069e-06, + "loss": 7.347054958343506, + "loss_alignment": 0.2861328125, + "loss_alignment_w": 0.14306640625, + "loss_sft": 0.3160024657845497, + "loss_total_v6": 0.4591909274458885, + "step": 427 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09870990990990991, + "grad_norm": 28.838956832885742, + "learning_rate": 9.925206705682961e-06, + "loss": 8.142644882202148, + "loss_alignment": 0.29052734375, + "loss_alignment_w": 0.145263671875, + "loss_sft": 0.3636211268603802, + "loss_total_v6": 0.5089153200387955, + "step": 428 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09894054054054054, + "grad_norm": 26.730464935302734, + "learning_rate": 9.924550844014039e-06, + "loss": 7.3794097900390625, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.3181009441614151, + "loss_total_v6": 0.4612131081521511, + "step": 429 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09917117117117118, + "grad_norm": 24.474353790283203, + "learning_rate": 9.923892141122324e-06, + "loss": 7.312655448913574, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.3153936415910721, + "loss_total_v6": 0.4570409767329693, + "step": 430 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.0994018018018018, + "grad_norm": 25.217803955078125, + "learning_rate": 9.923230597387856e-06, + "loss": 7.196147441864014, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.30882903560996056, + "loss_total_v6": 0.44975920394062996, + "step": 431 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09963243243243243, + "grad_norm": 25.212018966674805, + "learning_rate": 9.92256621319231e-06, + "loss": 7.219468116760254, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.3097830042243004, + "loss_total_v6": 0.4512167349457741, + "step": 432 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.09986306306306306, + "grad_norm": 27.188098907470703, + "learning_rate": 9.921898988919006e-06, + "loss": 7.238982677459717, + "loss_alignment": 0.285400390625, + "loss_alignment_w": 0.1427001953125, + "loss_sft": 0.3097362294793129, + "loss_total_v6": 0.4524364098906517, + "step": 433 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1000936936936937, + "grad_norm": 27.64120864868164, + "learning_rate": 9.9212289249529e-06, + "loss": 7.607364177703857, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.3339807651937008, + "loss_total_v6": 0.475460272282362, + "step": 434 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10032432432432432, + "grad_norm": 23.68056297302246, + "learning_rate": 9.92055602168058e-06, + "loss": 7.858199119567871, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.34649940207600594, + "loss_total_v6": 0.49113744869828224, + "step": 435 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10055495495495495, + "grad_norm": 25.967283248901367, + "learning_rate": 9.919880279490286e-06, + "loss": 7.700484275817871, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.34082309901714325, + "loss_total_v6": 0.48128025978803635, + "step": 436 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10078558558558559, + "grad_norm": 22.253768920898438, + "learning_rate": 9.919201698771882e-06, + "loss": 7.211543083190918, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.30945559218525887, + "loss_total_v6": 0.45072146132588387, + "step": 437 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10101621621621622, + "grad_norm": 25.953041076660156, + "learning_rate": 9.918520279916879e-06, + "loss": 7.106566429138184, + "loss_alignment": 0.28466796875, + "loss_alignment_w": 0.142333984375, + "loss_sft": 0.3019026927649975, + "loss_total_v6": 0.444160383194685, + "step": 438 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10124684684684684, + "grad_norm": 28.015933990478516, + "learning_rate": 9.91783602331842e-06, + "loss": 7.245784759521484, + "loss_alignment": 0.2841796875, + "loss_alignment_w": 0.14208984375, + "loss_sft": 0.3109853193163872, + "loss_total_v6": 0.4528615362942219, + "step": 439 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10147747747747747, + "grad_norm": 27.152992248535156, + "learning_rate": 9.917148929371287e-06, + "loss": 7.014365196228027, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.29754393361508846, + "loss_total_v6": 0.4383978024125099, + "step": 440 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10170810810810811, + "grad_norm": 25.223331451416016, + "learning_rate": 9.916458998471902e-06, + "loss": 7.104636192321777, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.3032164201140404, + "loss_total_v6": 0.4440397769212723, + "step": 441 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10193873873873874, + "grad_norm": 25.637144088745117, + "learning_rate": 9.915766231018317e-06, + "loss": 6.495797157287598, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.26540811359882355, + "loss_total_v6": 0.40598733723163605, + "step": 442 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10216936936936936, + "grad_norm": 24.069929122924805, + "learning_rate": 9.915070627410229e-06, + "loss": 7.094552040100098, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.3018994852900505, + "loss_total_v6": 0.4434094987809658, + "step": 443 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1024, + "grad_norm": 24.730363845825195, + "learning_rate": 9.914372188048964e-06, + "loss": 7.875697135925293, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3514076918363571, + "loss_total_v6": 0.4922310560941696, + "step": 444 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10263063063063063, + "grad_norm": 23.55607032775879, + "learning_rate": 9.91367091333749e-06, + "loss": 7.210235595703125, + "loss_alignment": 0.28564453125, + "loss_alignment_w": 0.142822265625, + "loss_sft": 0.3079242818057537, + "loss_total_v6": 0.4506397321820259, + "step": 445 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10286126126126126, + "grad_norm": 25.00507354736328, + "learning_rate": 9.912966803680404e-06, + "loss": 7.117639541625977, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.30381549149751663, + "loss_total_v6": 0.4448524974286556, + "step": 446 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1030918918918919, + "grad_norm": 25.43283462524414, + "learning_rate": 9.912259859483943e-06, + "loss": 7.098872184753418, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.3021084554493427, + "loss_total_v6": 0.44367948547005653, + "step": 447 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10332252252252253, + "grad_norm": 25.574352264404297, + "learning_rate": 9.911550081155983e-06, + "loss": 7.2906107902526855, + "loss_alignment": 0.2841796875, + "loss_alignment_w": 0.14208984375, + "loss_sft": 0.313695402815938, + "loss_total_v6": 0.45566317811608315, + "step": 448 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10355315315315315, + "grad_norm": 26.01457977294922, + "learning_rate": 9.910837469106028e-06, + "loss": 7.42388916015625, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.3226967044174671, + "loss_total_v6": 0.4639930836856365, + "step": 449 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10378378378378378, + "grad_norm": 22.14702033996582, + "learning_rate": 9.910122023745219e-06, + "loss": 7.28305721282959, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.31654975935816765, + "loss_total_v6": 0.45519109815359116, + "step": 450 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10401441441441442, + "grad_norm": 24.93172264099121, + "learning_rate": 9.909403745486335e-06, + "loss": 7.300642967224121, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.3165196552872658, + "loss_total_v6": 0.4562901742756367, + "step": 451 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10424504504504505, + "grad_norm": 25.29639434814453, + "learning_rate": 9.908682634743785e-06, + "loss": 6.929100036621094, + "loss_alignment": 0.2841796875, + "loss_alignment_w": 0.14208984375, + "loss_sft": 0.291116239503026, + "loss_total_v6": 0.43306873738765717, + "step": 452 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10447567567567567, + "grad_norm": 25.654848098754883, + "learning_rate": 9.907958691933616e-06, + "loss": 7.157703399658203, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3103935495018959, + "loss_total_v6": 0.4473564475774765, + "step": 453 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10470630630630631, + "grad_norm": 25.325759887695312, + "learning_rate": 9.907231917473507e-06, + "loss": 7.568625450134277, + "loss_alignment": 0.2841796875, + "loss_alignment_w": 0.14208984375, + "loss_sft": 0.33104081079363823, + "loss_total_v6": 0.47303910553455353, + "step": 454 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10493693693693694, + "grad_norm": 24.554676055908203, + "learning_rate": 9.90650231178277e-06, + "loss": 7.555966377258301, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.3310583084821701, + "loss_total_v6": 0.4722478836774826, + "step": 455 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10516756756756757, + "grad_norm": 25.747770309448242, + "learning_rate": 9.905769875282351e-06, + "loss": 7.091398239135742, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.3017939403653145, + "loss_total_v6": 0.4432123936712742, + "step": 456 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10539819819819819, + "grad_norm": 26.10896873474121, + "learning_rate": 9.905034608394835e-06, + "loss": 7.291767120361328, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.31565978936851025, + "loss_total_v6": 0.4557354599237442, + "step": 457 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10562882882882883, + "grad_norm": 22.537370681762695, + "learning_rate": 9.904296511544427e-06, + "loss": 7.149600028991699, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.3078424446284771, + "loss_total_v6": 0.4468500055372715, + "step": 458 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10585945945945946, + "grad_norm": 24.77789306640625, + "learning_rate": 9.903555585156977e-06, + "loss": 7.80611515045166, + "loss_alignment": 0.28564453125, + "loss_alignment_w": 0.142822265625, + "loss_sft": 0.34521250426769257, + "loss_total_v6": 0.48788217082619667, + "step": 459 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10609009009009009, + "grad_norm": 24.038419723510742, + "learning_rate": 9.902811829659962e-06, + "loss": 7.703800678253174, + "loss_alignment": 0.284912109375, + "loss_alignment_w": 0.1424560546875, + "loss_sft": 0.33881785720586777, + "loss_total_v6": 0.48148754239082336, + "step": 460 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10632072072072073, + "grad_norm": 24.705612182617188, + "learning_rate": 9.902065245482493e-06, + "loss": 7.616962432861328, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.3327038325369358, + "loss_total_v6": 0.4760601483285427, + "step": 461 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10655135135135135, + "grad_norm": 21.833175659179688, + "learning_rate": 9.90131583305531e-06, + "loss": 6.850624084472656, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.2890954241156578, + "loss_total_v6": 0.4281640164554119, + "step": 462 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10678198198198198, + "grad_norm": 23.027158737182617, + "learning_rate": 9.900563592810789e-06, + "loss": 7.166749954223633, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.30812086910009384, + "loss_total_v6": 0.44792189076542854, + "step": 463 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1070126126126126, + "grad_norm": 24.870441436767578, + "learning_rate": 9.899808525182935e-06, + "loss": 7.606190204620361, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.3308251202106476, + "loss_total_v6": 0.4753868766129017, + "step": 464 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10724324324324325, + "grad_norm": 26.1025390625, + "learning_rate": 9.899050630607386e-06, + "loss": 7.071601867675781, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.30127381533384323, + "loss_total_v6": 0.44197510182857513, + "step": 465 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10747387387387387, + "grad_norm": 28.987415313720703, + "learning_rate": 9.898289909521405e-06, + "loss": 7.147857189178467, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.3074588496237993, + "loss_total_v6": 0.4467410780489445, + "step": 466 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1077045045045045, + "grad_norm": 25.98288917541504, + "learning_rate": 9.897526362363896e-06, + "loss": 7.504622459411621, + "loss_alignment": 0.2880859375, + "loss_alignment_w": 0.14404296875, + "loss_sft": 0.3249196596443653, + "loss_total_v6": 0.4690389223396778, + "step": 467 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10793513513513514, + "grad_norm": 26.819744110107422, + "learning_rate": 9.896759989575386e-06, + "loss": 7.441749572753906, + "loss_alignment": 0.284912109375, + "loss_alignment_w": 0.1424560546875, + "loss_sft": 0.32276009023189545, + "loss_total_v6": 0.46510933712124825, + "step": 468 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10816576576576577, + "grad_norm": 26.1023006439209, + "learning_rate": 9.895990791598037e-06, + "loss": 6.919488430023193, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.2881503999233246, + "loss_total_v6": 0.4324680268764496, + "step": 469 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1083963963963964, + "grad_norm": 27.836475372314453, + "learning_rate": 9.89521876887563e-06, + "loss": 7.768131256103516, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.34486793354153633, + "loss_total_v6": 0.48550818860530853, + "step": 470 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10862702702702702, + "grad_norm": 23.021345138549805, + "learning_rate": 9.894443921853594e-06, + "loss": 7.469030857086182, + "loss_alignment": 0.288818359375, + "loss_alignment_w": 0.1444091796875, + "loss_sft": 0.32292405888438225, + "loss_total_v6": 0.46681443229317665, + "step": 471 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10885765765765766, + "grad_norm": 28.87868309020996, + "learning_rate": 9.893666250978971e-06, + "loss": 7.690742015838623, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.34024474024772644, + "loss_total_v6": 0.48067137971520424, + "step": 472 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10908828828828829, + "grad_norm": 29.42804718017578, + "learning_rate": 9.892885756700441e-06, + "loss": 7.2440643310546875, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.31272411346435547, + "loss_total_v6": 0.45275401696562767, + "step": 473 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10931891891891891, + "grad_norm": 25.17354965209961, + "learning_rate": 9.89210243946831e-06, + "loss": 7.382498264312744, + "loss_alignment": 0.28955078125, + "loss_alignment_w": 0.144775390625, + "loss_sft": 0.3163561001420021, + "loss_total_v6": 0.4614061526954174, + "step": 474 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10954954954954955, + "grad_norm": 24.046607971191406, + "learning_rate": 9.891316299734514e-06, + "loss": 7.038099765777588, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3027505073696375, + "loss_total_v6": 0.43988125771284103, + "step": 475 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.10978018018018018, + "grad_norm": 24.72098159790039, + "learning_rate": 9.890527337952617e-06, + "loss": 7.109096527099609, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.3050668276846409, + "loss_total_v6": 0.4443185329437256, + "step": 476 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11001081081081081, + "grad_norm": 26.233585357666016, + "learning_rate": 9.88973555457781e-06, + "loss": 6.957071304321289, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.2963586989790201, + "loss_total_v6": 0.43481694906949997, + "step": 477 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11024144144144143, + "grad_norm": 23.927278518676758, + "learning_rate": 9.888940950066915e-06, + "loss": 7.6566548347473145, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.33660366386175156, + "loss_total_v6": 0.47854093462228775, + "step": 478 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11047207207207208, + "grad_norm": 25.117835998535156, + "learning_rate": 9.888143524878378e-06, + "loss": 6.951703071594238, + "loss_alignment": 0.287353515625, + "loss_alignment_w": 0.1436767578125, + "loss_sft": 0.2906826101243496, + "loss_total_v6": 0.4344814345240593, + "step": 479 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1107027027027027, + "grad_norm": 24.451074600219727, + "learning_rate": 9.887343279472272e-06, + "loss": 7.300184726715088, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.3149346075952053, + "loss_total_v6": 0.4562615230679512, + "step": 480 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11093333333333333, + "grad_norm": 30.073286056518555, + "learning_rate": 9.886540214310303e-06, + "loss": 7.270021438598633, + "loss_alignment": 0.2861328125, + "loss_alignment_w": 0.14306640625, + "loss_sft": 0.31175244227051735, + "loss_total_v6": 0.45437632873654366, + "step": 481 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11116396396396397, + "grad_norm": 22.8901309967041, + "learning_rate": 9.885734329855798e-06, + "loss": 7.4568610191345215, + "loss_alignment": 0.283935546875, + "loss_alignment_w": 0.1419677734375, + "loss_sft": 0.3237045593559742, + "loss_total_v6": 0.4660538099706173, + "step": 482 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1113945945945946, + "grad_norm": 27.824722290039062, + "learning_rate": 9.884925626573714e-06, + "loss": 7.283446311950684, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.3140410967171192, + "loss_total_v6": 0.4552154019474983, + "step": 483 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11162522522522522, + "grad_norm": 24.96876335144043, + "learning_rate": 9.884114104930629e-06, + "loss": 7.390820026397705, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.32050779461860657, + "loss_total_v6": 0.46192625537514687, + "step": 484 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11185585585585586, + "grad_norm": 23.497150421142578, + "learning_rate": 9.883299765394756e-06, + "loss": 7.466849327087402, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.32554954662919044, + "loss_total_v6": 0.46667807921767235, + "step": 485 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11208648648648649, + "grad_norm": 24.748802185058594, + "learning_rate": 9.882482608435924e-06, + "loss": 7.433751106262207, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.32326729968190193, + "loss_total_v6": 0.46460945159196854, + "step": 486 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11231711711711712, + "grad_norm": 24.154613494873047, + "learning_rate": 9.881662634525596e-06, + "loss": 7.295945167541504, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.3162718378007412, + "loss_total_v6": 0.4559965692460537, + "step": 487 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11254774774774774, + "grad_norm": 23.442028045654297, + "learning_rate": 9.880839844136854e-06, + "loss": 7.471139907836914, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.3259092718362808, + "loss_total_v6": 0.4669462628662586, + "step": 488 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11277837837837838, + "grad_norm": 24.042509078979492, + "learning_rate": 9.88001423774441e-06, + "loss": 7.601711273193359, + "loss_alignment": 0.287109375, + "loss_alignment_w": 0.1435546875, + "loss_sft": 0.33146071806550026, + "loss_total_v6": 0.47510695457458496, + "step": 489 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11300900900900901, + "grad_norm": 26.36500358581543, + "learning_rate": 9.879185815824595e-06, + "loss": 7.2740478515625, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.31418611481785774, + "loss_total_v6": 0.45462799072265625, + "step": 490 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11323963963963964, + "grad_norm": 24.998645782470703, + "learning_rate": 9.87835457885537e-06, + "loss": 7.188430309295654, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3104219138622284, + "loss_total_v6": 0.4492768943309784, + "step": 491 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11347027027027028, + "grad_norm": 23.537193298339844, + "learning_rate": 9.877520527316317e-06, + "loss": 7.178150177001953, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.3088028449565172, + "loss_total_v6": 0.44863439723849297, + "step": 492 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1137009009009009, + "grad_norm": 23.557979583740234, + "learning_rate": 9.876683661688642e-06, + "loss": 7.2982587814331055, + "loss_alignment": 0.28564453125, + "loss_alignment_w": 0.142822265625, + "loss_sft": 0.313059501349926, + "loss_total_v6": 0.4561411701142788, + "step": 493 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11393153153153153, + "grad_norm": 23.628482818603516, + "learning_rate": 9.875843982455176e-06, + "loss": 7.435612678527832, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.323337871581316, + "loss_total_v6": 0.4647258035838604, + "step": 494 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11416216216216216, + "grad_norm": 26.56638526916504, + "learning_rate": 9.875001490100372e-06, + "loss": 6.996859550476074, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.30035608634352684, + "loss_total_v6": 0.43730372563004494, + "step": 495 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1143927927927928, + "grad_norm": 26.52386474609375, + "learning_rate": 9.874156185110307e-06, + "loss": 7.3460283279418945, + "loss_alignment": 0.287841796875, + "loss_alignment_w": 0.1439208984375, + "loss_sft": 0.3149617239832878, + "loss_total_v6": 0.4591267704963684, + "step": 496 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11462342342342342, + "grad_norm": 25.082643508911133, + "learning_rate": 9.873308067972679e-06, + "loss": 7.630853176116943, + "loss_alignment": 0.284912109375, + "loss_alignment_w": 0.1424560546875, + "loss_sft": 0.33416712284088135, + "loss_total_v6": 0.47692834213376045, + "step": 497 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11485405405405405, + "grad_norm": 22.996652603149414, + "learning_rate": 9.872457139176812e-06, + "loss": 7.510892391204834, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.3287905119359493, + "loss_total_v6": 0.4694307819008827, + "step": 498 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11508468468468469, + "grad_norm": 27.044641494750977, + "learning_rate": 9.871603399213647e-06, + "loss": 7.568991184234619, + "loss_alignment": 0.283935546875, + "loss_alignment_w": 0.1419677734375, + "loss_sft": 0.3310636170208454, + "loss_total_v6": 0.4730619415640831, + "step": 499 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11531531531531532, + "grad_norm": 25.207088470458984, + "learning_rate": 9.870746848575751e-06, + "loss": 7.187070846557617, + "loss_alignment": 0.283935546875, + "loss_alignment_w": 0.1419677734375, + "loss_sft": 0.30777350068092346, + "loss_total_v6": 0.44919194281101227, + "step": 500 + }, + { + "epoch": 0.11531531531531532, + "eval_loss": 0.6034350991249084, + "eval_loss_alignment": 0.2532507491438356, + "eval_loss_alignment_w": 0.1266253745719178, + "eval_loss_sft": 0.47680965244565926, + "eval_loss_total_v6": 0.6034350265753051, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.1378, + "eval_samples_per_second": 19.147, + "eval_steps_per_second": 2.403, + "step": 500 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11554594594594594, + "grad_norm": 24.586816787719727, + "learning_rate": 9.869887487757312e-06, + "loss": 7.702412128448486, + "loss_alignment": 0.286376953125, + "loss_alignment_w": 0.1431884765625, + "loss_sft": 0.3381054364144802, + "loss_total_v6": 0.4814007319509983, + "step": 501 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11577657657657657, + "grad_norm": 26.656160354614258, + "learning_rate": 9.869025317254139e-06, + "loss": 6.691450119018555, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.27958953380584717, + "loss_total_v6": 0.41821563243865967, + "step": 502 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11600720720720721, + "grad_norm": 24.530593872070312, + "learning_rate": 9.868160337563662e-06, + "loss": 7.458154201507568, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.324380487203598, + "loss_total_v6": 0.4661346413195133, + "step": 503 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11623783783783784, + "grad_norm": 22.608381271362305, + "learning_rate": 9.86729254918493e-06, + "loss": 6.854622840881348, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.2917409669607878, + "loss_total_v6": 0.4284139387309551, + "step": 504 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11646846846846846, + "grad_norm": 24.220348358154297, + "learning_rate": 9.866421952618616e-06, + "loss": 7.457753658294678, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.32885680347681046, + "loss_total_v6": 0.46610959619283676, + "step": 505 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1166990990990991, + "grad_norm": 27.95622444152832, + "learning_rate": 9.86554854836701e-06, + "loss": 7.804121017456055, + "loss_alignment": 0.28857421875, + "loss_alignment_w": 0.144287109375, + "loss_sft": 0.3434246852993965, + "loss_total_v6": 0.4877575524151325, + "step": 506 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11692972972972973, + "grad_norm": 31.502065658569336, + "learning_rate": 9.864672336934022e-06, + "loss": 7.100419998168945, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.30281554721295834, + "loss_total_v6": 0.4437762424349785, + "step": 507 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11716036036036036, + "grad_norm": 23.620718002319336, + "learning_rate": 9.863793318825186e-06, + "loss": 7.07127571105957, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.30113136023283005, + "loss_total_v6": 0.44195471704006195, + "step": 508 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11739099099099098, + "grad_norm": 25.029321670532227, + "learning_rate": 9.86291149454765e-06, + "loss": 7.051932334899902, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.302348542958498, + "loss_total_v6": 0.4407457560300827, + "step": 509 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11762162162162162, + "grad_norm": 26.45556640625, + "learning_rate": 9.862026864610183e-06, + "loss": 6.895318984985352, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.2926212288439274, + "loss_total_v6": 0.4309574291110039, + "step": 510 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11785225225225225, + "grad_norm": 25.88391876220703, + "learning_rate": 9.861139429523173e-06, + "loss": 7.577051162719727, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.3343750238418579, + "loss_total_v6": 0.4735656827688217, + "step": 511 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11808288288288288, + "grad_norm": 24.20660972595215, + "learning_rate": 9.860249189798628e-06, + "loss": 7.142551898956299, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3077833913266659, + "loss_total_v6": 0.4464094899594784, + "step": 512 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11831351351351352, + "grad_norm": 26.06028938293457, + "learning_rate": 9.859356145950169e-06, + "loss": 7.370942115783691, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.3184872530400753, + "loss_total_v6": 0.4606838896870613, + "step": 513 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11854414414414415, + "grad_norm": 24.397201538085938, + "learning_rate": 9.85846029849304e-06, + "loss": 7.058782577514648, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.30276142433285713, + "loss_total_v6": 0.4411739148199558, + "step": 514 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11877477477477477, + "grad_norm": 35.553810119628906, + "learning_rate": 9.857561647944101e-06, + "loss": 7.0239763259887695, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3027528189122677, + "loss_total_v6": 0.4389985427260399, + "step": 515 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11900540540540541, + "grad_norm": 23.976959228515625, + "learning_rate": 9.856660194821829e-06, + "loss": 6.819845676422119, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.2859968412667513, + "loss_total_v6": 0.42624036222696304, + "step": 516 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11923603603603604, + "grad_norm": 24.029550552368164, + "learning_rate": 9.855755939646318e-06, + "loss": 7.312995910644531, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.3155980221927166, + "loss_total_v6": 0.4570622481405735, + "step": 517 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11946666666666667, + "grad_norm": 24.343664169311523, + "learning_rate": 9.854848882939278e-06, + "loss": 7.4033403396606445, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.3214886710047722, + "loss_total_v6": 0.4627087563276291, + "step": 518 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11969729729729729, + "grad_norm": 27.172266006469727, + "learning_rate": 9.853939025224037e-06, + "loss": 7.242857456207275, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.31240454129874706, + "loss_total_v6": 0.452678594738245, + "step": 519 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.11992792792792793, + "grad_norm": 25.058486938476562, + "learning_rate": 9.853026367025535e-06, + "loss": 7.506157875061035, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.3284640833735466, + "loss_total_v6": 0.4691348746418953, + "step": 520 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12015855855855856, + "grad_norm": 23.573057174682617, + "learning_rate": 9.852110908870336e-06, + "loss": 7.166837692260742, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.3070277124643326, + "loss_total_v6": 0.447927363216877, + "step": 521 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12038918918918919, + "grad_norm": 24.67251968383789, + "learning_rate": 9.85119265128661e-06, + "loss": 7.088788032531738, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.30286674946546555, + "loss_total_v6": 0.44304925203323364, + "step": 522 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12061981981981983, + "grad_norm": 25.385372161865234, + "learning_rate": 9.850271594804147e-06, + "loss": 7.193750381469727, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.3091827891767025, + "loss_total_v6": 0.4496094137430191, + "step": 523 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12085045045045045, + "grad_norm": 23.572980880737305, + "learning_rate": 9.849347739954353e-06, + "loss": 7.413745403289795, + "loss_alignment": 0.28466796875, + "loss_alignment_w": 0.142333984375, + "loss_sft": 0.3213760443031788, + "loss_total_v6": 0.4633590877056122, + "step": 524 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12108108108108108, + "grad_norm": 24.196990966796875, + "learning_rate": 9.848421087270243e-06, + "loss": 7.31143856048584, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.3173317052423954, + "loss_total_v6": 0.4569648988544941, + "step": 525 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1213117117117117, + "grad_norm": 25.553979873657227, + "learning_rate": 9.847491637286455e-06, + "loss": 6.95958948135376, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.2963482663035393, + "loss_total_v6": 0.4349743612110615, + "step": 526 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12154234234234235, + "grad_norm": 23.520458221435547, + "learning_rate": 9.846559390539232e-06, + "loss": 7.470972537994385, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.32818761095404625, + "loss_total_v6": 0.46693577989935875, + "step": 527 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12177297297297297, + "grad_norm": 25.49165916442871, + "learning_rate": 9.845624347566434e-06, + "loss": 7.029649257659912, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.3002539500594139, + "loss_total_v6": 0.4393530674278736, + "step": 528 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1220036036036036, + "grad_norm": 29.377870559692383, + "learning_rate": 9.844686508907538e-06, + "loss": 7.303449630737305, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.3183125089854002, + "loss_total_v6": 0.45646557956933975, + "step": 529 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12223423423423424, + "grad_norm": 25.285655975341797, + "learning_rate": 9.843745875103628e-06, + "loss": 7.178755760192871, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.30838294699788094, + "loss_total_v6": 0.44867225363850594, + "step": 530 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12246486486486487, + "grad_norm": 23.645381927490234, + "learning_rate": 9.842802446697403e-06, + "loss": 7.1609039306640625, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.3092355653643608, + "loss_total_v6": 0.4475564807653427, + "step": 531 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1226954954954955, + "grad_norm": 23.80010223388672, + "learning_rate": 9.841856224233175e-06, + "loss": 7.03311824798584, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.3018745705485344, + "loss_total_v6": 0.4395698718726635, + "step": 532 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12292612612612612, + "grad_norm": 21.369426727294922, + "learning_rate": 9.840907208256869e-06, + "loss": 7.1811628341674805, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.30941840074956417, + "loss_total_v6": 0.4488226994872093, + "step": 533 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12315675675675676, + "grad_norm": 25.02350616455078, + "learning_rate": 9.83995539931602e-06, + "loss": 7.381142616271973, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.3194299302995205, + "loss_total_v6": 0.4613214284181595, + "step": 534 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12338738738738739, + "grad_norm": 27.01273536682129, + "learning_rate": 9.839000797959775e-06, + "loss": 7.399693489074707, + "loss_alignment": 0.283447265625, + "loss_alignment_w": 0.1417236328125, + "loss_sft": 0.32074197567999363, + "loss_total_v6": 0.4624808467924595, + "step": 535 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12361801801801801, + "grad_norm": 24.49747657775879, + "learning_rate": 9.838043404738889e-06, + "loss": 6.930263042449951, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.2945763822644949, + "loss_total_v6": 0.43314145132899284, + "step": 536 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12384864864864865, + "grad_norm": 24.571210861206055, + "learning_rate": 9.837083220205735e-06, + "loss": 6.974754810333252, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.2955260518938303, + "loss_total_v6": 0.43592218309640884, + "step": 537 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12407927927927928, + "grad_norm": 25.28897476196289, + "learning_rate": 9.83612024491429e-06, + "loss": 7.565377712249756, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.33215006813406944, + "loss_total_v6": 0.47283610701560974, + "step": 538 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12430990990990991, + "grad_norm": 23.484699249267578, + "learning_rate": 9.835154479420141e-06, + "loss": 6.781961441040039, + "loss_alignment": 0.285400390625, + "loss_alignment_w": 0.1427001953125, + "loss_sft": 0.28150810673832893, + "loss_total_v6": 0.42387260496616364, + "step": 539 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12454054054054053, + "grad_norm": 24.384992599487305, + "learning_rate": 9.83418592428049e-06, + "loss": 7.648437976837158, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.3378296084702015, + "loss_total_v6": 0.4780273661017418, + "step": 540 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12477117117117117, + "grad_norm": 28.004487991333008, + "learning_rate": 9.833214580054145e-06, + "loss": 6.810153961181641, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.28475018963217735, + "loss_total_v6": 0.42563461512327194, + "step": 541 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12500180180180182, + "grad_norm": 23.4300537109375, + "learning_rate": 9.832240447301523e-06, + "loss": 6.903972625732422, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.29282641410827637, + "loss_total_v6": 0.43149829283356667, + "step": 542 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12523243243243243, + "grad_norm": 24.34965705871582, + "learning_rate": 9.831263526584651e-06, + "loss": 7.299942493438721, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.31725407391786575, + "loss_total_v6": 0.45624638721346855, + "step": 543 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12546306306306307, + "grad_norm": 24.501039505004883, + "learning_rate": 9.830283818467164e-06, + "loss": 7.339157581329346, + "loss_alignment": 0.283203125, + "loss_alignment_w": 0.1416015625, + "loss_sft": 0.3170194737613201, + "loss_total_v6": 0.4586973413825035, + "step": 544 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12569369369369368, + "grad_norm": 25.941402435302734, + "learning_rate": 9.829301323514307e-06, + "loss": 7.669066905975342, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.34052273631095886, + "loss_total_v6": 0.47931668534874916, + "step": 545 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12592432432432432, + "grad_norm": 24.095489501953125, + "learning_rate": 9.828316042292927e-06, + "loss": 7.238667011260986, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.31264616549015045, + "loss_total_v6": 0.45241668075323105, + "step": 546 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12615495495495496, + "grad_norm": 22.56650161743164, + "learning_rate": 9.827327975371485e-06, + "loss": 7.780430793762207, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.34690314903855324, + "loss_total_v6": 0.48627692088484764, + "step": 547 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12638558558558557, + "grad_norm": 23.67683219909668, + "learning_rate": 9.826337123320046e-06, + "loss": 7.178365230560303, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.3082974776625633, + "loss_total_v6": 0.448647815734148, + "step": 548 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12661621621621622, + "grad_norm": 23.922574996948242, + "learning_rate": 9.825343486710286e-06, + "loss": 7.018298625946045, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.2975303716957569, + "loss_total_v6": 0.4386436641216278, + "step": 549 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12684684684684686, + "grad_norm": 24.983787536621094, + "learning_rate": 9.82434706611548e-06, + "loss": 7.129955291748047, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.30490563064813614, + "loss_total_v6": 0.44562218710780144, + "step": 550 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12707747747747747, + "grad_norm": 22.955747604370117, + "learning_rate": 9.823347862110515e-06, + "loss": 7.214948654174805, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.3111485317349434, + "loss_total_v6": 0.4509342908859253, + "step": 551 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1273081081081081, + "grad_norm": 26.606843948364258, + "learning_rate": 9.822345875271884e-06, + "loss": 7.318840503692627, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3188319280743599, + "loss_total_v6": 0.4574275277554989, + "step": 552 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12753873873873875, + "grad_norm": 25.651437759399414, + "learning_rate": 9.821341106177681e-06, + "loss": 7.320440292358398, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.3180469162762165, + "loss_total_v6": 0.4575275108218193, + "step": 553 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12776936936936936, + "grad_norm": 25.99298095703125, + "learning_rate": 9.820333555407611e-06, + "loss": 7.166207790374756, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.30697305500507355, + "loss_total_v6": 0.44788798689842224, + "step": 554 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.128, + "grad_norm": 25.134418487548828, + "learning_rate": 9.81932322354298e-06, + "loss": 7.5027923583984375, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.32781121507287025, + "loss_total_v6": 0.46892450377345085, + "step": 555 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12823063063063064, + "grad_norm": 24.632843017578125, + "learning_rate": 9.818310111166699e-06, + "loss": 7.475839614868164, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.32669126242399216, + "loss_total_v6": 0.46723996102809906, + "step": 556 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12846126126126126, + "grad_norm": 25.242271423339844, + "learning_rate": 9.817294218863285e-06, + "loss": 7.33851957321167, + "loss_alignment": 0.28173828125, + "loss_alignment_w": 0.140869140625, + "loss_sft": 0.31804772838950157, + "loss_total_v6": 0.45865747332572937, + "step": 557 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1286918918918919, + "grad_norm": 27.152273178100586, + "learning_rate": 9.816275547218855e-06, + "loss": 7.3658928871154785, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.319789070636034, + "loss_total_v6": 0.4603683091700077, + "step": 558 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1289225225225225, + "grad_norm": 23.608257293701172, + "learning_rate": 9.815254096821136e-06, + "loss": 7.073732376098633, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.3027039933949709, + "loss_total_v6": 0.44210827723145485, + "step": 559 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12915315315315315, + "grad_norm": 24.145742416381836, + "learning_rate": 9.814229868259453e-06, + "loss": 7.114845275878906, + "loss_alignment": 0.283203125, + "loss_alignment_w": 0.1416015625, + "loss_sft": 0.30316781252622604, + "loss_total_v6": 0.44467782601714134, + "step": 560 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1293837837837838, + "grad_norm": 23.814790725708008, + "learning_rate": 9.813202862124735e-06, + "loss": 6.956794738769531, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.29939316026866436, + "loss_total_v6": 0.4347996562719345, + "step": 561 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1296144144144144, + "grad_norm": 25.302284240722656, + "learning_rate": 9.812173079009512e-06, + "loss": 7.1258416175842285, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.3051673509180546, + "loss_total_v6": 0.4453651010990143, + "step": 562 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.12984504504504504, + "grad_norm": 24.49597930908203, + "learning_rate": 9.811140519507922e-06, + "loss": 7.25911808013916, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.3133140541613102, + "loss_total_v6": 0.4536948911845684, + "step": 563 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13007567567567568, + "grad_norm": 24.400983810424805, + "learning_rate": 9.8101051842157e-06, + "loss": 7.268233299255371, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.3142194114625454, + "loss_total_v6": 0.4542645700275898, + "step": 564 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1303063063063063, + "grad_norm": 26.40557861328125, + "learning_rate": 9.809067073730178e-06, + "loss": 7.674243927001953, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.33898473158478737, + "loss_total_v6": 0.47964026406407356, + "step": 565 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13053693693693694, + "grad_norm": 21.50457000732422, + "learning_rate": 9.808026188650303e-06, + "loss": 7.397364139556885, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.3216645009815693, + "loss_total_v6": 0.4623352661728859, + "step": 566 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13076756756756758, + "grad_norm": 23.6086368560791, + "learning_rate": 9.806982529576608e-06, + "loss": 6.964539527893066, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.29888540878891945, + "loss_total_v6": 0.43528374284505844, + "step": 567 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1309981981981982, + "grad_norm": 25.27796173095703, + "learning_rate": 9.805936097111235e-06, + "loss": 7.392313003540039, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.32218800857663155, + "loss_total_v6": 0.46201955154538155, + "step": 568 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13122882882882883, + "grad_norm": 24.37344741821289, + "learning_rate": 9.804886891857924e-06, + "loss": 7.611758232116699, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.3341180570423603, + "loss_total_v6": 0.4757348820567131, + "step": 569 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13145945945945947, + "grad_norm": 25.703598022460938, + "learning_rate": 9.803834914422015e-06, + "loss": 7.361396789550781, + "loss_alignment": 0.285888671875, + "loss_alignment_w": 0.1429443359375, + "loss_sft": 0.31743287295103073, + "loss_total_v6": 0.46008728817105293, + "step": 570 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13169009009009008, + "grad_norm": 25.071077346801758, + "learning_rate": 9.802780165410446e-06, + "loss": 7.390101432800293, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.3220498152077198, + "loss_total_v6": 0.4618813619017601, + "step": 571 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13192072072072072, + "grad_norm": 24.10201644897461, + "learning_rate": 9.801722645431755e-06, + "loss": 7.255832672119141, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3154280073940754, + "loss_total_v6": 0.4534895457327366, + "step": 572 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13215135135135136, + "grad_norm": 24.61787986755371, + "learning_rate": 9.80066235509608e-06, + "loss": 7.286660194396973, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.31640871800482273, + "loss_total_v6": 0.455416277050972, + "step": 573 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13238198198198198, + "grad_norm": 25.33001136779785, + "learning_rate": 9.799599295015154e-06, + "loss": 7.1044769287109375, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.3046407960355282, + "loss_total_v6": 0.4440298266708851, + "step": 574 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13261261261261262, + "grad_norm": 29.564586639404297, + "learning_rate": 9.798533465802313e-06, + "loss": 7.6014814376831055, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.334147110581398, + "loss_total_v6": 0.4750925563275814, + "step": 575 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13284324324324323, + "grad_norm": 24.116016387939453, + "learning_rate": 9.797464868072489e-06, + "loss": 7.290192604064941, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.314950967207551, + "loss_total_v6": 0.45563701540231705, + "step": 576 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13307387387387387, + "grad_norm": 23.431631088256836, + "learning_rate": 9.796393502442205e-06, + "loss": 7.128287315368652, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.30548803322017193, + "loss_total_v6": 0.445517934858799, + "step": 577 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1333045045045045, + "grad_norm": 22.85860824584961, + "learning_rate": 9.795319369529588e-06, + "loss": 7.324368476867676, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.32222918421030045, + "loss_total_v6": 0.45777301117777824, + "step": 578 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13353513513513512, + "grad_norm": 26.072036743164062, + "learning_rate": 9.794242469954362e-06, + "loss": 7.271856307983398, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3172992542386055, + "loss_total_v6": 0.4544910304248333, + "step": 579 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13376576576576577, + "grad_norm": 24.88744354248047, + "learning_rate": 9.793162804337844e-06, + "loss": 7.583467483520508, + "loss_alignment": 0.28173828125, + "loss_alignment_w": 0.140869140625, + "loss_sft": 0.3329908028244972, + "loss_total_v6": 0.4739667437970638, + "step": 580 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1339963963963964, + "grad_norm": 26.160608291625977, + "learning_rate": 9.792080373302949e-06, + "loss": 6.9948835372924805, + "loss_alignment": 0.278076171875, + "loss_alignment_w": 0.1390380859375, + "loss_sft": 0.2983405329287052, + "loss_total_v6": 0.4371802546083927, + "step": 581 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13422702702702702, + "grad_norm": 25.593055725097656, + "learning_rate": 9.790995177474181e-06, + "loss": 7.4038543701171875, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.3226957395672798, + "loss_total_v6": 0.4627408981323242, + "step": 582 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13445765765765766, + "grad_norm": 25.2871150970459, + "learning_rate": 9.789907217477649e-06, + "loss": 7.316670894622803, + "loss_alignment": 0.283935546875, + "loss_alignment_w": 0.1419677734375, + "loss_sft": 0.31553778052330017, + "loss_total_v6": 0.4572919234633446, + "step": 583 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1346882882882883, + "grad_norm": 23.50116539001465, + "learning_rate": 9.788816493941052e-06, + "loss": 6.777833938598633, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.285202132537961, + "loss_total_v6": 0.42361461743712425, + "step": 584 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1349189189189189, + "grad_norm": 24.03115463256836, + "learning_rate": 9.787723007493681e-06, + "loss": 7.488104820251465, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.32642025128006935, + "loss_total_v6": 0.46800655499100685, + "step": 585 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13514954954954955, + "grad_norm": 25.08713722229004, + "learning_rate": 9.786626758766427e-06, + "loss": 7.1126556396484375, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.30618954822421074, + "loss_total_v6": 0.44454099610447884, + "step": 586 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1353801801801802, + "grad_norm": 26.628889083862305, + "learning_rate": 9.78552774839177e-06, + "loss": 7.282400131225586, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.3176377862691879, + "loss_total_v6": 0.4551499933004379, + "step": 587 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1356108108108108, + "grad_norm": 25.362333297729492, + "learning_rate": 9.784425977003786e-06, + "loss": 7.0987935066223145, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.30317166447639465, + "loss_total_v6": 0.44367459416389465, + "step": 588 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13584144144144145, + "grad_norm": 23.691160202026367, + "learning_rate": 9.783321445238138e-06, + "loss": 6.5061540603637695, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.27051097340881824, + "loss_total_v6": 0.4066346250474453, + "step": 589 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13607207207207206, + "grad_norm": 24.630022048950195, + "learning_rate": 9.782214153732093e-06, + "loss": 7.159619331359863, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3125732559710741, + "loss_total_v6": 0.44747620075941086, + "step": 590 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1363027027027027, + "grad_norm": 22.87793731689453, + "learning_rate": 9.781104103124499e-06, + "loss": 7.024153709411621, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.3000020682811737, + "loss_total_v6": 0.439009640365839, + "step": 591 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13653333333333334, + "grad_norm": 25.289445877075195, + "learning_rate": 9.779991294055802e-06, + "loss": 7.52552604675293, + "loss_alignment": 0.285888671875, + "loss_alignment_w": 0.1429443359375, + "loss_sft": 0.3275383282452822, + "loss_total_v6": 0.4703453481197357, + "step": 592 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13676396396396395, + "grad_norm": 22.494972229003906, + "learning_rate": 9.778875727168038e-06, + "loss": 7.175385475158691, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.3108273111283779, + "loss_total_v6": 0.4484615884721279, + "step": 593 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1369945945945946, + "grad_norm": 24.575775146484375, + "learning_rate": 9.777757403104833e-06, + "loss": 7.219342231750488, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.31537511944770813, + "loss_total_v6": 0.4512088745832443, + "step": 594 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13722522522522523, + "grad_norm": 23.186786651611328, + "learning_rate": 9.776636322511405e-06, + "loss": 7.265187740325928, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.317950576543808, + "loss_total_v6": 0.4540742300450802, + "step": 595 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13745585585585585, + "grad_norm": 29.51439094543457, + "learning_rate": 9.775512486034564e-06, + "loss": 6.990046501159668, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.29874005913734436, + "loss_total_v6": 0.43687789514660835, + "step": 596 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1376864864864865, + "grad_norm": 27.567359924316406, + "learning_rate": 9.774385894322705e-06, + "loss": 7.288619041442871, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.31701940298080444, + "loss_total_v6": 0.45553868263959885, + "step": 597 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13791711711711713, + "grad_norm": 29.76551628112793, + "learning_rate": 9.773256548025817e-06, + "loss": 7.420660972595215, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.325272049754858, + "loss_total_v6": 0.4637913256883621, + "step": 598 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13814774774774774, + "grad_norm": 22.482343673706055, + "learning_rate": 9.772124447795482e-06, + "loss": 7.367092132568359, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.31986401975154877, + "loss_total_v6": 0.46044325456023216, + "step": 599 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13837837837837838, + "grad_norm": 28.892227172851562, + "learning_rate": 9.770989594284858e-06, + "loss": 7.214008331298828, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.30901454761624336, + "loss_total_v6": 0.45087552443146706, + "step": 600 + }, + { + "epoch": 0.13837837837837838, + "eval_loss": 0.6048235893249512, + "eval_loss_alignment": 0.251244113869863, + "eval_loss_alignment_w": 0.1256220569349315, + "eval_loss_sft": 0.47920142797982856, + "eval_loss_total_v6": 0.6048234849317705, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.8127, + "eval_samples_per_second": 19.006, + "eval_steps_per_second": 2.385, + "step": 600 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13860900900900902, + "grad_norm": 22.69908332824707, + "learning_rate": 9.769851988148704e-06, + "loss": 7.50676155090332, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.32836446538567543, + "loss_total_v6": 0.4691725820302963, + "step": 601 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13883963963963963, + "grad_norm": 24.832744598388672, + "learning_rate": 9.768711630043362e-06, + "loss": 7.1578779220581055, + "loss_alignment": 0.28173828125, + "loss_alignment_w": 0.140869140625, + "loss_sft": 0.30691024847328663, + "loss_total_v6": 0.4473673962056637, + "step": 602 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13907027027027027, + "grad_norm": 26.935569763183594, + "learning_rate": 9.767568520626762e-06, + "loss": 7.319861888885498, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3167442958801985, + "loss_total_v6": 0.45749136805534363, + "step": 603 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13930090090090091, + "grad_norm": 27.019351959228516, + "learning_rate": 9.76642266055842e-06, + "loss": 7.4204816818237305, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.32556599751114845, + "loss_total_v6": 0.46378009393811226, + "step": 604 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13953153153153153, + "grad_norm": 28.520618438720703, + "learning_rate": 9.765274050499445e-06, + "loss": 6.932879447937012, + "loss_alignment": 0.2841796875, + "loss_alignment_w": 0.14208984375, + "loss_sft": 0.29113882780075073, + "loss_total_v6": 0.43330497294664383, + "step": 605 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13976216216216217, + "grad_norm": 26.83087158203125, + "learning_rate": 9.764122691112525e-06, + "loss": 6.6591315269470215, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.2774780709296465, + "loss_total_v6": 0.41619570553302765, + "step": 606 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.13999279279279278, + "grad_norm": 24.29120445251465, + "learning_rate": 9.762968583061938e-06, + "loss": 7.674859523773193, + "loss_alignment": 0.285888671875, + "loss_alignment_w": 0.1429443359375, + "loss_sft": 0.3370090387761593, + "loss_total_v6": 0.4796787239611149, + "step": 607 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14022342342342342, + "grad_norm": 25.735872268676758, + "learning_rate": 9.761811727013548e-06, + "loss": 6.782725811004639, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.2855689190328121, + "loss_total_v6": 0.4239203706383705, + "step": 608 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14045405405405406, + "grad_norm": 24.375347137451172, + "learning_rate": 9.760652123634804e-06, + "loss": 7.107718467712402, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.30668970942497253, + "loss_total_v6": 0.44423241540789604, + "step": 609 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14068468468468467, + "grad_norm": 22.990690231323242, + "learning_rate": 9.759489773594739e-06, + "loss": 7.157663822174072, + "loss_alignment": 0.283935546875, + "loss_alignment_w": 0.1419677734375, + "loss_sft": 0.305554062128067, + "loss_total_v6": 0.4473539926111698, + "step": 610 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14091531531531531, + "grad_norm": 24.25033187866211, + "learning_rate": 9.758324677563972e-06, + "loss": 7.224820137023926, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.31222326681017876, + "loss_total_v6": 0.45155125111341476, + "step": 611 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14114594594594596, + "grad_norm": 23.99294090270996, + "learning_rate": 9.757156836214707e-06, + "loss": 7.828352928161621, + "loss_alignment": 0.28466796875, + "loss_alignment_w": 0.142333984375, + "loss_sft": 0.3467397168278694, + "loss_total_v6": 0.4892720505595207, + "step": 612 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14137657657657657, + "grad_norm": 26.26667594909668, + "learning_rate": 9.755986250220728e-06, + "loss": 7.76433801651001, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.3465687297284603, + "loss_total_v6": 0.4852711297571659, + "step": 613 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1416072072072072, + "grad_norm": 26.801095962524414, + "learning_rate": 9.754812920257409e-06, + "loss": 7.20583438873291, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3147292695939541, + "loss_total_v6": 0.4503646343946457, + "step": 614 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14183783783783785, + "grad_norm": 24.829845428466797, + "learning_rate": 9.753636847001701e-06, + "loss": 7.585324287414551, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.3347089923918247, + "loss_total_v6": 0.47408274561166763, + "step": 615 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14206846846846846, + "grad_norm": 23.415267944335938, + "learning_rate": 9.752458031132142e-06, + "loss": 7.324220657348633, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.31706248223781586, + "loss_total_v6": 0.45776378363370895, + "step": 616 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1422990990990991, + "grad_norm": 23.717805862426758, + "learning_rate": 9.751276473328848e-06, + "loss": 7.8838911056518555, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.35300322994589806, + "loss_total_v6": 0.49274320527911186, + "step": 617 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14252972972972974, + "grad_norm": 24.5263671875, + "learning_rate": 9.75009217427352e-06, + "loss": 6.84896993637085, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.2903805933892727, + "loss_total_v6": 0.428060632199049, + "step": 618 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14276036036036036, + "grad_norm": 24.662988662719727, + "learning_rate": 9.748905134649444e-06, + "loss": 7.124758720397949, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.30540481209754944, + "loss_total_v6": 0.44529738649725914, + "step": 619 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.142990990990991, + "grad_norm": 24.322162628173828, + "learning_rate": 9.747715355141478e-06, + "loss": 7.041816711425781, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.30223512649536133, + "loss_total_v6": 0.4401135630905628, + "step": 620 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1432216216216216, + "grad_norm": 23.379121780395508, + "learning_rate": 9.74652283643607e-06, + "loss": 7.414329528808594, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.3235335610806942, + "loss_total_v6": 0.463395606726408, + "step": 621 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14345225225225225, + "grad_norm": 25.345773696899414, + "learning_rate": 9.745327579221241e-06, + "loss": 7.289071083068848, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3150181993842125, + "loss_total_v6": 0.4555669203400612, + "step": 622 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1436828828828829, + "grad_norm": 28.26356315612793, + "learning_rate": 9.744129584186599e-06, + "loss": 7.272341728210449, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.314399890601635, + "loss_total_v6": 0.4545213580131531, + "step": 623 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1439135135135135, + "grad_norm": 25.51447105407715, + "learning_rate": 9.742928852023326e-06, + "loss": 7.466243743896484, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.32529808953404427, + "loss_total_v6": 0.46664024144411087, + "step": 624 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14414414414414414, + "grad_norm": 21.978727340698242, + "learning_rate": 9.741725383424185e-06, + "loss": 7.168520927429199, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.30865875631570816, + "loss_total_v6": 0.44803252816200256, + "step": 625 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14437477477477478, + "grad_norm": 23.660320281982422, + "learning_rate": 9.740519179083519e-06, + "loss": 7.082430362701416, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3043004311621189, + "loss_total_v6": 0.4426518902182579, + "step": 626 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1446054054054054, + "grad_norm": 25.990367889404297, + "learning_rate": 9.739310239697248e-06, + "loss": 7.942398548126221, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.3563699796795845, + "loss_total_v6": 0.4963998980820179, + "step": 627 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14483603603603604, + "grad_norm": 25.822603225708008, + "learning_rate": 9.73809856596287e-06, + "loss": 7.624983787536621, + "loss_alignment": 0.278076171875, + "loss_alignment_w": 0.1390380859375, + "loss_sft": 0.3374471142888069, + "loss_total_v6": 0.4765614904463291, + "step": 628 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14506666666666668, + "grad_norm": 25.213376998901367, + "learning_rate": 9.73688415857946e-06, + "loss": 7.535830974578857, + "loss_alignment": 0.283203125, + "loss_alignment_w": 0.1416015625, + "loss_sft": 0.329601489007473, + "loss_total_v6": 0.470989428460598, + "step": 629 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1452972972972973, + "grad_norm": 24.462278366088867, + "learning_rate": 9.735667018247674e-06, + "loss": 7.066629409790039, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3055101502686739, + "loss_total_v6": 0.44166431948542595, + "step": 630 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14552792792792793, + "grad_norm": 23.26820945739746, + "learning_rate": 9.73444714566974e-06, + "loss": 7.149508476257324, + "loss_alignment": 0.28369140625, + "loss_alignment_w": 0.141845703125, + "loss_sft": 0.30496807768940926, + "loss_total_v6": 0.44684429094195366, + "step": 631 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14575855855855857, + "grad_norm": 24.380151748657227, + "learning_rate": 9.733224541549464e-06, + "loss": 7.135506629943848, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.30584774166345596, + "loss_total_v6": 0.44596919044852257, + "step": 632 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14598918918918918, + "grad_norm": 23.655715942382812, + "learning_rate": 9.731999206592229e-06, + "loss": 6.951894283294678, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.29731688648462296, + "loss_total_v6": 0.43449339643120766, + "step": 633 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14621981981981982, + "grad_norm": 24.47887420654297, + "learning_rate": 9.730771141504993e-06, + "loss": 6.850419044494629, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.2887773811817169, + "loss_total_v6": 0.4281511791050434, + "step": 634 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14645045045045044, + "grad_norm": 25.70744514465332, + "learning_rate": 9.729540346996286e-06, + "loss": 7.396149635314941, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.32421309873461723, + "loss_total_v6": 0.46225937083363533, + "step": 635 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14668108108108108, + "grad_norm": 24.531143188476562, + "learning_rate": 9.72830682377622e-06, + "loss": 7.137683868408203, + "loss_alignment": 0.2750244140625, + "loss_alignment_w": 0.13751220703125, + "loss_sft": 0.3083946444094181, + "loss_total_v6": 0.4461052194237709, + "step": 636 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14691171171171172, + "grad_norm": 23.53561019897461, + "learning_rate": 9.727070572556476e-06, + "loss": 7.111209869384766, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.30678581446409225, + "loss_total_v6": 0.44445061683654785, + "step": 637 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14714234234234233, + "grad_norm": 25.175569534301758, + "learning_rate": 9.725831594050307e-06, + "loss": 7.43409538269043, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.32489099353551865, + "loss_total_v6": 0.46463097631931305, + "step": 638 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14737297297297297, + "grad_norm": 23.14017677307129, + "learning_rate": 9.724589888972547e-06, + "loss": 6.942498207092285, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.29837758652865887, + "loss_total_v6": 0.4339061304926872, + "step": 639 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1476036036036036, + "grad_norm": 24.992450714111328, + "learning_rate": 9.723345458039595e-06, + "loss": 6.997921943664551, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.2990186717361212, + "loss_total_v6": 0.4373701177537441, + "step": 640 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14783423423423422, + "grad_norm": 23.712621688842773, + "learning_rate": 9.722098301969427e-06, + "loss": 7.094903945922852, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.30608711391687393, + "loss_total_v6": 0.4434314854443073, + "step": 641 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14806486486486486, + "grad_norm": 25.338111877441406, + "learning_rate": 9.720848421481592e-06, + "loss": 6.948931694030762, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.2958957441151142, + "loss_total_v6": 0.434308223426342, + "step": 642 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1482954954954955, + "grad_norm": 25.418930053710938, + "learning_rate": 9.719595817297207e-06, + "loss": 7.126823425292969, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.30362654849886894, + "loss_total_v6": 0.44542647525668144, + "step": 643 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14852612612612612, + "grad_norm": 23.524860382080078, + "learning_rate": 9.718340490138965e-06, + "loss": 7.1304168701171875, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.3067960552871227, + "loss_total_v6": 0.4456510432064533, + "step": 644 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14875675675675676, + "grad_norm": 23.00027084350586, + "learning_rate": 9.717082440731125e-06, + "loss": 7.288444995880127, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.31697797775268555, + "loss_total_v6": 0.45552779734134674, + "step": 645 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1489873873873874, + "grad_norm": 24.172636032104492, + "learning_rate": 9.715821669799521e-06, + "loss": 6.969738960266113, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.2988746725022793, + "loss_total_v6": 0.4356086924672127, + "step": 646 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.149218018018018, + "grad_norm": 22.739334106445312, + "learning_rate": 9.714558178071559e-06, + "loss": 7.304135322570801, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.31789760664105415, + "loss_total_v6": 0.45650845393538475, + "step": 647 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.14944864864864865, + "grad_norm": 22.761573791503906, + "learning_rate": 9.713291966276205e-06, + "loss": 6.946493148803711, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.29522455483675003, + "loss_total_v6": 0.43415582180023193, + "step": 648 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1496792792792793, + "grad_norm": 23.15144157409668, + "learning_rate": 9.712023035144004e-06, + "loss": 6.979101181030273, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.298971526324749, + "loss_total_v6": 0.4361938200891018, + "step": 649 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1499099099099099, + "grad_norm": 22.35938835144043, + "learning_rate": 9.710751385407068e-06, + "loss": 7.274543762207031, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.31343893334269524, + "loss_total_v6": 0.45465901121497154, + "step": 650 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15014054054054055, + "grad_norm": 24.629899978637695, + "learning_rate": 9.709477017799076e-06, + "loss": 6.905644416809082, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.2942889332771301, + "loss_total_v6": 0.4316027872264385, + "step": 651 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15037117117117116, + "grad_norm": 25.463655471801758, + "learning_rate": 9.708199933055272e-06, + "loss": 6.733314037322998, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.2852577716112137, + "loss_total_v6": 0.4208321273326874, + "step": 652 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1506018018018018, + "grad_norm": 24.753475189208984, + "learning_rate": 9.706920131912473e-06, + "loss": 6.858580589294434, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.29009624011814594, + "loss_total_v6": 0.4286612942814827, + "step": 653 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15083243243243244, + "grad_norm": 25.291744232177734, + "learning_rate": 9.70563761510906e-06, + "loss": 7.209812164306641, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.31308577954769135, + "loss_total_v6": 0.45061324909329414, + "step": 654 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15106306306306305, + "grad_norm": 25.356306076049805, + "learning_rate": 9.704352383384984e-06, + "loss": 7.439563751220703, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.32497335597872734, + "loss_total_v6": 0.46497273072600365, + "step": 655 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1512936936936937, + "grad_norm": 27.174097061157227, + "learning_rate": 9.70306443748176e-06, + "loss": 7.325691223144531, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.3196873627603054, + "loss_total_v6": 0.4578556977212429, + "step": 656 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15152432432432433, + "grad_norm": 24.51626968383789, + "learning_rate": 9.701773778142467e-06, + "loss": 7.08250617980957, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.30557164922356606, + "loss_total_v6": 0.44265661388635635, + "step": 657 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15175495495495495, + "grad_norm": 25.665172576904297, + "learning_rate": 9.700480406111752e-06, + "loss": 7.4215922355651855, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.3265509307384491, + "loss_total_v6": 0.4638495109975338, + "step": 658 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1519855855855856, + "grad_norm": 22.829010009765625, + "learning_rate": 9.69918432213583e-06, + "loss": 7.920661449432373, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.3546147011220455, + "loss_total_v6": 0.495041336864233, + "step": 659 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15221621621621623, + "grad_norm": 24.41936492919922, + "learning_rate": 9.697885526962474e-06, + "loss": 7.637414932250977, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.33680498227477074, + "loss_total_v6": 0.47733843326568604, + "step": 660 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15244684684684684, + "grad_norm": 38.65446472167969, + "learning_rate": 9.696584021341025e-06, + "loss": 7.041236400604248, + "loss_alignment": 0.284912109375, + "loss_alignment_w": 0.1424560546875, + "loss_sft": 0.29752968065440655, + "loss_total_v6": 0.4400772713124752, + "step": 661 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15267747747747748, + "grad_norm": 26.688949584960938, + "learning_rate": 9.695279806022391e-06, + "loss": 7.493835926055908, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.32773976027965546, + "loss_total_v6": 0.46836475655436516, + "step": 662 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15290810810810812, + "grad_norm": 29.080453872680664, + "learning_rate": 9.693972881759036e-06, + "loss": 7.221242904663086, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.313327182084322, + "loss_total_v6": 0.451327670365572, + "step": 663 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15313873873873873, + "grad_norm": 29.382230758666992, + "learning_rate": 9.692663249304992e-06, + "loss": 7.375823974609375, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.32123372331261635, + "loss_total_v6": 0.46098897978663445, + "step": 664 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15336936936936937, + "grad_norm": 25.866748809814453, + "learning_rate": 9.691350909415853e-06, + "loss": 7.286056995391846, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.31608108058571815, + "loss_total_v6": 0.45537856593728065, + "step": 665 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1536, + "grad_norm": 25.871395111083984, + "learning_rate": 9.690035862848773e-06, + "loss": 7.0231218338012695, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.300257982686162, + "loss_total_v6": 0.43894509971141815, + "step": 666 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15383063063063063, + "grad_norm": 25.645320892333984, + "learning_rate": 9.688718110362472e-06, + "loss": 7.565120697021484, + "loss_alignment": 0.28173828125, + "loss_alignment_w": 0.140869140625, + "loss_sft": 0.33214927837252617, + "loss_total_v6": 0.4728200361132622, + "step": 667 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15406126126126127, + "grad_norm": 23.29018211364746, + "learning_rate": 9.687397652717223e-06, + "loss": 7.593719959259033, + "loss_alignment": 0.281982421875, + "loss_alignment_w": 0.1409912109375, + "loss_sft": 0.33378414064645767, + "loss_total_v6": 0.47460751235485077, + "step": 668 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15429189189189188, + "grad_norm": 23.537490844726562, + "learning_rate": 9.686074490674871e-06, + "loss": 7.4607038497924805, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.3266913518309593, + "loss_total_v6": 0.46629399433732033, + "step": 669 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15452252252252252, + "grad_norm": 23.779277801513672, + "learning_rate": 9.68474862499881e-06, + "loss": 7.512609004974365, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3331702724099159, + "loss_total_v6": 0.46953805908560753, + "step": 670 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15475315315315316, + "grad_norm": 23.180543899536133, + "learning_rate": 9.683420056454e-06, + "loss": 7.461039066314697, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.3263918124139309, + "loss_total_v6": 0.4663149416446686, + "step": 671 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15498378378378377, + "grad_norm": 25.869205474853516, + "learning_rate": 9.682088785806964e-06, + "loss": 7.384243488311768, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.32403354719281197, + "loss_total_v6": 0.46151522174477577, + "step": 672 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15521441441441441, + "grad_norm": 22.937070846557617, + "learning_rate": 9.680754813825774e-06, + "loss": 7.540336608886719, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.334521759301424, + "loss_total_v6": 0.4712710455060005, + "step": 673 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15544504504504505, + "grad_norm": 24.235013961791992, + "learning_rate": 9.679418141280068e-06, + "loss": 6.864472389221191, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.28895380161702633, + "loss_total_v6": 0.4290294982492924, + "step": 674 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15567567567567567, + "grad_norm": 25.842147827148438, + "learning_rate": 9.678078768941039e-06, + "loss": 7.335341930389404, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.32257934659719467, + "loss_total_v6": 0.45845886692404747, + "step": 675 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1559063063063063, + "grad_norm": 23.880857467651367, + "learning_rate": 9.676736697581441e-06, + "loss": 7.572358131408691, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.3350277803838253, + "loss_total_v6": 0.4732724018394947, + "step": 676 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15613693693693695, + "grad_norm": 23.809852600097656, + "learning_rate": 9.675391927975584e-06, + "loss": 7.264443397521973, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.3136926330626011, + "loss_total_v6": 0.4540277235209942, + "step": 677 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15636756756756756, + "grad_norm": 24.03036880493164, + "learning_rate": 9.674044460899329e-06, + "loss": 7.216693878173828, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.3163998071104288, + "loss_total_v6": 0.45104335620999336, + "step": 678 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1565981981981982, + "grad_norm": 27.848663330078125, + "learning_rate": 9.672694297130101e-06, + "loss": 7.327932834625244, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.31856097280979156, + "loss_total_v6": 0.45799579471349716, + "step": 679 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15682882882882884, + "grad_norm": 29.646162033081055, + "learning_rate": 9.671341437446878e-06, + "loss": 7.895075798034668, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.35232895612716675, + "loss_total_v6": 0.49344225600361824, + "step": 680 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15705945945945946, + "grad_norm": 27.668930053710938, + "learning_rate": 9.66998588263019e-06, + "loss": 7.29323148727417, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3173229470849037, + "loss_total_v6": 0.455826960504055, + "step": 681 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1572900900900901, + "grad_norm": 28.093029022216797, + "learning_rate": 9.66862763346213e-06, + "loss": 7.3688225746154785, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.3226272203028202, + "loss_total_v6": 0.4605514258146286, + "step": 682 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1575207207207207, + "grad_norm": 23.661611557006836, + "learning_rate": 9.66726669072634e-06, + "loss": 7.391627311706543, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.32435767352581024, + "loss_total_v6": 0.46197669953107834, + "step": 683 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15775135135135135, + "grad_norm": 25.109210968017578, + "learning_rate": 9.665903055208013e-06, + "loss": 7.257603645324707, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.31352453865110874, + "loss_total_v6": 0.4536002241075039, + "step": 684 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.157981981981982, + "grad_norm": 24.72564125061035, + "learning_rate": 9.664536727693903e-06, + "loss": 7.090549468994141, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.3048536591231823, + "loss_total_v6": 0.4431593343615532, + "step": 685 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1582126126126126, + "grad_norm": 25.69887351989746, + "learning_rate": 9.663167708972312e-06, + "loss": 7.07271146774292, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3049900010228157, + "loss_total_v6": 0.4420444592833519, + "step": 686 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15844324324324324, + "grad_norm": 27.18357276916504, + "learning_rate": 9.661795999833098e-06, + "loss": 7.173650741577148, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.31238211132586, + "loss_total_v6": 0.4483531787991524, + "step": 687 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15867387387387388, + "grad_norm": 26.454753875732422, + "learning_rate": 9.660421601067667e-06, + "loss": 7.485564231872559, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.3291454091668129, + "loss_total_v6": 0.4678477942943573, + "step": 688 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1589045045045045, + "grad_norm": 26.89518165588379, + "learning_rate": 9.659044513468978e-06, + "loss": 7.350733757019043, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.3196961395442486, + "loss_total_v6": 0.4594208709895611, + "step": 689 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15913513513513514, + "grad_norm": 26.13521957397461, + "learning_rate": 9.657664737831548e-06, + "loss": 7.3694915771484375, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.32294370234012604, + "loss_total_v6": 0.46059323847293854, + "step": 690 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15936576576576578, + "grad_norm": 25.56412696838379, + "learning_rate": 9.656282274951433e-06, + "loss": 6.975528717041016, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.29694772884249687, + "loss_total_v6": 0.43597056716680527, + "step": 691 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1595963963963964, + "grad_norm": 24.366300582885742, + "learning_rate": 9.654897125626253e-06, + "loss": 6.879057884216309, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.2900332696735859, + "loss_total_v6": 0.4299411103129387, + "step": 692 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.15982702702702703, + "grad_norm": 23.01405143737793, + "learning_rate": 9.653509290655166e-06, + "loss": 6.985747337341309, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.2974948473274708, + "loss_total_v6": 0.436609223484993, + "step": 693 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16005765765765767, + "grad_norm": 23.413570404052734, + "learning_rate": 9.652118770838884e-06, + "loss": 6.872347831726074, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.29640408232808113, + "loss_total_v6": 0.42952174320816994, + "step": 694 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16028828828828828, + "grad_norm": 23.609786987304688, + "learning_rate": 9.650725566979671e-06, + "loss": 6.963265419006348, + "loss_alignment": 0.285888671875, + "loss_alignment_w": 0.1429443359375, + "loss_sft": 0.29212240502238274, + "loss_total_v6": 0.43520407751202583, + "step": 695 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16051891891891892, + "grad_norm": 23.629623413085938, + "learning_rate": 9.649329679881334e-06, + "loss": 6.682189464569092, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.28169628232717514, + "loss_total_v6": 0.41763683781027794, + "step": 696 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16074954954954954, + "grad_norm": 24.13742446899414, + "learning_rate": 9.647931110349233e-06, + "loss": 7.077353000640869, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.3028082177042961, + "loss_total_v6": 0.4423345699906349, + "step": 697 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16098018018018018, + "grad_norm": 24.799121856689453, + "learning_rate": 9.646529859190273e-06, + "loss": 7.39961051940918, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.32099615409970284, + "loss_total_v6": 0.46247563511133194, + "step": 698 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16121081081081082, + "grad_norm": 27.59730339050293, + "learning_rate": 9.645125927212907e-06, + "loss": 7.3528218269348145, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3221764750778675, + "loss_total_v6": 0.4595513604581356, + "step": 699 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16144144144144143, + "grad_norm": 23.57408905029297, + "learning_rate": 9.643719315227134e-06, + "loss": 7.388763427734375, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.32037927955389023, + "loss_total_v6": 0.46179772168397903, + "step": 700 + }, + { + "epoch": 0.16144144144144143, + "eval_loss": 0.6064326167106628, + "eval_loss_alignment": 0.24960536172945205, + "eval_loss_alignment_w": 0.12480268086472603, + "eval_loss_sft": 0.4816299304196007, + "eval_loss_total_v6": 0.6064326120838183, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.9707, + "eval_samples_per_second": 19.182, + "eval_steps_per_second": 2.407, + "step": 700 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16167207207207207, + "grad_norm": 24.638072967529297, + "learning_rate": 9.642310024044499e-06, + "loss": 7.236987113952637, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3154861219227314, + "loss_total_v6": 0.4523116759955883, + "step": 701 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1619027027027027, + "grad_norm": 25.662982940673828, + "learning_rate": 9.640898054478095e-06, + "loss": 6.788204193115234, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.2862012777477503, + "loss_total_v6": 0.42426278814673424, + "step": 702 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16213333333333332, + "grad_norm": 22.313671112060547, + "learning_rate": 9.639483407342557e-06, + "loss": 7.060609817504883, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3044015094637871, + "loss_total_v6": 0.4412881098687649, + "step": 703 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16236396396396396, + "grad_norm": 25.676807403564453, + "learning_rate": 9.63806608345407e-06, + "loss": 6.908287048339844, + "loss_alignment": 0.2679443359375, + "loss_alignment_w": 0.13397216796875, + "loss_sft": 0.29747533798217773, + "loss_total_v6": 0.43176794797182083, + "step": 704 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1625945945945946, + "grad_norm": 23.560396194458008, + "learning_rate": 9.636646083630356e-06, + "loss": 6.999075889587402, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.30276815965771675, + "loss_total_v6": 0.43744224309921265, + "step": 705 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16282522522522522, + "grad_norm": 22.534929275512695, + "learning_rate": 9.635223408690688e-06, + "loss": 7.66396427154541, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.34127194434404373, + "loss_total_v6": 0.47899777814745903, + "step": 706 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16305585585585586, + "grad_norm": 23.09499740600586, + "learning_rate": 9.633798059455877e-06, + "loss": 7.176980972290039, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.31054553762078285, + "loss_total_v6": 0.44856129214167595, + "step": 707 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1632864864864865, + "grad_norm": 23.052820205688477, + "learning_rate": 9.63237003674828e-06, + "loss": 7.6993818283081055, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.3397166356444359, + "loss_total_v6": 0.4812113828957081, + "step": 708 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1635171171171171, + "grad_norm": 24.701278686523438, + "learning_rate": 9.630939341391796e-06, + "loss": 7.003816604614258, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.2998753674328327, + "loss_total_v6": 0.4377385415136814, + "step": 709 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16374774774774775, + "grad_norm": 23.84186553955078, + "learning_rate": 9.629505974211863e-06, + "loss": 7.285475730895996, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.32004252821207047, + "loss_total_v6": 0.45534222200512886, + "step": 710 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1639783783783784, + "grad_norm": 27.055479049682617, + "learning_rate": 9.628069936035467e-06, + "loss": 7.072357177734375, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.30533410236239433, + "loss_total_v6": 0.44202231988310814, + "step": 711 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.164209009009009, + "grad_norm": 25.48093605041504, + "learning_rate": 9.626631227691128e-06, + "loss": 7.4758620262146, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.3277913071215153, + "loss_total_v6": 0.4672413840889931, + "step": 712 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16443963963963965, + "grad_norm": 25.00847625732422, + "learning_rate": 9.625189850008912e-06, + "loss": 7.081247329711914, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.30558453500270844, + "loss_total_v6": 0.44257795438170433, + "step": 713 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16467027027027026, + "grad_norm": 24.063554763793945, + "learning_rate": 9.623745803820422e-06, + "loss": 7.0111985206604, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3021830338984728, + "loss_total_v6": 0.43819989636540413, + "step": 714 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1649009009009009, + "grad_norm": 23.53645896911621, + "learning_rate": 9.6222990899588e-06, + "loss": 6.939296722412109, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.29788755998015404, + "loss_total_v6": 0.43370603770017624, + "step": 715 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16513153153153154, + "grad_norm": 22.053682327270508, + "learning_rate": 9.62084970925873e-06, + "loss": 7.002923011779785, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.29971268586814404, + "loss_total_v6": 0.4376826584339142, + "step": 716 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16536216216216215, + "grad_norm": 22.4423885345459, + "learning_rate": 9.619397662556434e-06, + "loss": 7.521102428436279, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.33582208678126335, + "loss_total_v6": 0.47006892040371895, + "step": 717 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1655927927927928, + "grad_norm": 23.537778854370117, + "learning_rate": 9.617942950689671e-06, + "loss": 7.2574462890625, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.31507111340761185, + "loss_total_v6": 0.45359040424227715, + "step": 718 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16582342342342343, + "grad_norm": 24.960094451904297, + "learning_rate": 9.616485574497734e-06, + "loss": 7.088696479797363, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.307637020945549, + "loss_total_v6": 0.4430435188114643, + "step": 719 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16605405405405405, + "grad_norm": 26.398345947265625, + "learning_rate": 9.615025534821463e-06, + "loss": 7.416609764099121, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.32620903104543686, + "loss_total_v6": 0.46353812888264656, + "step": 720 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16628468468468469, + "grad_norm": 25.940919876098633, + "learning_rate": 9.613562832503225e-06, + "loss": 7.34088134765625, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3209571950137615, + "loss_total_v6": 0.4588050916790962, + "step": 721 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16651531531531533, + "grad_norm": 25.962627410888672, + "learning_rate": 9.612097468386927e-06, + "loss": 7.171144962310791, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.3104859795421362, + "loss_total_v6": 0.44819656014442444, + "step": 722 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16674594594594594, + "grad_norm": 28.351797103881836, + "learning_rate": 9.610629443318015e-06, + "loss": 7.477998733520508, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.32755863294005394, + "loss_total_v6": 0.46737490594387054, + "step": 723 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16697657657657658, + "grad_norm": 21.834238052368164, + "learning_rate": 9.609158758143464e-06, + "loss": 7.311892986297607, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.31531544402241707, + "loss_total_v6": 0.45699330046772957, + "step": 724 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16720720720720722, + "grad_norm": 26.00486946105957, + "learning_rate": 9.607685413711786e-06, + "loss": 7.33190393447876, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.3198010176420212, + "loss_total_v6": 0.4582439959049225, + "step": 725 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16743783783783783, + "grad_norm": 24.249195098876953, + "learning_rate": 9.606209410873032e-06, + "loss": 7.295772075653076, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.31900760531425476, + "loss_total_v6": 0.45598575472831726, + "step": 726 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16766846846846847, + "grad_norm": 26.266225814819336, + "learning_rate": 9.60473075047878e-06, + "loss": 7.293049335479736, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.3168537840247154, + "loss_total_v6": 0.4558155760169029, + "step": 727 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1678990990990991, + "grad_norm": 24.417339324951172, + "learning_rate": 9.603249433382145e-06, + "loss": 7.094038009643555, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.30430879816412926, + "loss_total_v6": 0.44337740167975426, + "step": 728 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16812972972972973, + "grad_norm": 26.628122329711914, + "learning_rate": 9.601765460437774e-06, + "loss": 7.199338912963867, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.312888965010643, + "loss_total_v6": 0.4499586895108223, + "step": 729 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16836036036036037, + "grad_norm": 30.282804489135742, + "learning_rate": 9.600278832501843e-06, + "loss": 7.400684833526611, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.3233521059155464, + "loss_total_v6": 0.462542787194252, + "step": 730 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16859099099099098, + "grad_norm": 30.368488311767578, + "learning_rate": 9.598789550432065e-06, + "loss": 6.785353183746338, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.28593150712549686, + "loss_total_v6": 0.4240845777094364, + "step": 731 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16882162162162162, + "grad_norm": 26.744842529296875, + "learning_rate": 9.597297615087686e-06, + "loss": 7.398472785949707, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3240073174238205, + "loss_total_v6": 0.462404552847147, + "step": 732 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16905225225225226, + "grad_norm": 23.842966079711914, + "learning_rate": 9.595803027329473e-06, + "loss": 7.28032112121582, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.31680595874786377, + "loss_total_v6": 0.4550200514495373, + "step": 733 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16928288288288287, + "grad_norm": 23.074840545654297, + "learning_rate": 9.594305788019736e-06, + "loss": 7.733161449432373, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.3451389744877815, + "loss_total_v6": 0.4833225831389427, + "step": 734 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16951351351351351, + "grad_norm": 23.46688461303711, + "learning_rate": 9.5928058980223e-06, + "loss": 7.179591178894043, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3138062320649624, + "loss_total_v6": 0.4487244561314583, + "step": 735 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16974414414414415, + "grad_norm": 24.24445152282715, + "learning_rate": 9.591303358202535e-06, + "loss": 6.981223106384277, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.29608288034796715, + "loss_total_v6": 0.43632641434669495, + "step": 736 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.16997477477477477, + "grad_norm": 22.19243812561035, + "learning_rate": 9.589798169427327e-06, + "loss": 6.922547817230225, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.2923089098185301, + "loss_total_v6": 0.43265925347805023, + "step": 737 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1702054054054054, + "grad_norm": 22.622791290283203, + "learning_rate": 9.588290332565098e-06, + "loss": 7.394185543060303, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3239072374999523, + "loss_total_v6": 0.4621366001665592, + "step": 738 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17043603603603605, + "grad_norm": 23.24901008605957, + "learning_rate": 9.586779848485797e-06, + "loss": 7.158230781555176, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.308717530220747, + "loss_total_v6": 0.4473894201219082, + "step": 739 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17066666666666666, + "grad_norm": 23.877830505371094, + "learning_rate": 9.585266718060897e-06, + "loss": 7.103506088256836, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.30631957948207855, + "loss_total_v6": 0.44396911934018135, + "step": 740 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1708972972972973, + "grad_norm": 23.360008239746094, + "learning_rate": 9.583750942163398e-06, + "loss": 7.298788070678711, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.31954706087708473, + "loss_total_v6": 0.45617425441741943, + "step": 741 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17112792792792791, + "grad_norm": 26.408796310424805, + "learning_rate": 9.58223252166783e-06, + "loss": 6.983349800109863, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.29911500960588455, + "loss_total_v6": 0.43645935878157616, + "step": 742 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17135855855855855, + "grad_norm": 24.78398895263672, + "learning_rate": 9.580711457450246e-06, + "loss": 6.60591459274292, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.277692049741745, + "loss_total_v6": 0.4128696583211422, + "step": 743 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1715891891891892, + "grad_norm": 23.12201499938965, + "learning_rate": 9.579187750388227e-06, + "loss": 7.500580310821533, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.3321590758860111, + "loss_total_v6": 0.4687862806022167, + "step": 744 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1718198198198198, + "grad_norm": 25.296777725219727, + "learning_rate": 9.577661401360873e-06, + "loss": 7.2683329582214355, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3189100809395313, + "loss_total_v6": 0.4542708061635494, + "step": 745 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17205045045045045, + "grad_norm": 24.86825180053711, + "learning_rate": 9.576132411248815e-06, + "loss": 7.18985652923584, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.3091682493686676, + "loss_total_v6": 0.4493660219013691, + "step": 746 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1722810810810811, + "grad_norm": 21.801349639892578, + "learning_rate": 9.574600780934204e-06, + "loss": 7.345915794372559, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.3215464986860752, + "loss_total_v6": 0.4591197408735752, + "step": 747 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1725117117117117, + "grad_norm": 23.94152069091797, + "learning_rate": 9.573066511300715e-06, + "loss": 7.31500768661499, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.32115588150918484, + "loss_total_v6": 0.457187969237566, + "step": 748 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17274234234234234, + "grad_norm": 24.867849349975586, + "learning_rate": 9.571529603233545e-06, + "loss": 7.536200523376465, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.33467527851462364, + "loss_total_v6": 0.47101254761219025, + "step": 749 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17297297297297298, + "grad_norm": 24.670053482055664, + "learning_rate": 9.569990057619414e-06, + "loss": 7.0291748046875, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3041000310331583, + "loss_total_v6": 0.43932342156767845, + "step": 750 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1732036036036036, + "grad_norm": 24.81081771850586, + "learning_rate": 9.568447875346567e-06, + "loss": 7.2289581298828125, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3135652467608452, + "loss_total_v6": 0.4518098719418049, + "step": 751 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17343423423423424, + "grad_norm": 24.936107635498047, + "learning_rate": 9.566903057304764e-06, + "loss": 7.200794219970703, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.31264423951506615, + "loss_total_v6": 0.45004963129758835, + "step": 752 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17366486486486488, + "grad_norm": 25.289960861206055, + "learning_rate": 9.56535560438529e-06, + "loss": 7.315777778625488, + "loss_alignment": 0.281494140625, + "loss_alignment_w": 0.1407470703125, + "loss_sft": 0.3163974732160568, + "loss_total_v6": 0.4572360962629318, + "step": 753 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1738954954954955, + "grad_norm": 24.94016456604004, + "learning_rate": 9.563805517480946e-06, + "loss": 7.5244245529174805, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.33622805401682854, + "loss_total_v6": 0.47027650848031044, + "step": 754 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17412612612612613, + "grad_norm": 24.279857635498047, + "learning_rate": 9.56225279748606e-06, + "loss": 6.655339241027832, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.27981978841125965, + "loss_total_v6": 0.4159587025642395, + "step": 755 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17435675675675677, + "grad_norm": 25.397178649902344, + "learning_rate": 9.560697445296474e-06, + "loss": 7.02215576171875, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3038902226835489, + "loss_total_v6": 0.4388847388327122, + "step": 756 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17458738738738738, + "grad_norm": 23.41543960571289, + "learning_rate": 9.559139461809547e-06, + "loss": 6.58159065246582, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.27623284235596657, + "loss_total_v6": 0.4113494083285332, + "step": 757 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17481801801801802, + "grad_norm": 23.12144660949707, + "learning_rate": 9.557578847924161e-06, + "loss": 7.252588272094727, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.314431793987751, + "loss_total_v6": 0.453286774456501, + "step": 758 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17504864864864864, + "grad_norm": 23.313623428344727, + "learning_rate": 9.556015604540712e-06, + "loss": 6.965331077575684, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.29968256689608097, + "loss_total_v6": 0.4353332035243511, + "step": 759 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17527927927927928, + "grad_norm": 23.199909210205078, + "learning_rate": 9.554449732561114e-06, + "loss": 7.347964763641357, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.32051488384604454, + "loss_total_v6": 0.45924778282642365, + "step": 760 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17550990990990992, + "grad_norm": 23.352428436279297, + "learning_rate": 9.5528812328888e-06, + "loss": 7.12160587310791, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3066268302500248, + "loss_total_v6": 0.445100337266922, + "step": 761 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17574054054054053, + "grad_norm": 26.811861038208008, + "learning_rate": 9.551310106428713e-06, + "loss": 7.223580837249756, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.31367165222764015, + "loss_total_v6": 0.45147378370165825, + "step": 762 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17597117117117117, + "grad_norm": 23.209890365600586, + "learning_rate": 9.549736354087317e-06, + "loss": 7.134970664978027, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.3084692284464836, + "loss_total_v6": 0.4459356591105461, + "step": 763 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1762018018018018, + "grad_norm": 28.39348793029785, + "learning_rate": 9.548159976772593e-06, + "loss": 6.849673271179199, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.29190462455153465, + "loss_total_v6": 0.42810457199811935, + "step": 764 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17643243243243242, + "grad_norm": 25.861614227294922, + "learning_rate": 9.546580975394029e-06, + "loss": 7.547506809234619, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.3334287591278553, + "loss_total_v6": 0.4717191718518734, + "step": 765 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17666306306306306, + "grad_norm": 22.61585807800293, + "learning_rate": 9.54499935086263e-06, + "loss": 7.234643936157227, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.31436310708522797, + "loss_total_v6": 0.45216522365808487, + "step": 766 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1768936936936937, + "grad_norm": 25.17121124267578, + "learning_rate": 9.54341510409092e-06, + "loss": 7.0059332847595215, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.3004349283874035, + "loss_total_v6": 0.4378708340227604, + "step": 767 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17712432432432432, + "grad_norm": 23.97315788269043, + "learning_rate": 9.541828235992927e-06, + "loss": 7.24859619140625, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.31938554532825947, + "loss_total_v6": 0.4530372768640518, + "step": 768 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17735495495495496, + "grad_norm": 24.03539276123047, + "learning_rate": 9.540238747484197e-06, + "loss": 7.3114166259765625, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.31876470148563385, + "loss_total_v6": 0.45696355774998665, + "step": 769 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1775855855855856, + "grad_norm": 23.138002395629883, + "learning_rate": 9.538646639481785e-06, + "loss": 7.337723731994629, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.31917291693389416, + "loss_total_v6": 0.4586077220737934, + "step": 770 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1778162162162162, + "grad_norm": 23.512596130371094, + "learning_rate": 9.537051912904263e-06, + "loss": 7.248204708099365, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3169196583330631, + "loss_total_v6": 0.4530127942562103, + "step": 771 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17804684684684685, + "grad_norm": 23.394147872924805, + "learning_rate": 9.535454568671705e-06, + "loss": 6.914125442504883, + "loss_alignment": 0.278076171875, + "loss_alignment_w": 0.1390380859375, + "loss_sft": 0.2930794972926378, + "loss_total_v6": 0.4321328289806843, + "step": 772 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17827747747747746, + "grad_norm": 21.966001510620117, + "learning_rate": 9.533854607705702e-06, + "loss": 7.702850341796875, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3462505117058754, + "loss_total_v6": 0.4814281389117241, + "step": 773 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1785081081081081, + "grad_norm": 22.971940994262695, + "learning_rate": 9.532252030929351e-06, + "loss": 6.839685440063477, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.2899223603308201, + "loss_total_v6": 0.427480336278677, + "step": 774 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17873873873873874, + "grad_norm": 23.906557083129883, + "learning_rate": 9.530646839267262e-06, + "loss": 7.1212263107299805, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.304680535569787, + "loss_total_v6": 0.4450766555964947, + "step": 775 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17896936936936936, + "grad_norm": 24.332780838012695, + "learning_rate": 9.529039033645548e-06, + "loss": 7.387455940246582, + "loss_alignment": 0.280517578125, + "loss_alignment_w": 0.1402587890625, + "loss_sft": 0.32151826098561287, + "loss_total_v6": 0.4617160074412823, + "step": 776 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1792, + "grad_norm": 26.56561279296875, + "learning_rate": 9.527428614991835e-06, + "loss": 7.304253578186035, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3218875601887703, + "loss_total_v6": 0.4565158635377884, + "step": 777 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17943063063063064, + "grad_norm": 22.329500198364258, + "learning_rate": 9.525815584235258e-06, + "loss": 7.356873989105225, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.32438284158706665, + "loss_total_v6": 0.45980460569262505, + "step": 778 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.17966126126126125, + "grad_norm": 24.133255004882812, + "learning_rate": 9.524199942306454e-06, + "loss": 7.245771408081055, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3156841993331909, + "loss_total_v6": 0.4528606981039047, + "step": 779 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1798918918918919, + "grad_norm": 24.015195846557617, + "learning_rate": 9.522581690137567e-06, + "loss": 7.663888454437256, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.3414045386016369, + "loss_total_v6": 0.4789930284023285, + "step": 780 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18012252252252253, + "grad_norm": 26.76654815673828, + "learning_rate": 9.52096082866225e-06, + "loss": 7.2986226081848145, + "loss_alignment": 0.2822265625, + "loss_alignment_w": 0.14111328125, + "loss_sft": 0.3150811605155468, + "loss_total_v6": 0.4561639241874218, + "step": 781 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18035315315315315, + "grad_norm": 23.140933990478516, + "learning_rate": 9.51933735881566e-06, + "loss": 7.373495101928711, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.32232414558529854, + "loss_total_v6": 0.46084342896938324, + "step": 782 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18058378378378379, + "grad_norm": 24.679475784301758, + "learning_rate": 9.51771128153446e-06, + "loss": 7.184541702270508, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.31100286543369293, + "loss_total_v6": 0.44903387129306793, + "step": 783 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18081441441441443, + "grad_norm": 22.962017059326172, + "learning_rate": 9.516082597756817e-06, + "loss": 6.981627941131592, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.2982902191579342, + "loss_total_v6": 0.4363517463207245, + "step": 784 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18104504504504504, + "grad_norm": 25.064498901367188, + "learning_rate": 9.514451308422397e-06, + "loss": 7.654791831970215, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.34050029143691063, + "loss_total_v6": 0.478424496948719, + "step": 785 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18127567567567568, + "grad_norm": 23.78114128112793, + "learning_rate": 9.512817414472376e-06, + "loss": 6.904546737670898, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.29666173458099365, + "loss_total_v6": 0.43153416737914085, + "step": 786 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18150630630630632, + "grad_norm": 22.933822631835938, + "learning_rate": 9.511180916849431e-06, + "loss": 7.09158992767334, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.30818410590291023, + "loss_total_v6": 0.44322438538074493, + "step": 787 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18173693693693693, + "grad_norm": 26.551753997802734, + "learning_rate": 9.509541816497738e-06, + "loss": 6.903048038482666, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.2964154854416847, + "loss_total_v6": 0.43144049122929573, + "step": 788 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18196756756756757, + "grad_norm": 24.7861328125, + "learning_rate": 9.50790011436298e-06, + "loss": 7.07211971282959, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.30281680822372437, + "loss_total_v6": 0.44200748577713966, + "step": 789 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18219819819819819, + "grad_norm": 23.97046661376953, + "learning_rate": 9.506255811392338e-06, + "loss": 7.662881374359131, + "loss_alignment": 0.28173828125, + "loss_alignment_w": 0.140869140625, + "loss_sft": 0.33830510824918747, + "loss_total_v6": 0.4789300933480263, + "step": 790 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18242882882882883, + "grad_norm": 26.330759048461914, + "learning_rate": 9.504608908534489e-06, + "loss": 6.834634780883789, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.29091894812881947, + "loss_total_v6": 0.4271646700799465, + "step": 791 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18265945945945947, + "grad_norm": 23.929569244384766, + "learning_rate": 9.50295940673962e-06, + "loss": 7.031583309173584, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.30411321111023426, + "loss_total_v6": 0.4394739381968975, + "step": 792 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18289009009009008, + "grad_norm": 25.52804946899414, + "learning_rate": 9.501307306959411e-06, + "loss": 7.025256156921387, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.30214615166187286, + "loss_total_v6": 0.43907850980758667, + "step": 793 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18312072072072072, + "grad_norm": 24.691930770874023, + "learning_rate": 9.49965261014704e-06, + "loss": 7.128931999206543, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.30723730847239494, + "loss_total_v6": 0.44555824249982834, + "step": 794 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18335135135135136, + "grad_norm": 22.892822265625, + "learning_rate": 9.497995317257187e-06, + "loss": 6.820432662963867, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.2885359786450863, + "loss_total_v6": 0.4262770712375641, + "step": 795 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18358198198198197, + "grad_norm": 21.61021614074707, + "learning_rate": 9.496335429246026e-06, + "loss": 6.787701606750488, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.2911137081682682, + "loss_total_v6": 0.4242313764989376, + "step": 796 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1838126126126126, + "grad_norm": 24.607372283935547, + "learning_rate": 9.494672947071236e-06, + "loss": 7.132047176361084, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.30889688059687614, + "loss_total_v6": 0.44575295597314835, + "step": 797 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18404324324324325, + "grad_norm": 24.182552337646484, + "learning_rate": 9.49300787169198e-06, + "loss": 7.709855079650879, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.34323983266949654, + "loss_total_v6": 0.48186592757701874, + "step": 798 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18427387387387387, + "grad_norm": 25.102516174316406, + "learning_rate": 9.491340204068929e-06, + "loss": 7.269367218017578, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.31886792555451393, + "loss_total_v6": 0.45433544740080833, + "step": 799 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1845045045045045, + "grad_norm": 25.557086944580078, + "learning_rate": 9.489669945164242e-06, + "loss": 7.329100608825684, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.3183288015425205, + "loss_total_v6": 0.4580687806010246, + "step": 800 + }, + { + "epoch": 0.1845045045045045, + "eval_loss": 0.6026487946510315, + "eval_loss_alignment": 0.24815612514269406, + "eval_loss_alignment_w": 0.12407806257134703, + "eval_loss_sft": 0.47857077878189686, + "eval_loss_total_v6": 0.6026488412937073, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.5772, + "eval_samples_per_second": 19.265, + "eval_steps_per_second": 2.418, + "step": 800 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18473513513513515, + "grad_norm": 25.11588478088379, + "learning_rate": 9.487997095941579e-06, + "loss": 7.168918609619141, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.3091566823422909, + "loss_total_v6": 0.4480574354529381, + "step": 801 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18496576576576576, + "grad_norm": 23.52119255065918, + "learning_rate": 9.48632165736609e-06, + "loss": 7.327577590942383, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.3191491588950157, + "loss_total_v6": 0.4579736106097698, + "step": 802 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1851963963963964, + "grad_norm": 24.28586769104004, + "learning_rate": 9.48464363040442e-06, + "loss": 7.232993125915527, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.31621308997273445, + "loss_total_v6": 0.45206209644675255, + "step": 803 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.185427027027027, + "grad_norm": 25.284147262573242, + "learning_rate": 9.482963016024709e-06, + "loss": 7.277033805847168, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.3151509203016758, + "loss_total_v6": 0.454814612865448, + "step": 804 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18565765765765765, + "grad_norm": 21.81777000427246, + "learning_rate": 9.481279815196587e-06, + "loss": 7.204213619232178, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.3112405315041542, + "loss_total_v6": 0.4502633549273014, + "step": 805 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1858882882882883, + "grad_norm": 30.478952407836914, + "learning_rate": 9.479594028891178e-06, + "loss": 7.331459999084473, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.32030732557177544, + "loss_total_v6": 0.45821626111865044, + "step": 806 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1861189189189189, + "grad_norm": 24.371051788330078, + "learning_rate": 9.477905658081098e-06, + "loss": 7.257650375366211, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3174184449017048, + "loss_total_v6": 0.4536031410098076, + "step": 807 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18634954954954955, + "grad_norm": 22.787752151489258, + "learning_rate": 9.476214703740454e-06, + "loss": 7.090266227722168, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.3044239766895771, + "loss_total_v6": 0.4431416504085064, + "step": 808 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1865801801801802, + "grad_norm": 25.637781143188477, + "learning_rate": 9.474521166844843e-06, + "loss": 7.143834114074707, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3092215731739998, + "loss_total_v6": 0.4464896395802498, + "step": 809 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1868108108108108, + "grad_norm": 23.572261810302734, + "learning_rate": 9.472825048371353e-06, + "loss": 6.961542129516602, + "loss_alignment": 0.28173828125, + "loss_alignment_w": 0.140869140625, + "loss_sft": 0.29436458460986614, + "loss_total_v6": 0.4350964054465294, + "step": 810 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18704144144144144, + "grad_norm": 24.550907135009766, + "learning_rate": 9.471126349298557e-06, + "loss": 6.846390724182129, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.28934958577156067, + "loss_total_v6": 0.42789939790964127, + "step": 811 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18727207207207208, + "grad_norm": 24.754457473754883, + "learning_rate": 9.469425070606525e-06, + "loss": 7.590785503387451, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3374917209148407, + "loss_total_v6": 0.4744240827858448, + "step": 812 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1875027027027027, + "grad_norm": 22.938772201538086, + "learning_rate": 9.467721213276808e-06, + "loss": 7.408852577209473, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.32807404547929764, + "loss_total_v6": 0.46305329352617264, + "step": 813 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18773333333333334, + "grad_norm": 22.580297470092773, + "learning_rate": 9.46601477829245e-06, + "loss": 7.377330780029297, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.3229758720844984, + "loss_total_v6": 0.46108317002654076, + "step": 814 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18796396396396398, + "grad_norm": 22.30919647216797, + "learning_rate": 9.464305766637978e-06, + "loss": 7.1690521240234375, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.31105708703398705, + "loss_total_v6": 0.44806577265262604, + "step": 815 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1881945945945946, + "grad_norm": 23.07061767578125, + "learning_rate": 9.462594179299408e-06, + "loss": 6.996876239776611, + "loss_alignment": 0.2686767578125, + "loss_alignment_w": 0.13433837890625, + "loss_sft": 0.30301216058433056, + "loss_total_v6": 0.4373047649860382, + "step": 816 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18842522522522523, + "grad_norm": 25.09157371520996, + "learning_rate": 9.46088001726424e-06, + "loss": 7.116680145263672, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.30755496956408024, + "loss_total_v6": 0.4447925128042698, + "step": 817 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18865585585585587, + "grad_norm": 23.38579559326172, + "learning_rate": 9.459163281521462e-06, + "loss": 7.396327972412109, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.32330871000885963, + "loss_total_v6": 0.46227050945162773, + "step": 818 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18888648648648648, + "grad_norm": 23.375873565673828, + "learning_rate": 9.457443973061546e-06, + "loss": 7.279664039611816, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3186570033431053, + "loss_total_v6": 0.4549790285527706, + "step": 819 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18911711711711712, + "grad_norm": 23.901294708251953, + "learning_rate": 9.45572209287645e-06, + "loss": 6.879481792449951, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.2937371227890253, + "loss_total_v6": 0.42996759712696075, + "step": 820 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18934774774774774, + "grad_norm": 21.5498046875, + "learning_rate": 9.45399764195961e-06, + "loss": 7.685112953186035, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.3403201773762703, + "loss_total_v6": 0.4803195632994175, + "step": 821 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18957837837837838, + "grad_norm": 22.659526824951172, + "learning_rate": 9.452270621305952e-06, + "loss": 7.845632553100586, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3537553437054157, + "loss_total_v6": 0.4903520308434963, + "step": 822 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.18980900900900902, + "grad_norm": 23.15182113647461, + "learning_rate": 9.450541031911879e-06, + "loss": 7.778017044067383, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3503228574991226, + "loss_total_v6": 0.4861260838806629, + "step": 823 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19003963963963963, + "grad_norm": 22.89985466003418, + "learning_rate": 9.448808874775282e-06, + "loss": 7.2290191650390625, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3146829716861248, + "loss_total_v6": 0.4518137089908123, + "step": 824 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19027027027027027, + "grad_norm": 23.9815673828125, + "learning_rate": 9.447074150895528e-06, + "loss": 6.977939605712891, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.299844978377223, + "loss_total_v6": 0.43612121418118477, + "step": 825 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1905009009009009, + "grad_norm": 22.471908569335938, + "learning_rate": 9.445336861273468e-06, + "loss": 7.425594329833984, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.330600518733263, + "loss_total_v6": 0.4640996567904949, + "step": 826 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19073153153153152, + "grad_norm": 21.371427536010742, + "learning_rate": 9.443597006911432e-06, + "loss": 6.993454456329346, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.2977476492524147, + "loss_total_v6": 0.4370909109711647, + "step": 827 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19096216216216216, + "grad_norm": 24.821887969970703, + "learning_rate": 9.441854588813228e-06, + "loss": 7.412995338439941, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.32430465146899223, + "loss_total_v6": 0.46331223100423813, + "step": 828 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1911927927927928, + "grad_norm": 24.480812072753906, + "learning_rate": 9.440109607984148e-06, + "loss": 7.154157638549805, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.31301007978618145, + "loss_total_v6": 0.4471348412334919, + "step": 829 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19142342342342342, + "grad_norm": 24.503244400024414, + "learning_rate": 9.438362065430961e-06, + "loss": 6.67576265335083, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.27949406020343304, + "loss_total_v6": 0.4172351621091366, + "step": 830 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19165405405405406, + "grad_norm": 23.21803092956543, + "learning_rate": 9.43661196216191e-06, + "loss": 7.115450859069824, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.30711192823946476, + "loss_total_v6": 0.444715678691864, + "step": 831 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1918846846846847, + "grad_norm": 24.103008270263672, + "learning_rate": 9.434859299186719e-06, + "loss": 7.347009658813477, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.3199669159948826, + "loss_total_v6": 0.4591881036758423, + "step": 832 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1921153153153153, + "grad_norm": 23.35542869567871, + "learning_rate": 9.433104077516588e-06, + "loss": 6.794672012329102, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.2870479766279459, + "loss_total_v6": 0.42466700077056885, + "step": 833 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19234594594594595, + "grad_norm": 31.051639556884766, + "learning_rate": 9.431346298164195e-06, + "loss": 7.476103782653809, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3289203457534313, + "loss_total_v6": 0.46725650504231453, + "step": 834 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19257657657657656, + "grad_norm": 27.984769821166992, + "learning_rate": 9.429585962143687e-06, + "loss": 7.474880695343018, + "loss_alignment": 0.282470703125, + "loss_alignment_w": 0.1412353515625, + "loss_sft": 0.3258836455643177, + "loss_total_v6": 0.4671800397336483, + "step": 835 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1928072072072072, + "grad_norm": 22.934358596801758, + "learning_rate": 9.4278230704707e-06, + "loss": 6.919014930725098, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.2941785641014576, + "loss_total_v6": 0.4324384517967701, + "step": 836 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19303783783783784, + "grad_norm": 22.652206420898438, + "learning_rate": 9.426057624162324e-06, + "loss": 7.484976768493652, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.33353370428085327, + "loss_total_v6": 0.4678110405802727, + "step": 837 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19326846846846846, + "grad_norm": 22.929840087890625, + "learning_rate": 9.424289624237143e-06, + "loss": 7.262887954711914, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3190275691449642, + "loss_total_v6": 0.4539305195212364, + "step": 838 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1934990990990991, + "grad_norm": 23.491687774658203, + "learning_rate": 9.422519071715204e-06, + "loss": 7.690555572509766, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.34084339812397957, + "loss_total_v6": 0.48065970465540886, + "step": 839 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19372972972972974, + "grad_norm": 24.796791076660156, + "learning_rate": 9.420745967618026e-06, + "loss": 7.5211334228515625, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.33095648512244225, + "loss_total_v6": 0.47007086500525475, + "step": 840 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19396036036036035, + "grad_norm": 24.141069412231445, + "learning_rate": 9.418970312968603e-06, + "loss": 7.070819854736328, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3050396256148815, + "loss_total_v6": 0.4419262260198593, + "step": 841 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.194190990990991, + "grad_norm": 25.014413833618164, + "learning_rate": 9.417192108791402e-06, + "loss": 7.235541343688965, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.31414457596838474, + "loss_total_v6": 0.4522213526070118, + "step": 842 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19442162162162163, + "grad_norm": 24.019577026367188, + "learning_rate": 9.41541135611236e-06, + "loss": 7.563320159912109, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.33418821170926094, + "loss_total_v6": 0.47270751744508743, + "step": 843 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19465225225225224, + "grad_norm": 23.772653579711914, + "learning_rate": 9.413628055958878e-06, + "loss": 7.450672626495361, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.3280022479593754, + "loss_total_v6": 0.4656670428812504, + "step": 844 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19488288288288289, + "grad_norm": 24.460315704345703, + "learning_rate": 9.411842209359838e-06, + "loss": 7.387714862823486, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.3215039409697056, + "loss_total_v6": 0.4617321938276291, + "step": 845 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19511351351351353, + "grad_norm": 22.861913681030273, + "learning_rate": 9.410053817345584e-06, + "loss": 7.462444305419922, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3279445320367813, + "loss_total_v6": 0.4664027877151966, + "step": 846 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19534414414414414, + "grad_norm": 25.198047637939453, + "learning_rate": 9.408262880947927e-06, + "loss": 7.135989189147949, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.3081971798092127, + "loss_total_v6": 0.44599931687116623, + "step": 847 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19557477477477478, + "grad_norm": 22.958452224731445, + "learning_rate": 9.406469401200151e-06, + "loss": 7.1739397048950195, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3154213950037956, + "loss_total_v6": 0.4483712166547775, + "step": 848 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1958054054054054, + "grad_norm": 22.522083282470703, + "learning_rate": 9.404673379137007e-06, + "loss": 6.991242408752441, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.2986470013856888, + "loss_total_v6": 0.4369526617228985, + "step": 849 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19603603603603603, + "grad_norm": 23.763469696044922, + "learning_rate": 9.402874815794708e-06, + "loss": 7.079322814941406, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3060135841369629, + "loss_total_v6": 0.4424576722085476, + "step": 850 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19626666666666667, + "grad_norm": 24.633420944213867, + "learning_rate": 9.401073712210937e-06, + "loss": 6.919668197631836, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.29716434702277184, + "loss_total_v6": 0.43247928842902184, + "step": 851 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19649729729729729, + "grad_norm": 23.192651748657227, + "learning_rate": 9.399270069424843e-06, + "loss": 7.893838405609131, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3567682094871998, + "loss_total_v6": 0.4933648817241192, + "step": 852 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19672792792792793, + "grad_norm": 23.959613800048828, + "learning_rate": 9.39746388847704e-06, + "loss": 7.035414695739746, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3040475398302078, + "loss_total_v6": 0.43971342593431473, + "step": 853 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19695855855855857, + "grad_norm": 22.79491424560547, + "learning_rate": 9.3956551704096e-06, + "loss": 7.298015594482422, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.318796843290329, + "loss_total_v6": 0.45612596347928047, + "step": 854 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19718918918918918, + "grad_norm": 22.11050796508789, + "learning_rate": 9.39384391626607e-06, + "loss": 7.473455429077148, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.33137930929660797, + "loss_total_v6": 0.46709098294377327, + "step": 855 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19741981981981982, + "grad_norm": 24.221012115478516, + "learning_rate": 9.392030127091452e-06, + "loss": 7.235605239868164, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.31195129454135895, + "loss_total_v6": 0.45222534984350204, + "step": 856 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19765045045045046, + "grad_norm": 25.354915618896484, + "learning_rate": 9.39021380393221e-06, + "loss": 7.1940999031066895, + "loss_alignment": 0.2669677734375, + "loss_alignment_w": 0.13348388671875, + "loss_sft": 0.3163304589688778, + "loss_total_v6": 0.4496312513947487, + "step": 857 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19788108108108107, + "grad_norm": 24.855257034301758, + "learning_rate": 9.388394947836278e-06, + "loss": 7.0546159744262695, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.30298932641744614, + "loss_total_v6": 0.44091351330280304, + "step": 858 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1981117117117117, + "grad_norm": 24.409915924072266, + "learning_rate": 9.386573559853046e-06, + "loss": 7.111321449279785, + "loss_alignment": 0.278076171875, + "loss_alignment_w": 0.1390380859375, + "loss_sft": 0.3054347485303879, + "loss_total_v6": 0.4444575682282448, + "step": 859 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19834234234234235, + "grad_norm": 24.51997947692871, + "learning_rate": 9.384749641033358e-06, + "loss": 7.448563098907471, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.3281145393848419, + "loss_total_v6": 0.4655351936817169, + "step": 860 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19857297297297297, + "grad_norm": 24.50995635986328, + "learning_rate": 9.382923192429532e-06, + "loss": 6.746048927307129, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.28634366393089294, + "loss_total_v6": 0.42162807658314705, + "step": 861 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1988036036036036, + "grad_norm": 24.694971084594727, + "learning_rate": 9.381094215095336e-06, + "loss": 7.298945903778076, + "loss_alignment": 0.2667236328125, + "loss_alignment_w": 0.13336181640625, + "loss_sft": 0.32273074984550476, + "loss_total_v6": 0.45618413016200066, + "step": 862 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19903423423423425, + "grad_norm": 25.685771942138672, + "learning_rate": 9.379262710086002e-06, + "loss": 6.965340614318848, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.29940848983824253, + "loss_total_v6": 0.4353337809443474, + "step": 863 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19926486486486486, + "grad_norm": 27.18695068359375, + "learning_rate": 9.377428678458215e-06, + "loss": 7.235050201416016, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3152429759502411, + "loss_total_v6": 0.4521906077861786, + "step": 864 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1994954954954955, + "grad_norm": 24.626724243164062, + "learning_rate": 9.375592121270122e-06, + "loss": 7.295762062072754, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3205175865441561, + "loss_total_v6": 0.4559851288795471, + "step": 865 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.1997261261261261, + "grad_norm": 24.1707706451416, + "learning_rate": 9.373753039581326e-06, + "loss": 7.17184591293335, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.31091128662228584, + "loss_total_v6": 0.44824038073420525, + "step": 866 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.19995675675675675, + "grad_norm": 29.28748893737793, + "learning_rate": 9.371911434452885e-06, + "loss": 6.792362213134766, + "loss_alignment": 0.2642822265625, + "loss_alignment_w": 0.13214111328125, + "loss_sft": 0.2922136653214693, + "loss_total_v6": 0.42452262714505196, + "step": 867 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2001873873873874, + "grad_norm": 22.38194465637207, + "learning_rate": 9.370067306947317e-06, + "loss": 7.0589704513549805, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.305962273851037, + "loss_total_v6": 0.4411856606602669, + "step": 868 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.200418018018018, + "grad_norm": 22.826560974121094, + "learning_rate": 9.36822065812859e-06, + "loss": 6.984838008880615, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.30077966302633286, + "loss_total_v6": 0.43655238300561905, + "step": 869 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20064864864864865, + "grad_norm": 23.7735652923584, + "learning_rate": 9.36637148906213e-06, + "loss": 7.33643913269043, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.32090841978788376, + "loss_total_v6": 0.45852744206786156, + "step": 870 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2008792792792793, + "grad_norm": 25.904102325439453, + "learning_rate": 9.364519800814818e-06, + "loss": 6.963071346282959, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.29944977164268494, + "loss_total_v6": 0.43519196659326553, + "step": 871 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2011099099099099, + "grad_norm": 23.052852630615234, + "learning_rate": 9.362665594454985e-06, + "loss": 7.371308326721191, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.32160765677690506, + "loss_total_v6": 0.46070678532123566, + "step": 872 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20134054054054054, + "grad_norm": 24.384729385375977, + "learning_rate": 9.360808871052415e-06, + "loss": 7.079678535461426, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.30818728916347027, + "loss_total_v6": 0.4424798972904682, + "step": 873 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20157117117117118, + "grad_norm": 23.298776626586914, + "learning_rate": 9.35894963167835e-06, + "loss": 7.437570571899414, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.32933484390378, + "loss_total_v6": 0.4648481458425522, + "step": 874 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2018018018018018, + "grad_norm": 23.212865829467773, + "learning_rate": 9.357087877405475e-06, + "loss": 7.320503234863281, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3235897980630398, + "loss_total_v6": 0.4575314596295357, + "step": 875 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20203243243243243, + "grad_norm": 24.25543785095215, + "learning_rate": 9.355223609307932e-06, + "loss": 7.703491687774658, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.34383395314216614, + "loss_total_v6": 0.48146822303533554, + "step": 876 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20226306306306308, + "grad_norm": 22.82959747314453, + "learning_rate": 9.353356828461313e-06, + "loss": 6.748510837554932, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.2893661633133888, + "loss_total_v6": 0.4217819310724735, + "step": 877 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2024936936936937, + "grad_norm": 23.487377166748047, + "learning_rate": 9.351487535942658e-06, + "loss": 7.114376068115234, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3061292301863432, + "loss_total_v6": 0.44464851543307304, + "step": 878 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20272432432432433, + "grad_norm": 25.299646377563477, + "learning_rate": 9.349615732830455e-06, + "loss": 7.3023271560668945, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.3176472783088684, + "loss_total_v6": 0.4563954472541809, + "step": 879 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20295495495495494, + "grad_norm": 21.659578323364258, + "learning_rate": 9.347741420204643e-06, + "loss": 7.078652381896973, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.3028131164610386, + "loss_total_v6": 0.442415788769722, + "step": 880 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20318558558558558, + "grad_norm": 23.02499008178711, + "learning_rate": 9.345864599146608e-06, + "loss": 7.168112754821777, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.3111662119626999, + "loss_total_v6": 0.4480070322751999, + "step": 881 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20341621621621622, + "grad_norm": 24.211362838745117, + "learning_rate": 9.343985270739184e-06, + "loss": 7.39654541015625, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.32643506303429604, + "loss_total_v6": 0.46228406578302383, + "step": 882 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20364684684684684, + "grad_norm": 24.784103393554688, + "learning_rate": 9.342103436066647e-06, + "loss": 6.954733371734619, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.2983945794403553, + "loss_total_v6": 0.4346708320081234, + "step": 883 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20387747747747748, + "grad_norm": 24.254329681396484, + "learning_rate": 9.340219096214728e-06, + "loss": 7.228929042816162, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3147383853793144, + "loss_total_v6": 0.4518080838024616, + "step": 884 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20410810810810812, + "grad_norm": 23.44577980041504, + "learning_rate": 9.338332252270594e-06, + "loss": 7.44122314453125, + "loss_alignment": 0.282958984375, + "loss_alignment_w": 0.1414794921875, + "loss_sft": 0.3235511798411608, + "loss_total_v6": 0.4650764428079128, + "step": 885 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20433873873873873, + "grad_norm": 26.975374221801758, + "learning_rate": 9.336442905322862e-06, + "loss": 7.197182655334473, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.31136567890644073, + "loss_total_v6": 0.44982394203543663, + "step": 886 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20456936936936937, + "grad_norm": 26.899099349975586, + "learning_rate": 9.334551056461593e-06, + "loss": 7.281010627746582, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3186190687119961, + "loss_total_v6": 0.4550631567835808, + "step": 887 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2048, + "grad_norm": 23.785127639770508, + "learning_rate": 9.33265670677829e-06, + "loss": 7.184414863586426, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3127191625535488, + "loss_total_v6": 0.4490259289741516, + "step": 888 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20503063063063062, + "grad_norm": 26.4122314453125, + "learning_rate": 9.330759857365899e-06, + "loss": 7.260825157165527, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.3162588719278574, + "loss_total_v6": 0.45380157232284546, + "step": 889 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20526126126126126, + "grad_norm": 23.899768829345703, + "learning_rate": 9.328860509318807e-06, + "loss": 6.574178695678711, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.2727026008069515, + "loss_total_v6": 0.4108861833810806, + "step": 890 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2054918918918919, + "grad_norm": 22.721256256103516, + "learning_rate": 9.326958663732848e-06, + "loss": 7.412054538726807, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.3234829194843769, + "loss_total_v6": 0.4632534310221672, + "step": 891 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20572252252252252, + "grad_norm": 21.548267364501953, + "learning_rate": 9.32505432170529e-06, + "loss": 7.073696613311768, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3065927419811487, + "loss_total_v6": 0.4421060346066952, + "step": 892 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20595315315315316, + "grad_norm": 26.41758918762207, + "learning_rate": 9.323147484334843e-06, + "loss": 6.9853644371032715, + "loss_alignment": 0.281005859375, + "loss_alignment_w": 0.1405029296875, + "loss_sft": 0.29597552865743637, + "loss_total_v6": 0.43658527731895447, + "step": 893 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2061837837837838, + "grad_norm": 23.8134708404541, + "learning_rate": 9.32123815272166e-06, + "loss": 6.846484184265137, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.2895538378506899, + "loss_total_v6": 0.42790527269244194, + "step": 894 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2064144144144144, + "grad_norm": 22.192440032958984, + "learning_rate": 9.319326327967332e-06, + "loss": 7.400468826293945, + "loss_alignment": 0.2783203125, + "loss_alignment_w": 0.13916015625, + "loss_sft": 0.3236285448074341, + "loss_total_v6": 0.4625293090939522, + "step": 895 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20664504504504505, + "grad_norm": 23.384418487548828, + "learning_rate": 9.317412011174886e-06, + "loss": 7.13239049911499, + "loss_alignment": 0.2747802734375, + "loss_alignment_w": 0.13739013671875, + "loss_sft": 0.3082774616777897, + "loss_total_v6": 0.4457744061946869, + "step": 896 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20687567567567566, + "grad_norm": 25.839984893798828, + "learning_rate": 9.315495203448792e-06, + "loss": 6.959521770477295, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.29835816472768784, + "loss_total_v6": 0.43497011065483093, + "step": 897 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2071063063063063, + "grad_norm": 22.544151306152344, + "learning_rate": 9.313575905894947e-06, + "loss": 7.028319358825684, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.30113215185701847, + "loss_total_v6": 0.4392699673771858, + "step": 898 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20733693693693694, + "grad_norm": 23.74683952331543, + "learning_rate": 9.311654119620697e-06, + "loss": 7.288507461547852, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.31734813190996647, + "loss_total_v6": 0.4555317163467407, + "step": 899 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20756756756756756, + "grad_norm": 23.644025802612305, + "learning_rate": 9.309729845734813e-06, + "loss": 6.931960105895996, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.29658978804945946, + "loss_total_v6": 0.43324751406908035, + "step": 900 + }, + { + "epoch": 0.20756756756756756, + "eval_loss": 0.6044149994850159, + "eval_loss_alignment": 0.24625205122716895, + "eval_loss_alignment_w": 0.12312602561358447, + "eval_loss_sft": 0.4812889589362493, + "eval_loss_total_v6": 0.6044149844817919, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.2601, + "eval_samples_per_second": 19.121, + "eval_steps_per_second": 2.4, + "step": 900 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2077981981981982, + "grad_norm": 24.93939971923828, + "learning_rate": 9.307803085347508e-06, + "loss": 7.35106086730957, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.32481303066015244, + "loss_total_v6": 0.45944131165742874, + "step": 901 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20802882882882884, + "grad_norm": 25.40445327758789, + "learning_rate": 9.305873839570429e-06, + "loss": 7.177141189575195, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3117915391921997, + "loss_total_v6": 0.4485713243484497, + "step": 902 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20825945945945945, + "grad_norm": 23.219289779663086, + "learning_rate": 9.303942109516655e-06, + "loss": 7.406335830688477, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.325505830347538, + "loss_total_v6": 0.4628959707915783, + "step": 903 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2084900900900901, + "grad_norm": 23.82453155517578, + "learning_rate": 9.302007896300697e-06, + "loss": 7.137380123138428, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.30718550086021423, + "loss_total_v6": 0.44608626514673233, + "step": 904 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20872072072072073, + "grad_norm": 23.922067642211914, + "learning_rate": 9.300071201038503e-06, + "loss": 7.512930870056152, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3334192633628845, + "loss_total_v6": 0.4695581793785095, + "step": 905 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20895135135135134, + "grad_norm": 22.16087532043457, + "learning_rate": 9.29813202484745e-06, + "loss": 7.02012825012207, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.30081856995821, + "loss_total_v6": 0.4387580081820488, + "step": 906 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20918198198198198, + "grad_norm": 21.577566146850586, + "learning_rate": 9.296190368846345e-06, + "loss": 7.014657497406006, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.30091916024684906, + "loss_total_v6": 0.43841610476374626, + "step": 907 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20941261261261263, + "grad_norm": 24.36880874633789, + "learning_rate": 9.29424623415543e-06, + "loss": 6.826986312866211, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.2931417301297188, + "loss_total_v6": 0.4266866482794285, + "step": 908 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20964324324324324, + "grad_norm": 25.21582794189453, + "learning_rate": 9.292299621896375e-06, + "loss": 7.1713361740112305, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3119627609848976, + "loss_total_v6": 0.448208499699831, + "step": 909 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.20987387387387388, + "grad_norm": 23.890897750854492, + "learning_rate": 9.290350533192277e-06, + "loss": 7.013001918792725, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3057290054857731, + "loss_total_v6": 0.4383126236498356, + "step": 910 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2101045045045045, + "grad_norm": 25.069547653198242, + "learning_rate": 9.288398969167665e-06, + "loss": 7.4395222663879395, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3279309682548046, + "loss_total_v6": 0.4649701453745365, + "step": 911 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21033513513513513, + "grad_norm": 23.916261672973633, + "learning_rate": 9.286444930948496e-06, + "loss": 7.473777770996094, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.33126210421323776, + "loss_total_v6": 0.46711110323667526, + "step": 912 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21056576576576577, + "grad_norm": 24.039581298828125, + "learning_rate": 9.284488419662151e-06, + "loss": 7.303046703338623, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.31772277504205704, + "loss_total_v6": 0.45644042640924454, + "step": 913 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21079639639639638, + "grad_norm": 23.814409255981445, + "learning_rate": 9.282529436437445e-06, + "loss": 7.048999786376953, + "loss_alignment": 0.2755126953125, + "loss_alignment_w": 0.13775634765625, + "loss_sft": 0.3027450777590275, + "loss_total_v6": 0.440562479197979, + "step": 914 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21102702702702703, + "grad_norm": 28.424297332763672, + "learning_rate": 9.280567982404611e-06, + "loss": 7.465554237365723, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.32699451223015785, + "loss_total_v6": 0.46659716591238976, + "step": 915 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21125765765765767, + "grad_norm": 24.421630859375, + "learning_rate": 9.278604058695313e-06, + "loss": 7.466895580291748, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.33046574890613556, + "loss_total_v6": 0.46668096259236336, + "step": 916 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21148828828828828, + "grad_norm": 24.11271095275879, + "learning_rate": 9.276637666442636e-06, + "loss": 6.7838640213012695, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.2869218271225691, + "loss_total_v6": 0.42399152368307114, + "step": 917 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21171891891891892, + "grad_norm": 24.87441635131836, + "learning_rate": 9.274668806781093e-06, + "loss": 7.497722148895264, + "loss_alignment": 0.278076171875, + "loss_alignment_w": 0.1390380859375, + "loss_sft": 0.3298899866640568, + "loss_total_v6": 0.4686076268553734, + "step": 918 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21194954954954956, + "grad_norm": 24.186565399169922, + "learning_rate": 9.272697480846618e-06, + "loss": 7.361722469329834, + "loss_alignment": 0.284423828125, + "loss_alignment_w": 0.1422119140625, + "loss_sft": 0.31804831326007843, + "loss_total_v6": 0.460107646882534, + "step": 919 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21218018018018017, + "grad_norm": 24.775936126708984, + "learning_rate": 9.270723689776568e-06, + "loss": 7.155124664306641, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.30873705819249153, + "loss_total_v6": 0.44719531014561653, + "step": 920 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2124108108108108, + "grad_norm": 24.038700103759766, + "learning_rate": 9.268747434709726e-06, + "loss": 6.77979850769043, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.2861641254276037, + "loss_total_v6": 0.42373738437891006, + "step": 921 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21264144144144145, + "grad_norm": 22.80710220336914, + "learning_rate": 9.26676871678629e-06, + "loss": 7.4118828773498535, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.32920949906110764, + "loss_total_v6": 0.46324268355965614, + "step": 922 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21287207207207207, + "grad_norm": 24.99994659423828, + "learning_rate": 9.264787537147886e-06, + "loss": 7.065951347351074, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3054372314363718, + "loss_total_v6": 0.44162194058299065, + "step": 923 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2131027027027027, + "grad_norm": 23.839214324951172, + "learning_rate": 9.262803896937556e-06, + "loss": 7.304960250854492, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3212451059371233, + "loss_total_v6": 0.45656004175543785, + "step": 924 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21333333333333335, + "grad_norm": 25.29839324951172, + "learning_rate": 9.26081779729976e-06, + "loss": 7.820987701416016, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.35061289742588997, + "loss_total_v6": 0.4888117350637913, + "step": 925 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21356396396396396, + "grad_norm": 24.16594696044922, + "learning_rate": 9.25882923938038e-06, + "loss": 6.784292221069336, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.2875283695757389, + "loss_total_v6": 0.4240182526409626, + "step": 926 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2137945945945946, + "grad_norm": 21.997079849243164, + "learning_rate": 9.256838224326722e-06, + "loss": 7.085813045501709, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3071668930351734, + "loss_total_v6": 0.4428633116185665, + "step": 927 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2140252252252252, + "grad_norm": 25.192245483398438, + "learning_rate": 9.254844753287493e-06, + "loss": 7.326826095581055, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.3214062415063381, + "loss_total_v6": 0.4579266421496868, + "step": 928 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21425585585585585, + "grad_norm": 24.137401580810547, + "learning_rate": 9.252848827412835e-06, + "loss": 7.030789852142334, + "loss_alignment": 0.2711181640625, + "loss_alignment_w": 0.13555908203125, + "loss_sft": 0.3038805387914181, + "loss_total_v6": 0.4394243620336056, + "step": 929 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2144864864864865, + "grad_norm": 21.85614013671875, + "learning_rate": 9.250850447854296e-06, + "loss": 7.05927038192749, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.30422625690698624, + "loss_total_v6": 0.44120440632104874, + "step": 930 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2147171171171171, + "grad_norm": 23.602949142456055, + "learning_rate": 9.248849615764846e-06, + "loss": 7.176816463470459, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3118933141231537, + "loss_total_v6": 0.448551032692194, + "step": 931 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21494774774774775, + "grad_norm": 24.496973037719727, + "learning_rate": 9.24684633229886e-06, + "loss": 7.077799320220947, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.3023630678653717, + "loss_total_v6": 0.4423624500632286, + "step": 932 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2151783783783784, + "grad_norm": 23.952775955200195, + "learning_rate": 9.244840598612137e-06, + "loss": 7.052988052368164, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.30407773330807686, + "loss_total_v6": 0.44081173464655876, + "step": 933 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.215409009009009, + "grad_norm": 24.29390525817871, + "learning_rate": 9.242832415861888e-06, + "loss": 6.777052402496338, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.2894410453736782, + "loss_total_v6": 0.4235657900571823, + "step": 934 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21563963963963964, + "grad_norm": 23.080228805541992, + "learning_rate": 9.240821785206732e-06, + "loss": 7.259625434875488, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3188541643321514, + "loss_total_v6": 0.4537265934050083, + "step": 935 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21587027027027028, + "grad_norm": 25.03331756591797, + "learning_rate": 9.238808707806706e-06, + "loss": 7.303825378417969, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.31851913034915924, + "loss_total_v6": 0.45648909360170364, + "step": 936 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2161009009009009, + "grad_norm": 23.019277572631836, + "learning_rate": 9.236793184823257e-06, + "loss": 7.0064778327941895, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3009724784642458, + "loss_total_v6": 0.43790485709905624, + "step": 937 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21633153153153153, + "grad_norm": 22.403825759887695, + "learning_rate": 9.23477521741924e-06, + "loss": 7.153481483459473, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.31222012639045715, + "loss_total_v6": 0.44709257036447525, + "step": 938 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21656216216216218, + "grad_norm": 24.864999771118164, + "learning_rate": 9.232754806758923e-06, + "loss": 6.850374221801758, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.2914296109229326, + "loss_total_v6": 0.42814837396144867, + "step": 939 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2167927927927928, + "grad_norm": 22.637422561645508, + "learning_rate": 9.230731954007983e-06, + "loss": 7.404547691345215, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3270725514739752, + "loss_total_v6": 0.4627842307090759, + "step": 940 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21702342342342343, + "grad_norm": 24.192995071411133, + "learning_rate": 9.228706660333508e-06, + "loss": 6.826220512390137, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.29469603300094604, + "loss_total_v6": 0.42663878202438354, + "step": 941 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21725405405405404, + "grad_norm": 22.315885543823242, + "learning_rate": 9.226678926903991e-06, + "loss": 6.978635787963867, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3004225268959999, + "loss_total_v6": 0.4361647143959999, + "step": 942 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21748468468468468, + "grad_norm": 32.08873748779297, + "learning_rate": 9.224648754889337e-06, + "loss": 7.215943336486816, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3157578147947788, + "loss_total_v6": 0.4509964697062969, + "step": 943 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21771531531531532, + "grad_norm": 24.459980010986328, + "learning_rate": 9.22261614546085e-06, + "loss": 7.683439254760742, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3417719751596451, + "loss_total_v6": 0.4802149794995785, + "step": 944 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21794594594594593, + "grad_norm": 24.92308807373047, + "learning_rate": 9.220581099791248e-06, + "loss": 7.1359758377075195, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.30987482890486717, + "loss_total_v6": 0.44599849358201027, + "step": 945 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21817657657657658, + "grad_norm": 23.11790657043457, + "learning_rate": 9.218543619054652e-06, + "loss": 7.06906795501709, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3074783459305763, + "loss_total_v6": 0.4418167248368263, + "step": 946 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21840720720720722, + "grad_norm": 21.09989356994629, + "learning_rate": 9.216503704426587e-06, + "loss": 7.40805196762085, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.3236142061650753, + "loss_total_v6": 0.4630032517015934, + "step": 947 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21863783783783783, + "grad_norm": 22.979907989501953, + "learning_rate": 9.214461357083986e-06, + "loss": 7.126716613769531, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3114628829061985, + "loss_total_v6": 0.4454197846353054, + "step": 948 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21886846846846847, + "grad_norm": 24.530529022216797, + "learning_rate": 9.212416578205178e-06, + "loss": 7.314089775085449, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.31780263036489487, + "loss_total_v6": 0.4571306183934212, + "step": 949 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2190990990990991, + "grad_norm": 23.6677188873291, + "learning_rate": 9.210369368969904e-06, + "loss": 6.616133689880371, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.2776288613677025, + "loss_total_v6": 0.4135083705186844, + "step": 950 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21932972972972972, + "grad_norm": 25.214977264404297, + "learning_rate": 9.2083197305593e-06, + "loss": 7.029906272888184, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3043746445327997, + "loss_total_v6": 0.43936915323138237, + "step": 951 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.21956036036036036, + "grad_norm": 27.069028854370117, + "learning_rate": 9.206267664155906e-06, + "loss": 6.51434326171875, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.27248763106763363, + "loss_total_v6": 0.407146442681551, + "step": 952 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.219790990990991, + "grad_norm": 23.043437957763672, + "learning_rate": 9.204213170943666e-06, + "loss": 7.361830711364746, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.32441799715161324, + "loss_total_v6": 0.46011440828442574, + "step": 953 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22002162162162162, + "grad_norm": 22.649463653564453, + "learning_rate": 9.202156252107916e-06, + "loss": 6.938024520874023, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.29773175716400146, + "loss_total_v6": 0.43362653255462646, + "step": 954 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22025225225225226, + "grad_norm": 24.51274299621582, + "learning_rate": 9.200096908835401e-06, + "loss": 7.359385013580322, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.32595887780189514, + "loss_total_v6": 0.45996156334877014, + "step": 955 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22048288288288287, + "grad_norm": 27.256269454956055, + "learning_rate": 9.198035142314259e-06, + "loss": 7.341322898864746, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.3189096041023731, + "loss_total_v6": 0.45883268117904663, + "step": 956 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2207135135135135, + "grad_norm": 22.758081436157227, + "learning_rate": 9.195970953734027e-06, + "loss": 7.423681259155273, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.3275665044784546, + "loss_total_v6": 0.4639800973236561, + "step": 957 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22094414414414415, + "grad_norm": 24.575664520263672, + "learning_rate": 9.193904344285643e-06, + "loss": 6.899669170379639, + "loss_alignment": 0.2664794921875, + "loss_alignment_w": 0.13323974609375, + "loss_sft": 0.2977759465575218, + "loss_total_v6": 0.4312293194234371, + "step": 958 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22117477477477476, + "grad_norm": 25.057632446289062, + "learning_rate": 9.191835315161432e-06, + "loss": 7.138950824737549, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.3083823211491108, + "loss_total_v6": 0.4461844265460968, + "step": 959 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2214054054054054, + "grad_norm": 22.385326385498047, + "learning_rate": 9.18976386755513e-06, + "loss": 7.044180870056152, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3031153380870819, + "loss_total_v6": 0.440261323004961, + "step": 960 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22163603603603604, + "grad_norm": 25.478151321411133, + "learning_rate": 9.187690002661853e-06, + "loss": 7.236872673034668, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.31380053237080574, + "loss_total_v6": 0.45230456814169884, + "step": 961 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22186666666666666, + "grad_norm": 23.42085838317871, + "learning_rate": 9.185613721678123e-06, + "loss": 6.8929338455200195, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.29578336142003536, + "loss_total_v6": 0.4308083839714527, + "step": 962 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2220972972972973, + "grad_norm": 23.68161392211914, + "learning_rate": 9.18353502580185e-06, + "loss": 7.307032585144043, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3206269107758999, + "loss_total_v6": 0.4566895440220833, + "step": 963 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22232792792792794, + "grad_norm": 24.35551643371582, + "learning_rate": 9.181453916232339e-06, + "loss": 7.118260383605957, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.3110106736421585, + "loss_total_v6": 0.4448912739753723, + "step": 964 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22255855855855855, + "grad_norm": 26.513973236083984, + "learning_rate": 9.179370394170289e-06, + "loss": 7.331345081329346, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.3193540796637535, + "loss_total_v6": 0.4582090638577938, + "step": 965 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2227891891891892, + "grad_norm": 23.125701904296875, + "learning_rate": 9.177284460817788e-06, + "loss": 6.945684909820557, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.2975848913192749, + "loss_total_v6": 0.4341052919626236, + "step": 966 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22301981981981983, + "grad_norm": 24.866804122924805, + "learning_rate": 9.175196117378316e-06, + "loss": 7.368197917938232, + "loss_alignment": 0.278076171875, + "loss_alignment_w": 0.1390380859375, + "loss_sft": 0.32132167741656303, + "loss_total_v6": 0.46051235496997833, + "step": 967 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22325045045045044, + "grad_norm": 25.3885498046875, + "learning_rate": 9.173105365056742e-06, + "loss": 6.614959239959717, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.27483937330543995, + "loss_total_v6": 0.4134349636733532, + "step": 968 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22348108108108108, + "grad_norm": 26.940715789794922, + "learning_rate": 9.171012205059333e-06, + "loss": 6.955265045166016, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.29812265560030937, + "loss_total_v6": 0.43470407649874687, + "step": 969 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22371171171171172, + "grad_norm": 25.141427993774414, + "learning_rate": 9.168916638593736e-06, + "loss": 6.938648223876953, + "loss_alignment": 0.28271484375, + "loss_alignment_w": 0.141357421875, + "loss_sft": 0.29233864694833755, + "loss_total_v6": 0.43366553634405136, + "step": 970 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22394234234234234, + "grad_norm": 23.857177734375, + "learning_rate": 9.166818666868988e-06, + "loss": 7.4165263175964355, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3283094950020313, + "loss_total_v6": 0.4635328873991966, + "step": 971 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22417297297297298, + "grad_norm": 25.650930404663086, + "learning_rate": 9.164718291095515e-06, + "loss": 7.035210132598877, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3041568212211132, + "loss_total_v6": 0.4397006370127201, + "step": 972 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2244036036036036, + "grad_norm": 24.63576316833496, + "learning_rate": 9.162615512485135e-06, + "loss": 7.012528419494629, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3019151873886585, + "loss_total_v6": 0.4382830001413822, + "step": 973 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22463423423423423, + "grad_norm": 20.649538040161133, + "learning_rate": 9.16051033225104e-06, + "loss": 7.273676872253418, + "loss_alignment": 0.2691650390625, + "loss_alignment_w": 0.13458251953125, + "loss_sft": 0.32003757543861866, + "loss_total_v6": 0.4546048305928707, + "step": 974 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22486486486486487, + "grad_norm": 23.518798828125, + "learning_rate": 9.158402751607822e-06, + "loss": 7.275275230407715, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.31745188124477863, + "loss_total_v6": 0.454704686999321, + "step": 975 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22509549549549548, + "grad_norm": 23.016040802001953, + "learning_rate": 9.156292771771446e-06, + "loss": 7.180117130279541, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3117181360721588, + "loss_total_v6": 0.4487573206424713, + "step": 976 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22532612612612613, + "grad_norm": 20.06507110595703, + "learning_rate": 9.15418039395927e-06, + "loss": 6.782174110412598, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2911191303282976, + "loss_total_v6": 0.42388586699962616, + "step": 977 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22555675675675677, + "grad_norm": 22.194561004638672, + "learning_rate": 9.15206561939003e-06, + "loss": 7.109041690826416, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3102361224591732, + "loss_total_v6": 0.4443151019513607, + "step": 978 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22578738738738738, + "grad_norm": 23.107486724853516, + "learning_rate": 9.149948449283846e-06, + "loss": 7.024462699890137, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.30426329374313354, + "loss_total_v6": 0.43902892246842384, + "step": 979 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22601801801801802, + "grad_norm": 32.45387649536133, + "learning_rate": 9.14782888486222e-06, + "loss": 7.313509464263916, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.31961265951395035, + "loss_total_v6": 0.45709433779120445, + "step": 980 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22624864864864866, + "grad_norm": 23.199119567871094, + "learning_rate": 9.14570692734804e-06, + "loss": 7.3308796882629395, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3216901160776615, + "loss_total_v6": 0.4581799805164337, + "step": 981 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22647927927927927, + "grad_norm": 24.403669357299805, + "learning_rate": 9.143582577965567e-06, + "loss": 7.324155807495117, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3232230097055435, + "loss_total_v6": 0.4577597416937351, + "step": 982 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2267099099099099, + "grad_norm": 25.11560821533203, + "learning_rate": 9.141455837940446e-06, + "loss": 7.046595573425293, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3018319234251976, + "loss_total_v6": 0.440412238240242, + "step": 983 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22694054054054055, + "grad_norm": 26.2879695892334, + "learning_rate": 9.139326708499701e-06, + "loss": 7.24272346496582, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3174620680510998, + "loss_total_v6": 0.4526702016592026, + "step": 984 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22717117117117117, + "grad_norm": 25.32889175415039, + "learning_rate": 9.137195190871737e-06, + "loss": 7.49298620223999, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.3310130573809147, + "loss_total_v6": 0.4683116376399994, + "step": 985 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2274018018018018, + "grad_norm": 22.513593673706055, + "learning_rate": 9.135061286286333e-06, + "loss": 7.207869529724121, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.3170384783297777, + "loss_total_v6": 0.45049185305833817, + "step": 986 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22763243243243242, + "grad_norm": 23.873647689819336, + "learning_rate": 9.132924995974646e-06, + "loss": 7.4226789474487305, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.32805320993065834, + "loss_total_v6": 0.46391746029257774, + "step": 987 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22786306306306306, + "grad_norm": 25.134164810180664, + "learning_rate": 9.130786321169209e-06, + "loss": 7.498241424560547, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.3299682214856148, + "loss_total_v6": 0.4686400964856148, + "step": 988 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2280936936936937, + "grad_norm": 22.81717872619629, + "learning_rate": 9.128645263103937e-06, + "loss": 7.465930938720703, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.33214496448636055, + "loss_total_v6": 0.46662066876888275, + "step": 989 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2283243243243243, + "grad_norm": 24.17848014831543, + "learning_rate": 9.12650182301411e-06, + "loss": 7.222921848297119, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.31338636204600334, + "loss_total_v6": 0.45143262296915054, + "step": 990 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22855495495495495, + "grad_norm": 22.283092498779297, + "learning_rate": 9.124356002136386e-06, + "loss": 7.366006851196289, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.32358039170503616, + "loss_total_v6": 0.46037542447447777, + "step": 991 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2287855855855856, + "grad_norm": 26.291202545166016, + "learning_rate": 9.122207801708802e-06, + "loss": 7.178830146789551, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.31558969989418983, + "loss_total_v6": 0.4486768916249275, + "step": 992 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2290162162162162, + "grad_norm": 23.5235652923584, + "learning_rate": 9.120057222970763e-06, + "loss": 7.229353904724121, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.31745046004652977, + "loss_total_v6": 0.4518346078693867, + "step": 993 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22924684684684685, + "grad_norm": 22.80474281311035, + "learning_rate": 9.117904267163044e-06, + "loss": 6.780423164367676, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.2863405477255583, + "loss_total_v6": 0.42377646639943123, + "step": 994 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2294774774774775, + "grad_norm": 25.787532806396484, + "learning_rate": 9.115748935527795e-06, + "loss": 7.589056968688965, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3394131362438202, + "loss_total_v6": 0.4743160866200924, + "step": 995 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2297081081081081, + "grad_norm": 24.462724685668945, + "learning_rate": 9.113591229308538e-06, + "loss": 7.064664840698242, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.3034647610038519, + "loss_total_v6": 0.44154154881834984, + "step": 996 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.22993873873873874, + "grad_norm": 24.02509307861328, + "learning_rate": 9.111431149750164e-06, + "loss": 7.522653102874756, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.33297404646873474, + "loss_total_v6": 0.47016581520438194, + "step": 997 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23016936936936938, + "grad_norm": 23.772361755371094, + "learning_rate": 9.109268698098927e-06, + "loss": 7.689481735229492, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.34089842066168785, + "loss_total_v6": 0.48059263825416565, + "step": 998 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2304, + "grad_norm": 24.6663818359375, + "learning_rate": 9.107103875602458e-06, + "loss": 7.46080207824707, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3314582295715809, + "loss_total_v6": 0.466300118714571, + "step": 999 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23063063063063063, + "grad_norm": 23.802095413208008, + "learning_rate": 9.104936683509755e-06, + "loss": 7.301963806152344, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.31881478428840637, + "loss_total_v6": 0.4563727416098118, + "step": 1000 + }, + { + "epoch": 0.23063063063063063, + "eval_loss": 0.6027373671531677, + "eval_loss_alignment": 0.2458083618721461, + "eval_loss_alignment_w": 0.12290418093607305, + "eval_loss_sft": 0.47983315974882207, + "eval_loss_total_v6": 0.6027373409570624, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.2985, + "eval_samples_per_second": 19.113, + "eval_steps_per_second": 2.399, + "step": 1000 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23086126126126127, + "grad_norm": 23.17778778076172, + "learning_rate": 9.102767123071177e-06, + "loss": 7.252513885498047, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3167922478169203, + "loss_total_v6": 0.45328211784362793, + "step": 1001 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2310918918918919, + "grad_norm": 22.794248580932617, + "learning_rate": 9.100595195538456e-06, + "loss": 7.349677085876465, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.32249871641397476, + "loss_total_v6": 0.45935481041669846, + "step": 1002 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23132252252252253, + "grad_norm": 24.31747817993164, + "learning_rate": 9.098420902164684e-06, + "loss": 7.234851837158203, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.3148033656179905, + "loss_total_v6": 0.4521782509982586, + "step": 1003 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23155315315315314, + "grad_norm": 24.58460807800293, + "learning_rate": 9.096244244204325e-06, + "loss": 6.956292152404785, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.2962642312049866, + "loss_total_v6": 0.43476827815175056, + "step": 1004 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23178378378378378, + "grad_norm": 24.042720794677734, + "learning_rate": 9.094065222913199e-06, + "loss": 6.880568981170654, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.2950257994234562, + "loss_total_v6": 0.4300355613231659, + "step": 1005 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23201441441441442, + "grad_norm": 22.17411231994629, + "learning_rate": 9.091883839548497e-06, + "loss": 7.117405891418457, + "loss_alignment": 0.2750244140625, + "loss_alignment_w": 0.13751220703125, + "loss_sft": 0.30735617503523827, + "loss_total_v6": 0.44483786076307297, + "step": 1006 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23224504504504503, + "grad_norm": 23.5369873046875, + "learning_rate": 9.089700095368768e-06, + "loss": 6.777456283569336, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.28745211847126484, + "loss_total_v6": 0.423591036349535, + "step": 1007 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23247567567567567, + "grad_norm": 23.43639373779297, + "learning_rate": 9.087513991633926e-06, + "loss": 7.443058490753174, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.32836560532450676, + "loss_total_v6": 0.46519115567207336, + "step": 1008 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23270630630630632, + "grad_norm": 25.809412002563477, + "learning_rate": 9.085325529605242e-06, + "loss": 7.023364067077637, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3052932657301426, + "loss_total_v6": 0.438960250467062, + "step": 1009 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23293693693693693, + "grad_norm": 23.619075775146484, + "learning_rate": 9.083134710545354e-06, + "loss": 7.20265007019043, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.31385887786746025, + "loss_total_v6": 0.45016563683748245, + "step": 1010 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23316756756756757, + "grad_norm": 25.325105667114258, + "learning_rate": 9.080941535718254e-06, + "loss": 7.331366539001465, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.32197991386055946, + "loss_total_v6": 0.45821039378643036, + "step": 1011 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2333981981981982, + "grad_norm": 21.999826431274414, + "learning_rate": 9.078746006389299e-06, + "loss": 7.148412227630615, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.31011804565787315, + "loss_total_v6": 0.44677576050162315, + "step": 1012 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23362882882882882, + "grad_norm": 25.113462448120117, + "learning_rate": 9.076548123825195e-06, + "loss": 7.459209442138672, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.3279712572693825, + "loss_total_v6": 0.4662006162106991, + "step": 1013 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23385945945945946, + "grad_norm": 24.387447357177734, + "learning_rate": 9.074347889294017e-06, + "loss": 7.186807632446289, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.31413520872592926, + "loss_total_v6": 0.44917549192905426, + "step": 1014 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2340900900900901, + "grad_norm": 23.010536193847656, + "learning_rate": 9.072145304065193e-06, + "loss": 7.061408996582031, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.30617570504546165, + "loss_total_v6": 0.44133804738521576, + "step": 1015 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23432072072072072, + "grad_norm": 22.811128616333008, + "learning_rate": 9.0699403694095e-06, + "loss": 7.166318416595459, + "loss_alignment": 0.279296875, + "loss_alignment_w": 0.1396484375, + "loss_sft": 0.3080938868224621, + "loss_total_v6": 0.4478949122130871, + "step": 1016 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23455135135135136, + "grad_norm": 23.20501708984375, + "learning_rate": 9.06773308659908e-06, + "loss": 6.930508613586426, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.2951257489621639, + "loss_total_v6": 0.4331567771732807, + "step": 1017 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23478198198198197, + "grad_norm": 24.995643615722656, + "learning_rate": 9.065523456907425e-06, + "loss": 7.125, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.3139801248908043, + "loss_total_v6": 0.4453125223517418, + "step": 1018 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2350126126126126, + "grad_norm": 23.877368927001953, + "learning_rate": 9.063311481609387e-06, + "loss": 7.382889747619629, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.3289385549724102, + "loss_total_v6": 0.4614306204020977, + "step": 1019 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23524324324324325, + "grad_norm": 25.0568790435791, + "learning_rate": 9.061097161981159e-06, + "loss": 7.2996015548706055, + "loss_alignment": 0.27978515625, + "loss_alignment_w": 0.139892578125, + "loss_sft": 0.31662243977189064, + "loss_total_v6": 0.45622508972883224, + "step": 1020 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23547387387387386, + "grad_norm": 23.08808135986328, + "learning_rate": 9.0588804993003e-06, + "loss": 7.316226482391357, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.32240698114037514, + "loss_total_v6": 0.45726415887475014, + "step": 1021 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2357045045045045, + "grad_norm": 23.715734481811523, + "learning_rate": 9.05666149484571e-06, + "loss": 7.622123718261719, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3401522599160671, + "loss_total_v6": 0.4763827435672283, + "step": 1022 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23593513513513514, + "grad_norm": 25.253549575805664, + "learning_rate": 9.054440149897652e-06, + "loss": 7.49928092956543, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.33223044127225876, + "loss_total_v6": 0.46870507299900055, + "step": 1023 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23616576576576576, + "grad_norm": 23.19316864013672, + "learning_rate": 9.052216465737727e-06, + "loss": 6.440119743347168, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.2672535665333271, + "loss_total_v6": 0.4025074616074562, + "step": 1024 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2363963963963964, + "grad_norm": 21.89507293701172, + "learning_rate": 9.04999044364889e-06, + "loss": 7.0955305099487305, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.30463093891739845, + "loss_total_v6": 0.44347064942121506, + "step": 1025 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23662702702702704, + "grad_norm": 26.6163272857666, + "learning_rate": 9.04776208491545e-06, + "loss": 7.65644645690918, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.3401611503213644, + "loss_total_v6": 0.47852787375450134, + "step": 1026 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23685765765765765, + "grad_norm": 23.66944122314453, + "learning_rate": 9.045531390823058e-06, + "loss": 6.8426947593688965, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.29194149002432823, + "loss_total_v6": 0.42766841500997543, + "step": 1027 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2370882882882883, + "grad_norm": 25.716449737548828, + "learning_rate": 9.043298362658715e-06, + "loss": 7.781721115112305, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.35037126764655113, + "loss_total_v6": 0.48635757341980934, + "step": 1028 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23731891891891893, + "grad_norm": 24.09894561767578, + "learning_rate": 9.041063001710767e-06, + "loss": 6.703008651733398, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.28633918426930904, + "loss_total_v6": 0.4189380556344986, + "step": 1029 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23754954954954954, + "grad_norm": 24.912425994873047, + "learning_rate": 9.038825309268908e-06, + "loss": 7.375519752502441, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.3268147148191929, + "loss_total_v6": 0.4609699845314026, + "step": 1030 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23778018018018018, + "grad_norm": 21.97924041748047, + "learning_rate": 9.036585286624175e-06, + "loss": 7.157590866088867, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.31211075745522976, + "loss_total_v6": 0.447349414229393, + "step": 1031 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23801081081081082, + "grad_norm": 22.533878326416016, + "learning_rate": 9.034342935068953e-06, + "loss": 7.269911766052246, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.3170861564576626, + "loss_total_v6": 0.4543694704771042, + "step": 1032 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23824144144144144, + "grad_norm": 23.349056243896484, + "learning_rate": 9.032098255896965e-06, + "loss": 6.940308570861816, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.2971420884132385, + "loss_total_v6": 0.4337693005800247, + "step": 1033 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23847207207207208, + "grad_norm": 21.80149269104004, + "learning_rate": 9.029851250403282e-06, + "loss": 7.149692535400391, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.30790927074849606, + "loss_total_v6": 0.4468557946383953, + "step": 1034 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2387027027027027, + "grad_norm": 23.44370460510254, + "learning_rate": 9.027601919884314e-06, + "loss": 7.236804008483887, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3159629553556442, + "loss_total_v6": 0.4523002430796623, + "step": 1035 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23893333333333333, + "grad_norm": 22.94554901123047, + "learning_rate": 9.025350265637816e-06, + "loss": 7.094198703765869, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3049291577190161, + "loss_total_v6": 0.4433874152600765, + "step": 1036 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23916396396396397, + "grad_norm": 22.170795440673828, + "learning_rate": 9.023096288962878e-06, + "loss": 7.042516708374023, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3091453090310097, + "loss_total_v6": 0.44015728309750557, + "step": 1037 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23939459459459458, + "grad_norm": 21.030845642089844, + "learning_rate": 9.020839991159937e-06, + "loss": 7.23630428314209, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.3147873356938362, + "loss_total_v6": 0.4522690176963806, + "step": 1038 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23962522522522522, + "grad_norm": 22.353675842285156, + "learning_rate": 9.018581373530762e-06, + "loss": 7.361452579498291, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.32575240544974804, + "loss_total_v6": 0.4600907862186432, + "step": 1039 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.23985585585585587, + "grad_norm": 24.477567672729492, + "learning_rate": 9.016320437378466e-06, + "loss": 7.712026119232178, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3468392677605152, + "loss_total_v6": 0.4820016250014305, + "step": 1040 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24008648648648648, + "grad_norm": 22.76725959777832, + "learning_rate": 9.014057184007497e-06, + "loss": 7.05885124206543, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.305298687890172, + "loss_total_v6": 0.44117819890379906, + "step": 1041 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24031711711711712, + "grad_norm": 21.14908218383789, + "learning_rate": 9.011791614723642e-06, + "loss": 7.583495616912842, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3379668928682804, + "loss_total_v6": 0.4739684760570526, + "step": 1042 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24054774774774776, + "grad_norm": 22.26528549194336, + "learning_rate": 9.00952373083402e-06, + "loss": 7.562374114990234, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.33638738840818405, + "loss_total_v6": 0.47264838218688965, + "step": 1043 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24077837837837837, + "grad_norm": 23.82785415649414, + "learning_rate": 9.007253533647088e-06, + "loss": 7.381798267364502, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.3243079371750355, + "loss_total_v6": 0.46136239916086197, + "step": 1044 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.241009009009009, + "grad_norm": 24.381433486938477, + "learning_rate": 9.004981024472641e-06, + "loss": 7.292712211608887, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3199455011636019, + "loss_total_v6": 0.4557945020496845, + "step": 1045 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24123963963963965, + "grad_norm": 26.26142120361328, + "learning_rate": 9.002706204621802e-06, + "loss": 7.295964241027832, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3206828534603119, + "loss_total_v6": 0.4559977799654007, + "step": 1046 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24147027027027027, + "grad_norm": 24.59134292602539, + "learning_rate": 9.00042907540703e-06, + "loss": 7.385979175567627, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.32189893908798695, + "loss_total_v6": 0.4616236835718155, + "step": 1047 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2417009009009009, + "grad_norm": 23.125043869018555, + "learning_rate": 8.99814963814212e-06, + "loss": 6.73052978515625, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.28456495702266693, + "loss_total_v6": 0.42065810412168503, + "step": 1048 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24193153153153152, + "grad_norm": 23.89820671081543, + "learning_rate": 8.995867894142187e-06, + "loss": 6.90004825592041, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.29506832361221313, + "loss_total_v6": 0.43125301226973534, + "step": 1049 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24216216216216216, + "grad_norm": 23.383657455444336, + "learning_rate": 8.993583844723693e-06, + "loss": 7.116683006286621, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.3082875572144985, + "loss_total_v6": 0.4447926878929138, + "step": 1050 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2423927927927928, + "grad_norm": 23.032684326171875, + "learning_rate": 8.991297491204415e-06, + "loss": 7.250757217407227, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.31694187596440315, + "loss_total_v6": 0.45317233726382256, + "step": 1051 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2426234234234234, + "grad_norm": 31.43733787536621, + "learning_rate": 8.98900883490347e-06, + "loss": 7.210237503051758, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.31387530639767647, + "loss_total_v6": 0.4506398178637028, + "step": 1052 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24285405405405405, + "grad_norm": 23.802032470703125, + "learning_rate": 8.986717877141297e-06, + "loss": 7.460331916809082, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.32889584824442863, + "loss_total_v6": 0.46627073362469673, + "step": 1053 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2430846846846847, + "grad_norm": 23.719337463378906, + "learning_rate": 8.984424619239671e-06, + "loss": 7.184748649597168, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.31251115910708904, + "loss_total_v6": 0.4490467980504036, + "step": 1054 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2433153153153153, + "grad_norm": 24.470212936401367, + "learning_rate": 8.982129062521683e-06, + "loss": 7.178015232086182, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.3131889346987009, + "loss_total_v6": 0.44862595945596695, + "step": 1055 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24354594594594595, + "grad_norm": 26.14818000793457, + "learning_rate": 8.979831208311758e-06, + "loss": 6.7829694747924805, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.289551455527544, + "loss_total_v6": 0.4239356070756912, + "step": 1056 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2437765765765766, + "grad_norm": 23.902545928955078, + "learning_rate": 8.977531057935644e-06, + "loss": 7.30244255065918, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.31754766404628754, + "loss_total_v6": 0.45640265941619873, + "step": 1057 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2440072072072072, + "grad_norm": 23.44525146484375, + "learning_rate": 8.975228612720415e-06, + "loss": 7.157785892486572, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3119245767593384, + "loss_total_v6": 0.44736161455512047, + "step": 1058 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24423783783783784, + "grad_norm": 25.604331970214844, + "learning_rate": 8.972923873994468e-06, + "loss": 7.4152679443359375, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3276052288711071, + "loss_total_v6": 0.4634542241692543, + "step": 1059 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24446846846846848, + "grad_norm": 25.811288833618164, + "learning_rate": 8.970616843087526e-06, + "loss": 7.249579429626465, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.31506771221756935, + "loss_total_v6": 0.45309872552752495, + "step": 1060 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2446990990990991, + "grad_norm": 22.634689331054688, + "learning_rate": 8.968307521330626e-06, + "loss": 6.960927486419678, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.30087218061089516, + "loss_total_v6": 0.43505798280239105, + "step": 1061 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24492972972972973, + "grad_norm": 25.127212524414062, + "learning_rate": 8.965995910056135e-06, + "loss": 7.27579402923584, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3183693513274193, + "loss_total_v6": 0.4547371380031109, + "step": 1062 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24516036036036035, + "grad_norm": 24.292551040649414, + "learning_rate": 8.963682010597744e-06, + "loss": 7.256462097167969, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.31609295681118965, + "loss_total_v6": 0.45352886617183685, + "step": 1063 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.245390990990991, + "grad_norm": 23.018918991088867, + "learning_rate": 8.961365824290451e-06, + "loss": 7.383659839630127, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.32552293315529823, + "loss_total_v6": 0.46147873252630234, + "step": 1064 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24562162162162163, + "grad_norm": 26.714805603027344, + "learning_rate": 8.959047352470584e-06, + "loss": 7.086267471313477, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.3126731850206852, + "loss_total_v6": 0.4428916983306408, + "step": 1065 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24585225225225224, + "grad_norm": 24.975379943847656, + "learning_rate": 8.956726596475792e-06, + "loss": 7.4522175788879395, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3302197866141796, + "loss_total_v6": 0.4657635875046253, + "step": 1066 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24608288288288288, + "grad_norm": 26.64065170288086, + "learning_rate": 8.954403557645029e-06, + "loss": 7.201793670654297, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.31406474113464355, + "loss_total_v6": 0.45011211931705475, + "step": 1067 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24631351351351352, + "grad_norm": 25.13797950744629, + "learning_rate": 8.952078237318576e-06, + "loss": 7.239932537078857, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.3150140941143036, + "loss_total_v6": 0.452495776116848, + "step": 1068 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24654414414414413, + "grad_norm": 23.04503059387207, + "learning_rate": 8.94975063683803e-06, + "loss": 6.691961288452148, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.2837413474917412, + "loss_total_v6": 0.4182475842535496, + "step": 1069 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24677477477477477, + "grad_norm": 21.919471740722656, + "learning_rate": 8.947420757546301e-06, + "loss": 6.824467658996582, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.2924655079841614, + "loss_total_v6": 0.426529236137867, + "step": 1070 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24700540540540541, + "grad_norm": 22.061145782470703, + "learning_rate": 8.945088600787614e-06, + "loss": 7.731888771057129, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.34579184278845787, + "loss_total_v6": 0.4832430072128773, + "step": 1071 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24723603603603603, + "grad_norm": 23.036590576171875, + "learning_rate": 8.942754167907506e-06, + "loss": 6.601680755615234, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.28015877678990364, + "loss_total_v6": 0.41260505840182304, + "step": 1072 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24746666666666667, + "grad_norm": 23.313379287719727, + "learning_rate": 8.940417460252834e-06, + "loss": 7.355046272277832, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3215678371489048, + "loss_total_v6": 0.4596904143691063, + "step": 1073 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2476972972972973, + "grad_norm": 24.601577758789062, + "learning_rate": 8.938078479171761e-06, + "loss": 7.096470832824707, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3090384714305401, + "loss_total_v6": 0.4435294307768345, + "step": 1074 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24792792792792792, + "grad_norm": 23.704404830932617, + "learning_rate": 8.935737226013763e-06, + "loss": 6.896193027496338, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.2944306544959545, + "loss_total_v6": 0.4310120642185211, + "step": 1075 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24815855855855856, + "grad_norm": 22.46504020690918, + "learning_rate": 8.933393702129627e-06, + "loss": 7.267546653747559, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.31847947277128696, + "loss_total_v6": 0.4542216584086418, + "step": 1076 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2483891891891892, + "grad_norm": 23.26019859313965, + "learning_rate": 8.931047908871455e-06, + "loss": 7.372697830200195, + "loss_alignment": 0.27880859375, + "loss_alignment_w": 0.139404296875, + "loss_sft": 0.3214198648929596, + "loss_total_v6": 0.4607936404645443, + "step": 1077 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24861981981981982, + "grad_norm": 23.7186222076416, + "learning_rate": 8.928699847592647e-06, + "loss": 7.207380294799805, + "loss_alignment": 0.2723388671875, + "loss_alignment_w": 0.13616943359375, + "loss_sft": 0.3140324167907238, + "loss_total_v6": 0.450461246073246, + "step": 1078 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24885045045045046, + "grad_norm": 25.343111038208008, + "learning_rate": 8.926349519647924e-06, + "loss": 7.402736186981201, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3240754157304764, + "loss_total_v6": 0.4626710042357445, + "step": 1079 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24908108108108107, + "grad_norm": 20.43378448486328, + "learning_rate": 8.923996926393306e-06, + "loss": 6.730903625488281, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.28478667326271534, + "loss_total_v6": 0.4206814579665661, + "step": 1080 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2493117117117117, + "grad_norm": 24.505332946777344, + "learning_rate": 8.921642069186125e-06, + "loss": 7.501039505004883, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.32883085310459137, + "loss_total_v6": 0.46881498396396637, + "step": 1081 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24954234234234235, + "grad_norm": 21.39565658569336, + "learning_rate": 8.919284949385016e-06, + "loss": 6.938204765319824, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.2996351234614849, + "loss_total_v6": 0.4336377941071987, + "step": 1082 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.24977297297297296, + "grad_norm": 24.063207626342773, + "learning_rate": 8.916925568349922e-06, + "loss": 6.9855241775512695, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.30063943937420845, + "loss_total_v6": 0.43659523874521255, + "step": 1083 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25000360360360363, + "grad_norm": 28.05557632446289, + "learning_rate": 8.91456392744209e-06, + "loss": 6.95284366607666, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.3040443006902933, + "loss_total_v6": 0.43455271422863007, + "step": 1084 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25023423423423424, + "grad_norm": 23.496116638183594, + "learning_rate": 8.912200028024069e-06, + "loss": 7.399381160736084, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.32732948660850525, + "loss_total_v6": 0.46246131509542465, + "step": 1085 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25046486486486486, + "grad_norm": 21.95627784729004, + "learning_rate": 8.909833871459711e-06, + "loss": 6.915806293487549, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.2969229482114315, + "loss_total_v6": 0.4322378784418106, + "step": 1086 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25069549549549547, + "grad_norm": 22.45117950439453, + "learning_rate": 8.907465459114175e-06, + "loss": 6.49677038192749, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.27199968695640564, + "loss_total_v6": 0.40604814141988754, + "step": 1087 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25092612612612614, + "grad_norm": 21.156553268432617, + "learning_rate": 8.905094792353917e-06, + "loss": 6.9315714836120605, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.2991594709455967, + "loss_total_v6": 0.4332232028245926, + "step": 1088 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25115675675675675, + "grad_norm": 23.18839454650879, + "learning_rate": 8.902721872546692e-06, + "loss": 6.895401954650879, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.293801324442029, + "loss_total_v6": 0.43096259981393814, + "step": 1089 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25138738738738736, + "grad_norm": 28.885807037353516, + "learning_rate": 8.90034670106156e-06, + "loss": 7.323631286621094, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.32418204471468925, + "loss_total_v6": 0.45772695541381836, + "step": 1090 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25161801801801803, + "grad_norm": 25.03797149658203, + "learning_rate": 8.897969279268877e-06, + "loss": 6.93665885925293, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.29537282697856426, + "loss_total_v6": 0.4335411675274372, + "step": 1091 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25184864864864864, + "grad_norm": 25.647329330444336, + "learning_rate": 8.895589608540299e-06, + "loss": 6.838556289672852, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2973896488547325, + "loss_total_v6": 0.4274097941815853, + "step": 1092 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25207927927927926, + "grad_norm": 27.061969757080078, + "learning_rate": 8.893207690248776e-06, + "loss": 7.028042793273926, + "loss_alignment": 0.278564453125, + "loss_alignment_w": 0.1392822265625, + "loss_sft": 0.30009252205491066, + "loss_total_v6": 0.43925267830491066, + "step": 1093 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2523099099099099, + "grad_norm": 23.867033004760742, + "learning_rate": 8.890823525768562e-06, + "loss": 6.671442031860352, + "loss_alignment": 0.2672119140625, + "loss_alignment_w": 0.13360595703125, + "loss_sft": 0.2834659833461046, + "loss_total_v6": 0.41696513816714287, + "step": 1094 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25254054054054054, + "grad_norm": 22.0096492767334, + "learning_rate": 8.888437116475196e-06, + "loss": 7.242502212524414, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3164564575999975, + "loss_total_v6": 0.45265640690922737, + "step": 1095 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25277117117117115, + "grad_norm": 25.00519561767578, + "learning_rate": 8.886048463745525e-06, + "loss": 7.250191688537598, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.31768471002578735, + "loss_total_v6": 0.45313697680830956, + "step": 1096 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2530018018018018, + "grad_norm": 25.926870346069336, + "learning_rate": 8.88365756895768e-06, + "loss": 6.9475555419921875, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.29580975510179996, + "loss_total_v6": 0.4342222362756729, + "step": 1097 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25323243243243243, + "grad_norm": 21.99610137939453, + "learning_rate": 8.881264433491089e-06, + "loss": 6.945046424865723, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.3000932354480028, + "loss_total_v6": 0.43406540527939796, + "step": 1098 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25346306306306304, + "grad_norm": 23.560728073120117, + "learning_rate": 8.878869058726474e-06, + "loss": 7.276493549346924, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3195879999548197, + "loss_total_v6": 0.45478086173534393, + "step": 1099 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2536936936936937, + "grad_norm": 22.286849975585938, + "learning_rate": 8.876471446045847e-06, + "loss": 7.221289157867432, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3170379772782326, + "loss_total_v6": 0.4513305835425854, + "step": 1100 + }, + { + "epoch": 0.2536936936936937, + "eval_loss": 0.6007767915725708, + "eval_loss_alignment": 0.24482065139840184, + "eval_loss_alignment_w": 0.12241032569920092, + "eval_loss_sft": 0.4783664458350504, + "eval_loss_total_v6": 0.6007767715342512, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.3199, + "eval_samples_per_second": 19.109, + "eval_steps_per_second": 2.398, + "step": 1100 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2539243243243243, + "grad_norm": 22.825197219848633, + "learning_rate": 8.874071596832514e-06, + "loss": 7.318735122680664, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3201528638601303, + "loss_total_v6": 0.4574209339916706, + "step": 1101 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25415495495495494, + "grad_norm": 23.976497650146484, + "learning_rate": 8.871669512471068e-06, + "loss": 7.287543296813965, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3223537877202034, + "loss_total_v6": 0.4554714672267437, + "step": 1102 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2543855855855856, + "grad_norm": 27.372623443603516, + "learning_rate": 8.869265194347394e-06, + "loss": 7.253933429718018, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.31651476211845875, + "loss_total_v6": 0.4533708468079567, + "step": 1103 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2546162162162162, + "grad_norm": 23.60000228881836, + "learning_rate": 8.866858643848664e-06, + "loss": 6.826461315155029, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.2911863140761852, + "loss_total_v6": 0.42665383592247963, + "step": 1104 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.25484684684684683, + "grad_norm": 21.992883682250977, + "learning_rate": 8.86444986236334e-06, + "loss": 7.042566299438477, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3035637252032757, + "loss_total_v6": 0.440160408616066, + "step": 1105 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2550774774774775, + "grad_norm": 23.21491241455078, + "learning_rate": 8.862038851281171e-06, + "loss": 6.741860866546631, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.28705843538045883, + "loss_total_v6": 0.42136630043387413, + "step": 1106 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2553081081081081, + "grad_norm": 23.329092025756836, + "learning_rate": 8.859625611993189e-06, + "loss": 7.327019691467285, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3237071596086025, + "loss_total_v6": 0.4579387307167053, + "step": 1107 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2555387387387387, + "grad_norm": 22.13730239868164, + "learning_rate": 8.857210145891714e-06, + "loss": 6.898428916931152, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.2980646472424269, + "loss_total_v6": 0.4311517998576164, + "step": 1108 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2557693693693694, + "grad_norm": 21.379175186157227, + "learning_rate": 8.854792454370354e-06, + "loss": 7.211792945861816, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.31813820078969, + "loss_total_v6": 0.4507370740175247, + "step": 1109 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.256, + "grad_norm": 24.130908966064453, + "learning_rate": 8.852372538823992e-06, + "loss": 6.786557674407959, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.28811250254511833, + "loss_total_v6": 0.42415985465049744, + "step": 1110 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2562306306306306, + "grad_norm": 23.886011123657227, + "learning_rate": 8.849950400648807e-06, + "loss": 6.988512992858887, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.303054042160511, + "loss_total_v6": 0.4367820620536804, + "step": 1111 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2564612612612613, + "grad_norm": 25.019819259643555, + "learning_rate": 8.847526041242247e-06, + "loss": 6.724538803100586, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.2847398463636637, + "loss_total_v6": 0.4202836602926254, + "step": 1112 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2566918918918919, + "grad_norm": 22.32982635498047, + "learning_rate": 8.845099462003049e-06, + "loss": 6.896383762359619, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.29544965736567974, + "loss_total_v6": 0.4310239776968956, + "step": 1113 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2569225225225225, + "grad_norm": 25.840717315673828, + "learning_rate": 8.842670664331233e-06, + "loss": 6.966947555541992, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.29981412179768085, + "loss_total_v6": 0.4354342296719551, + "step": 1114 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2571531531531532, + "grad_norm": 22.088424682617188, + "learning_rate": 8.84023964962809e-06, + "loss": 7.108872413635254, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3060293663293123, + "loss_total_v6": 0.44430453702807426, + "step": 1115 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2573837837837838, + "grad_norm": 22.031103134155273, + "learning_rate": 8.8378064192962e-06, + "loss": 7.146750450134277, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.31025832518935204, + "loss_total_v6": 0.44667190313339233, + "step": 1116 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2576144144144144, + "grad_norm": 26.035547256469727, + "learning_rate": 8.835370974739417e-06, + "loss": 6.987924575805664, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.30184233002364635, + "loss_total_v6": 0.4367452934384346, + "step": 1117 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.257845045045045, + "grad_norm": 24.206565856933594, + "learning_rate": 8.832933317362869e-06, + "loss": 7.139947414398193, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3129459507763386, + "loss_total_v6": 0.4462467208504677, + "step": 1118 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2580756756756757, + "grad_norm": 25.044282913208008, + "learning_rate": 8.830493448572964e-06, + "loss": 7.404094696044922, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.329256784170866, + "loss_total_v6": 0.4627559222280979, + "step": 1119 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2583063063063063, + "grad_norm": 24.254968643188477, + "learning_rate": 8.82805136977739e-06, + "loss": 7.498128414154053, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3328603319823742, + "loss_total_v6": 0.4686330407857895, + "step": 1120 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2585369369369369, + "grad_norm": 20.891979217529297, + "learning_rate": 8.825607082385101e-06, + "loss": 7.074794769287109, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3076226655393839, + "loss_total_v6": 0.44217466935515404, + "step": 1121 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2587675675675676, + "grad_norm": 23.78748321533203, + "learning_rate": 8.823160587806335e-06, + "loss": 7.303188323974609, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3225381299853325, + "loss_total_v6": 0.4564492553472519, + "step": 1122 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2589981981981982, + "grad_norm": 23.34002685546875, + "learning_rate": 8.820711887452596e-06, + "loss": 7.172015190124512, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.31517902202904224, + "loss_total_v6": 0.4482509307563305, + "step": 1123 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2592288288288288, + "grad_norm": 22.296680450439453, + "learning_rate": 8.818260982736662e-06, + "loss": 6.722251892089844, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.2824606988579035, + "loss_total_v6": 0.42014074698090553, + "step": 1124 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2594594594594595, + "grad_norm": 22.54844093322754, + "learning_rate": 8.815807875072586e-06, + "loss": 7.057109832763672, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3048846907913685, + "loss_total_v6": 0.4410693719983101, + "step": 1125 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2596900900900901, + "grad_norm": 23.217220306396484, + "learning_rate": 8.813352565875688e-06, + "loss": 7.274051666259766, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.31765011325478554, + "loss_total_v6": 0.45462824776768684, + "step": 1126 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2599207207207207, + "grad_norm": 26.145143508911133, + "learning_rate": 8.810895056562563e-06, + "loss": 6.95673942565918, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3005646727979183, + "loss_total_v6": 0.4347962364554405, + "step": 1127 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26015135135135137, + "grad_norm": 23.397022247314453, + "learning_rate": 8.808435348551072e-06, + "loss": 7.2249908447265625, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.31528565660119057, + "loss_total_v6": 0.45156191289424896, + "step": 1128 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.260381981981982, + "grad_norm": 25.853586196899414, + "learning_rate": 8.805973443260345e-06, + "loss": 6.738569259643555, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.28424347937107086, + "loss_total_v6": 0.42116058617830276, + "step": 1129 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2606126126126126, + "grad_norm": 22.1503849029541, + "learning_rate": 8.803509342110779e-06, + "loss": 6.960506916046143, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.2992131859064102, + "loss_total_v6": 0.4350316822528839, + "step": 1130 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26084324324324326, + "grad_norm": 22.658447265625, + "learning_rate": 8.801043046524039e-06, + "loss": 7.348489761352539, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.32311118580400944, + "loss_total_v6": 0.4592806175351143, + "step": 1131 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2610738738738739, + "grad_norm": 26.699851989746094, + "learning_rate": 8.798574557923055e-06, + "loss": 7.0860090255737305, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.30751486867666245, + "loss_total_v6": 0.44287557899951935, + "step": 1132 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2613045045045045, + "grad_norm": 23.377288818359375, + "learning_rate": 8.796103877732024e-06, + "loss": 7.100391387939453, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3067963160574436, + "loss_total_v6": 0.4437744729220867, + "step": 1133 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26153513513513515, + "grad_norm": 24.16975975036621, + "learning_rate": 8.793631007376407e-06, + "loss": 7.6637282371521, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3418370224535465, + "loss_total_v6": 0.4789830148220062, + "step": 1134 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26176576576576577, + "grad_norm": 23.984127044677734, + "learning_rate": 8.791155948282927e-06, + "loss": 7.099698543548584, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.3070582114160061, + "loss_total_v6": 0.4437311701476574, + "step": 1135 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2619963963963964, + "grad_norm": 26.945093154907227, + "learning_rate": 8.788678701879573e-06, + "loss": 7.519684791564941, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.33555038273334503, + "loss_total_v6": 0.46998030692338943, + "step": 1136 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26222702702702705, + "grad_norm": 24.50301742553711, + "learning_rate": 8.786199269595593e-06, + "loss": 7.069522380828857, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3047907017171383, + "loss_total_v6": 0.4418451376259327, + "step": 1137 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26245765765765766, + "grad_norm": 25.60744285583496, + "learning_rate": 8.783717652861495e-06, + "loss": 6.67482852935791, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.2827468551695347, + "loss_total_v6": 0.4171767756342888, + "step": 1138 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2626882882882883, + "grad_norm": 25.035247802734375, + "learning_rate": 8.781233853109052e-06, + "loss": 6.910348892211914, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.2979399058967829, + "loss_total_v6": 0.4318968169391155, + "step": 1139 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26291891891891894, + "grad_norm": 23.05251121520996, + "learning_rate": 8.778747871771293e-06, + "loss": 7.5138959884643555, + "loss_alignment": 0.280029296875, + "loss_alignment_w": 0.1400146484375, + "loss_sft": 0.329542838037014, + "loss_total_v6": 0.4696185179054737, + "step": 1140 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26314954954954956, + "grad_norm": 22.919635772705078, + "learning_rate": 8.776259710282504e-06, + "loss": 7.583380222320557, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.3358539715409279, + "loss_total_v6": 0.4739612601697445, + "step": 1141 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26338018018018017, + "grad_norm": 24.39705467224121, + "learning_rate": 8.773769370078238e-06, + "loss": 7.440708160400391, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3298361226916313, + "loss_total_v6": 0.4650442339479923, + "step": 1142 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26361081081081084, + "grad_norm": 23.685455322265625, + "learning_rate": 8.77127685259529e-06, + "loss": 7.406929969787598, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.325695613399148, + "loss_total_v6": 0.46293314173817635, + "step": 1143 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26384144144144145, + "grad_norm": 23.41263198852539, + "learning_rate": 8.768782159271727e-06, + "loss": 7.147839069366455, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.31232528015971184, + "loss_total_v6": 0.44673994556069374, + "step": 1144 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26407207207207206, + "grad_norm": 27.270889282226562, + "learning_rate": 8.766285291546858e-06, + "loss": 7.246458053588867, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3174208328127861, + "loss_total_v6": 0.4529036320745945, + "step": 1145 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26430270270270273, + "grad_norm": 25.98453140258789, + "learning_rate": 8.763786250861258e-06, + "loss": 7.2208051681518555, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.31473416090011597, + "loss_total_v6": 0.45130031555891037, + "step": 1146 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26453333333333334, + "grad_norm": 22.657512664794922, + "learning_rate": 8.761285038656747e-06, + "loss": 7.233180999755859, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.31643843837082386, + "loss_total_v6": 0.4520738199353218, + "step": 1147 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26476396396396396, + "grad_norm": 24.668317794799805, + "learning_rate": 8.758781656376398e-06, + "loss": 7.6011199951171875, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3407163806259632, + "loss_total_v6": 0.4750700034201145, + "step": 1148 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26499459459459457, + "grad_norm": 26.67666244506836, + "learning_rate": 8.756276105464545e-06, + "loss": 6.908168315887451, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.2990700900554657, + "loss_total_v6": 0.4317605309188366, + "step": 1149 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26522522522522524, + "grad_norm": 25.887744903564453, + "learning_rate": 8.753768387366764e-06, + "loss": 7.309537887573242, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.3191050235182047, + "loss_total_v6": 0.45684611797332764, + "step": 1150 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26545585585585585, + "grad_norm": 25.68214988708496, + "learning_rate": 8.751258503529885e-06, + "loss": 7.398448944091797, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3274237923324108, + "loss_total_v6": 0.462403055280447, + "step": 1151 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26568648648648646, + "grad_norm": 23.16781234741211, + "learning_rate": 8.748746455401986e-06, + "loss": 7.47852897644043, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.33108603581786156, + "loss_total_v6": 0.46740804240107536, + "step": 1152 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26591711711711713, + "grad_norm": 24.567136764526367, + "learning_rate": 8.746232244432397e-06, + "loss": 7.136324882507324, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.30753153935074806, + "loss_total_v6": 0.44602030143141747, + "step": 1153 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26614774774774774, + "grad_norm": 23.97275733947754, + "learning_rate": 8.743715872071694e-06, + "loss": 6.7469682693481445, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.2843106538057327, + "loss_total_v6": 0.42168551683425903, + "step": 1154 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26637837837837836, + "grad_norm": 23.646881103515625, + "learning_rate": 8.741197339771697e-06, + "loss": 7.553345680236816, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.3333206884562969, + "loss_total_v6": 0.4720841087400913, + "step": 1155 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.266609009009009, + "grad_norm": 23.892181396484375, + "learning_rate": 8.738676648985476e-06, + "loss": 6.8387908935546875, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.29090406000614166, + "loss_total_v6": 0.42742444947361946, + "step": 1156 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26683963963963964, + "grad_norm": 24.801645278930664, + "learning_rate": 8.736153801167346e-06, + "loss": 7.098830223083496, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3052491396665573, + "loss_total_v6": 0.4436768740415573, + "step": 1157 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26707027027027025, + "grad_norm": 23.42781639099121, + "learning_rate": 8.733628797772864e-06, + "loss": 7.2985944747924805, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3202368374913931, + "loss_total_v6": 0.45616213977336884, + "step": 1158 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2673009009009009, + "grad_norm": 22.763885498046875, + "learning_rate": 8.731101640258835e-06, + "loss": 7.498316764831543, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.33140723779797554, + "loss_total_v6": 0.46864478662610054, + "step": 1159 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26753153153153153, + "grad_norm": 24.753097534179688, + "learning_rate": 8.7285723300833e-06, + "loss": 7.07688045501709, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.30441136099398136, + "loss_total_v6": 0.4423050507903099, + "step": 1160 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26776216216216214, + "grad_norm": 28.303409576416016, + "learning_rate": 8.72604086870555e-06, + "loss": 7.479242324829102, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.33114588260650635, + "loss_total_v6": 0.46745265275239944, + "step": 1161 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2679927927927928, + "grad_norm": 25.348514556884766, + "learning_rate": 8.723507257586112e-06, + "loss": 7.8149847984313965, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.3518398627638817, + "loss_total_v6": 0.4884365499019623, + "step": 1162 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2682234234234234, + "grad_norm": 22.613887786865234, + "learning_rate": 8.720971498186754e-06, + "loss": 6.895544052124023, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.29516829177737236, + "loss_total_v6": 0.43097150698304176, + "step": 1163 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26845405405405404, + "grad_norm": 23.171398162841797, + "learning_rate": 8.718433591970486e-06, + "loss": 7.142897605895996, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.30955975130200386, + "loss_total_v6": 0.44643110036849976, + "step": 1164 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2686846846846847, + "grad_norm": 23.862911224365234, + "learning_rate": 8.715893540401552e-06, + "loss": 7.735307216644287, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.34504422172904015, + "loss_total_v6": 0.48345669358968735, + "step": 1165 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2689153153153153, + "grad_norm": 22.291942596435547, + "learning_rate": 8.713351344945438e-06, + "loss": 7.3140435218811035, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3228961452841759, + "loss_total_v6": 0.4571277126669884, + "step": 1166 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.26914594594594593, + "grad_norm": 22.310914993286133, + "learning_rate": 8.710807007068865e-06, + "loss": 7.055513858795166, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.30525791831314564, + "loss_total_v6": 0.440969604998827, + "step": 1167 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2693765765765766, + "grad_norm": 24.7689208984375, + "learning_rate": 8.708260528239788e-06, + "loss": 7.232020378112793, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.31903616711497307, + "loss_total_v6": 0.45200125873088837, + "step": 1168 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2696072072072072, + "grad_norm": 22.37305450439453, + "learning_rate": 8.705711909927404e-06, + "loss": 7.691858291625977, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.34234393015503883, + "loss_total_v6": 0.48074114322662354, + "step": 1169 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2698378378378378, + "grad_norm": 27.88308334350586, + "learning_rate": 8.703161153602137e-06, + "loss": 7.349720001220703, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.32450034469366074, + "loss_total_v6": 0.45935749635100365, + "step": 1170 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2700684684684685, + "grad_norm": 23.212970733642578, + "learning_rate": 8.700608260735648e-06, + "loss": 6.948514461517334, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.2970293499529362, + "loss_total_v6": 0.4342821575701237, + "step": 1171 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2702990990990991, + "grad_norm": 25.884294509887695, + "learning_rate": 8.698053232800831e-06, + "loss": 7.428017616271973, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.3261285126209259, + "loss_total_v6": 0.4642510861158371, + "step": 1172 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2705297297297297, + "grad_norm": 25.31771469116211, + "learning_rate": 8.695496071271811e-06, + "loss": 7.217898368835449, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3158647492527962, + "loss_total_v6": 0.45111866295337677, + "step": 1173 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2707603603603604, + "grad_norm": 25.153379440307617, + "learning_rate": 8.692936777623943e-06, + "loss": 6.444717884063721, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.2695093620568514, + "loss_total_v6": 0.40279487520456314, + "step": 1174 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.270990990990991, + "grad_norm": 23.406286239624023, + "learning_rate": 8.690375353333815e-06, + "loss": 7.601413249969482, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.33634013310074806, + "loss_total_v6": 0.47508831694722176, + "step": 1175 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2712216216216216, + "grad_norm": 25.537708282470703, + "learning_rate": 8.68781179987924e-06, + "loss": 7.4664812088012695, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.33149274811148643, + "loss_total_v6": 0.46665509045124054, + "step": 1176 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2714522522522523, + "grad_norm": 23.647558212280273, + "learning_rate": 8.685246118739264e-06, + "loss": 7.48605489730835, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.33102235198020935, + "loss_total_v6": 0.46787841990590096, + "step": 1177 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2716828828828829, + "grad_norm": 23.97349739074707, + "learning_rate": 8.682678311394155e-06, + "loss": 6.800876140594482, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.29121991619467735, + "loss_total_v6": 0.42505474761128426, + "step": 1178 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2719135135135135, + "grad_norm": 22.369476318359375, + "learning_rate": 8.680108379325413e-06, + "loss": 7.165534973144531, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.31518600694835186, + "loss_total_v6": 0.447845920920372, + "step": 1179 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2721441441441441, + "grad_norm": 27.230710983276367, + "learning_rate": 8.677536324015765e-06, + "loss": 7.170548439025879, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.31205084174871445, + "loss_total_v6": 0.44815924391150475, + "step": 1180 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2723747747747748, + "grad_norm": 24.620481491088867, + "learning_rate": 8.674962146949158e-06, + "loss": 7.033852577209473, + "loss_alignment": 0.2664794921875, + "loss_alignment_w": 0.13323974609375, + "loss_sft": 0.306498097255826, + "loss_total_v6": 0.43961578235030174, + "step": 1181 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2726054054054054, + "grad_norm": 25.444570541381836, + "learning_rate": 8.67238584961076e-06, + "loss": 7.17767858505249, + "loss_alignment": 0.279052734375, + "loss_alignment_w": 0.1395263671875, + "loss_sft": 0.3089259546250105, + "loss_total_v6": 0.44860490038990974, + "step": 1182 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.272836036036036, + "grad_norm": 26.22624397277832, + "learning_rate": 8.669807433486977e-06, + "loss": 7.52066707611084, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.3333382047712803, + "loss_total_v6": 0.4700416997075081, + "step": 1183 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2730666666666667, + "grad_norm": 26.447355270385742, + "learning_rate": 8.66722690006542e-06, + "loss": 7.367136001586914, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.32462751865386963, + "loss_total_v6": 0.4604460038244724, + "step": 1184 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2732972972972973, + "grad_norm": 21.89260482788086, + "learning_rate": 8.664644250834932e-06, + "loss": 7.015378952026367, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.3016966637223959, + "loss_total_v6": 0.43846118077635765, + "step": 1185 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2735279279279279, + "grad_norm": 21.755146026611328, + "learning_rate": 8.662059487285575e-06, + "loss": 7.223238945007324, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3156796954572201, + "loss_total_v6": 0.45145241916179657, + "step": 1186 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2737585585585586, + "grad_norm": 24.528942108154297, + "learning_rate": 8.659472610908628e-06, + "loss": 6.607970237731934, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.2798041757196188, + "loss_total_v6": 0.41299814730882645, + "step": 1187 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2739891891891892, + "grad_norm": 25.983911514282227, + "learning_rate": 8.656883623196592e-06, + "loss": 6.907777786254883, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.2979775369167328, + "loss_total_v6": 0.4317360818386078, + "step": 1188 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2742198198198198, + "grad_norm": 23.545080184936523, + "learning_rate": 8.654292525643188e-06, + "loss": 7.149837493896484, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.30855919420719147, + "loss_total_v6": 0.4468648433685303, + "step": 1189 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27445045045045047, + "grad_norm": 22.108449935913086, + "learning_rate": 8.651699319743348e-06, + "loss": 7.072606086730957, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.3115904778242111, + "loss_total_v6": 0.4420378729701042, + "step": 1190 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2746810810810811, + "grad_norm": 22.56694221496582, + "learning_rate": 8.649104006993223e-06, + "loss": 7.091485500335693, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.30788762867450714, + "loss_total_v6": 0.44321783259510994, + "step": 1191 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2749117117117117, + "grad_norm": 22.54511070251465, + "learning_rate": 8.646506588890184e-06, + "loss": 7.581777572631836, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.34121648594737053, + "loss_total_v6": 0.47386112064123154, + "step": 1192 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27514234234234236, + "grad_norm": 22.24480438232422, + "learning_rate": 8.64390706693281e-06, + "loss": 7.25906229019165, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.31591977924108505, + "loss_total_v6": 0.45369138568639755, + "step": 1193 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.275372972972973, + "grad_norm": 23.764772415161133, + "learning_rate": 8.641305442620897e-06, + "loss": 7.293567657470703, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.32025837525725365, + "loss_total_v6": 0.45584797114133835, + "step": 1194 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2756036036036036, + "grad_norm": 23.373144149780273, + "learning_rate": 8.638701717455457e-06, + "loss": 7.064990043640137, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.30714721977710724, + "loss_total_v6": 0.44156188517808914, + "step": 1195 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27583423423423425, + "grad_norm": 23.35517120361328, + "learning_rate": 8.636095892938707e-06, + "loss": 6.876317501068115, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.29733880050480366, + "loss_total_v6": 0.4297698326408863, + "step": 1196 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27606486486486487, + "grad_norm": 20.349145889282227, + "learning_rate": 8.633487970574082e-06, + "loss": 7.290887355804443, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3199688009917736, + "loss_total_v6": 0.4556804709136486, + "step": 1197 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2762954954954955, + "grad_norm": 24.152599334716797, + "learning_rate": 8.630877951866223e-06, + "loss": 6.856235504150391, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.2952291965484619, + "loss_total_v6": 0.4285147227346897, + "step": 1198 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27652612612612615, + "grad_norm": 24.984249114990234, + "learning_rate": 8.628265838320982e-06, + "loss": 7.829514980316162, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.3512832075357437, + "loss_total_v6": 0.4893447235226631, + "step": 1199 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27675675675675676, + "grad_norm": 24.443138122558594, + "learning_rate": 8.625651631445419e-06, + "loss": 7.113617897033691, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.3079586811363697, + "loss_total_v6": 0.4446011185646057, + "step": 1200 + }, + { + "epoch": 0.27675675675675676, + "eval_loss": 0.5984920263290405, + "eval_loss_alignment": 0.2440336044520548, + "eval_loss_alignment_w": 0.1220168022260274, + "eval_loss_sft": 0.47647517809822953, + "eval_loss_total_v6": 0.5984919801456471, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.3205, + "eval_samples_per_second": 19.109, + "eval_steps_per_second": 2.398, + "step": 1200 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2769873873873874, + "grad_norm": 22.6988582611084, + "learning_rate": 8.623035332747804e-06, + "loss": 7.118947505950928, + "loss_alignment": 0.2802734375, + "loss_alignment_w": 0.14013671875, + "loss_sft": 0.3049958571791649, + "loss_total_v6": 0.4449342153966427, + "step": 1201 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27721801801801804, + "grad_norm": 22.997800827026367, + "learning_rate": 8.620416943737613e-06, + "loss": 7.052372932434082, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.30635861307382584, + "loss_total_v6": 0.44077328965067863, + "step": 1202 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27744864864864865, + "grad_norm": 22.757421493530273, + "learning_rate": 8.617796465925522e-06, + "loss": 7.377205848693848, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3244786523282528, + "loss_total_v6": 0.4610753506422043, + "step": 1203 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27767927927927927, + "grad_norm": 22.12559700012207, + "learning_rate": 8.615173900823425e-06, + "loss": 7.361842155456543, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.32373206317424774, + "loss_total_v6": 0.46011511981487274, + "step": 1204 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27790990990990994, + "grad_norm": 22.721269607543945, + "learning_rate": 8.612549249944408e-06, + "loss": 7.232479095458984, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3171270228922367, + "loss_total_v6": 0.4520299658179283, + "step": 1205 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27814054054054055, + "grad_norm": 23.641998291015625, + "learning_rate": 8.609922514802765e-06, + "loss": 7.011031150817871, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.301958991214633, + "loss_total_v6": 0.43818945437669754, + "step": 1206 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27837117117117116, + "grad_norm": 25.269594192504883, + "learning_rate": 8.607293696913993e-06, + "loss": 7.528624057769775, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.33365242183208466, + "loss_total_v6": 0.47053901106119156, + "step": 1207 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27860180180180183, + "grad_norm": 24.200767517089844, + "learning_rate": 8.60466279779479e-06, + "loss": 7.118448257446289, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.30696357041597366, + "loss_total_v6": 0.44490302726626396, + "step": 1208 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27883243243243244, + "grad_norm": 26.02063751220703, + "learning_rate": 8.602029818963057e-06, + "loss": 6.916673183441162, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.29595478996634483, + "loss_total_v6": 0.43229207023978233, + "step": 1209 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27906306306306305, + "grad_norm": 23.712461471557617, + "learning_rate": 8.59939476193789e-06, + "loss": 7.429290771484375, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.32640649005770683, + "loss_total_v6": 0.46433068811893463, + "step": 1210 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27929369369369367, + "grad_norm": 21.02884864807129, + "learning_rate": 8.596757628239588e-06, + "loss": 7.223376274108887, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.31416241079568863, + "loss_total_v6": 0.45146098360419273, + "step": 1211 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27952432432432434, + "grad_norm": 23.313899993896484, + "learning_rate": 8.594118419389648e-06, + "loss": 6.7919111251831055, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.28879804350435734, + "loss_total_v6": 0.424494456499815, + "step": 1212 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27975495495495495, + "grad_norm": 28.309995651245117, + "learning_rate": 8.591477136910763e-06, + "loss": 7.4184770584106445, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3286603018641472, + "loss_total_v6": 0.4636548161506653, + "step": 1213 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.27998558558558556, + "grad_norm": 24.03061866760254, + "learning_rate": 8.588833782326818e-06, + "loss": 7.733516693115234, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.34891486167907715, + "loss_total_v6": 0.48334477096796036, + "step": 1214 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28021621621621623, + "grad_norm": 24.22528839111328, + "learning_rate": 8.586188357162906e-06, + "loss": 7.4601850509643555, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.330961886793375, + "loss_total_v6": 0.4662615619599819, + "step": 1215 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28044684684684684, + "grad_norm": 31.822113037109375, + "learning_rate": 8.583540862945301e-06, + "loss": 7.282430171966553, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3195775467902422, + "loss_total_v6": 0.45515188574790955, + "step": 1216 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28067747747747746, + "grad_norm": 23.009536743164062, + "learning_rate": 8.580891301201479e-06, + "loss": 7.388354301452637, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3298446498811245, + "loss_total_v6": 0.4617721326649189, + "step": 1217 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2809081081081081, + "grad_norm": 22.527816772460938, + "learning_rate": 8.578239673460106e-06, + "loss": 7.389993190765381, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.3215852752327919, + "loss_total_v6": 0.4618745744228363, + "step": 1218 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28113873873873874, + "grad_norm": 26.94561004638672, + "learning_rate": 8.57558598125104e-06, + "loss": 7.214770793914795, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.3138687200844288, + "loss_total_v6": 0.450923178344965, + "step": 1219 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28136936936936935, + "grad_norm": 23.677536010742188, + "learning_rate": 8.572930226105331e-06, + "loss": 7.017027854919434, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.3010367937386036, + "loss_total_v6": 0.4385642446577549, + "step": 1220 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2816, + "grad_norm": 21.02745246887207, + "learning_rate": 8.570272409555217e-06, + "loss": 7.095613479614258, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3094883859157562, + "loss_total_v6": 0.4434758313000202, + "step": 1221 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28183063063063063, + "grad_norm": 22.752717971801758, + "learning_rate": 8.567612533134128e-06, + "loss": 6.90347146987915, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.2949618510901928, + "loss_total_v6": 0.4314669668674469, + "step": 1222 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28206126126126124, + "grad_norm": 23.708093643188477, + "learning_rate": 8.564950598376683e-06, + "loss": 6.956885814666748, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.30086373537778854, + "loss_total_v6": 0.43480536714196205, + "step": 1223 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2822918918918919, + "grad_norm": 23.749324798583984, + "learning_rate": 8.562286606818684e-06, + "loss": 7.424626350402832, + "loss_alignment": 0.2777099609375, + "loss_alignment_w": 0.13885498046875, + "loss_sft": 0.3254435807466507, + "loss_total_v6": 0.4640391692519188, + "step": 1224 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2825225225225225, + "grad_norm": 24.89012908935547, + "learning_rate": 8.559620559997125e-06, + "loss": 7.089531898498535, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.30889469385147095, + "loss_total_v6": 0.44309573248028755, + "step": 1225 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28275315315315314, + "grad_norm": 25.805932998657227, + "learning_rate": 8.556952459450183e-06, + "loss": 7.192452430725098, + "loss_alignment": 0.28076171875, + "loss_alignment_w": 0.140380859375, + "loss_sft": 0.3086896352469921, + "loss_total_v6": 0.4495282582938671, + "step": 1226 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2829837837837838, + "grad_norm": 23.522981643676758, + "learning_rate": 8.554282306717218e-06, + "loss": 7.483460426330566, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3329811729490757, + "loss_total_v6": 0.4677162878215313, + "step": 1227 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2832144144144144, + "grad_norm": 24.068559646606445, + "learning_rate": 8.55161010333878e-06, + "loss": 7.037247657775879, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.3023310601711273, + "loss_total_v6": 0.4398279935121536, + "step": 1228 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28344504504504503, + "grad_norm": 22.592220306396484, + "learning_rate": 8.548935850856597e-06, + "loss": 7.210383415222168, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3147999681532383, + "loss_total_v6": 0.4506489783525467, + "step": 1229 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2836756756756757, + "grad_norm": 22.530363082885742, + "learning_rate": 8.54625955081358e-06, + "loss": 7.3540239334106445, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.326157845556736, + "loss_total_v6": 0.4596264809370041, + "step": 1230 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2839063063063063, + "grad_norm": 22.540565490722656, + "learning_rate": 8.543581204753822e-06, + "loss": 6.811432361602783, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2930240910500288, + "loss_total_v6": 0.42571451887488365, + "step": 1231 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2841369369369369, + "grad_norm": 22.442787170410156, + "learning_rate": 8.540900814222597e-06, + "loss": 7.315891265869141, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.3190443553030491, + "loss_total_v6": 0.4572432190179825, + "step": 1232 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2843675675675676, + "grad_norm": 22.990156173706055, + "learning_rate": 8.538218380766358e-06, + "loss": 7.394705772399902, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3268694132566452, + "loss_total_v6": 0.4621690958738327, + "step": 1233 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2845981981981982, + "grad_norm": 24.943437576293945, + "learning_rate": 8.535533905932739e-06, + "loss": 7.033618927001953, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3054153807461262, + "loss_total_v6": 0.43960117921233177, + "step": 1234 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2848288288288288, + "grad_norm": 26.403688430786133, + "learning_rate": 8.532847391270545e-06, + "loss": 7.032988548278809, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.30554383993148804, + "loss_total_v6": 0.43956178799271584, + "step": 1235 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2850594594594595, + "grad_norm": 23.432388305664062, + "learning_rate": 8.530158838329765e-06, + "loss": 7.746272087097168, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.347316462546587, + "loss_total_v6": 0.4841420240700245, + "step": 1236 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2852900900900901, + "grad_norm": 24.350269317626953, + "learning_rate": 8.527468248661562e-06, + "loss": 6.839145660400391, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.29255889542400837, + "loss_total_v6": 0.4274466037750244, + "step": 1237 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2855207207207207, + "grad_norm": 25.66450309753418, + "learning_rate": 8.52477562381827e-06, + "loss": 6.877553939819336, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.2938455380499363, + "loss_total_v6": 0.4298471137881279, + "step": 1238 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2857513513513513, + "grad_norm": 24.626794815063477, + "learning_rate": 8.522080965353405e-06, + "loss": 7.065519332885742, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3082941845059395, + "loss_total_v6": 0.4415949583053589, + "step": 1239 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.285981981981982, + "grad_norm": 24.740631103515625, + "learning_rate": 8.519384274821649e-06, + "loss": 7.083189964294434, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.30393592827022076, + "loss_total_v6": 0.4426993615925312, + "step": 1240 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2862126126126126, + "grad_norm": 25.867273330688477, + "learning_rate": 8.51668555377886e-06, + "loss": 7.040142059326172, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3039157222956419, + "loss_total_v6": 0.44000887498259544, + "step": 1241 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2864432432432432, + "grad_norm": 25.483257293701172, + "learning_rate": 8.513984803782069e-06, + "loss": 7.106750011444092, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.30643079429864883, + "loss_total_v6": 0.44417187571525574, + "step": 1242 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2866738738738739, + "grad_norm": 24.409143447875977, + "learning_rate": 8.511282026389474e-06, + "loss": 7.089870452880859, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3078324720263481, + "loss_total_v6": 0.4431168958544731, + "step": 1243 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2869045045045045, + "grad_norm": 23.5096492767334, + "learning_rate": 8.508577223160441e-06, + "loss": 6.930075645446777, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.2977689988911152, + "loss_total_v6": 0.4331297129392624, + "step": 1244 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2871351351351351, + "grad_norm": 20.753517150878906, + "learning_rate": 8.505870395655512e-06, + "loss": 6.443118095397949, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2713930103927851, + "loss_total_v6": 0.4026948846876621, + "step": 1245 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2873657657657658, + "grad_norm": 24.435964584350586, + "learning_rate": 8.503161545436394e-06, + "loss": 6.901890754699707, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.2973502352833748, + "loss_total_v6": 0.431368175894022, + "step": 1246 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2875963963963964, + "grad_norm": 25.519227981567383, + "learning_rate": 8.500450674065954e-06, + "loss": 7.040066242218018, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3051164448261261, + "loss_total_v6": 0.4400041438639164, + "step": 1247 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.287827027027027, + "grad_norm": 27.345497131347656, + "learning_rate": 8.497737783108238e-06, + "loss": 7.170576572418213, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3121289573609829, + "loss_total_v6": 0.4481610506772995, + "step": 1248 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2880576576576577, + "grad_norm": 24.30404281616211, + "learning_rate": 8.495022874128445e-06, + "loss": 6.8225908279418945, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.2898915484547615, + "loss_total_v6": 0.426411934196949, + "step": 1249 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2882882882882883, + "grad_norm": 27.262224197387695, + "learning_rate": 8.492305948692944e-06, + "loss": 7.295602798461914, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.31934795528650284, + "loss_total_v6": 0.45597516000270844, + "step": 1250 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2885189189189189, + "grad_norm": 27.80640983581543, + "learning_rate": 8.489587008369269e-06, + "loss": 7.388357639312744, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.324565339833498, + "loss_total_v6": 0.4617723636329174, + "step": 1251 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28874954954954957, + "grad_norm": 23.639816284179688, + "learning_rate": 8.486866054726115e-06, + "loss": 6.992911338806152, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.3017725460231304, + "loss_total_v6": 0.4370569810271263, + "step": 1252 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2889801801801802, + "grad_norm": 23.865245819091797, + "learning_rate": 8.484143089333334e-06, + "loss": 7.201281547546387, + "loss_alignment": 0.2713623046875, + "loss_alignment_w": 0.13568115234375, + "loss_sft": 0.31413955241441727, + "loss_total_v6": 0.45008009299635887, + "step": 1253 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2892108108108108, + "grad_norm": 24.322208404541016, + "learning_rate": 8.481418113761947e-06, + "loss": 6.740408897399902, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.2869066894054413, + "loss_total_v6": 0.4212755709886551, + "step": 1254 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.28944144144144146, + "grad_norm": 23.357940673828125, + "learning_rate": 8.478691129584128e-06, + "loss": 7.374388694763184, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.3249129615724087, + "loss_total_v6": 0.4608992859721184, + "step": 1255 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2896720720720721, + "grad_norm": 22.71914291381836, + "learning_rate": 8.475962138373212e-06, + "loss": 6.988134384155273, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.29996336437761784, + "loss_total_v6": 0.4367583990097046, + "step": 1256 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2899027027027027, + "grad_norm": 26.507535934448242, + "learning_rate": 8.473231141703694e-06, + "loss": 7.458277702331543, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.33136147633194923, + "loss_total_v6": 0.46614236012101173, + "step": 1257 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29013333333333335, + "grad_norm": 25.341880798339844, + "learning_rate": 8.470498141151223e-06, + "loss": 6.976737976074219, + "loss_alignment": 0.277099609375, + "loss_alignment_w": 0.1385498046875, + "loss_sft": 0.29767940752208233, + "loss_total_v6": 0.43604612350463867, + "step": 1258 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29036396396396397, + "grad_norm": 24.06606101989746, + "learning_rate": 8.467763138292605e-06, + "loss": 6.430680274963379, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.2731180712580681, + "loss_total_v6": 0.4019175209105015, + "step": 1259 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2905945945945946, + "grad_norm": 22.39508056640625, + "learning_rate": 8.465026134705805e-06, + "loss": 7.405568599700928, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.326754916459322, + "loss_total_v6": 0.4628480523824692, + "step": 1260 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29082522522522525, + "grad_norm": 23.96640396118164, + "learning_rate": 8.462287131969933e-06, + "loss": 7.298076152801514, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.321181021630764, + "loss_total_v6": 0.4561297632753849, + "step": 1261 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29105585585585586, + "grad_norm": 26.269269943237305, + "learning_rate": 8.459546131665265e-06, + "loss": 7.068235397338867, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3072432056069374, + "loss_total_v6": 0.4417647048830986, + "step": 1262 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2912864864864865, + "grad_norm": 24.62394142150879, + "learning_rate": 8.456803135373217e-06, + "loss": 7.147780418395996, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.31406112015247345, + "loss_total_v6": 0.44673628732562065, + "step": 1263 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29151711711711714, + "grad_norm": 25.71173095703125, + "learning_rate": 8.454058144676368e-06, + "loss": 7.218035697937012, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3144237343221903, + "loss_total_v6": 0.45112723112106323, + "step": 1264 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29174774774774775, + "grad_norm": 25.724573135375977, + "learning_rate": 8.451311161158437e-06, + "loss": 6.649350166320801, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.28275661543011665, + "loss_total_v6": 0.41558437421917915, + "step": 1265 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29197837837837837, + "grad_norm": 23.456241607666016, + "learning_rate": 8.448562186404301e-06, + "loss": 6.708845138549805, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.2845371961593628, + "loss_total_v6": 0.4193028248846531, + "step": 1266 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29220900900900904, + "grad_norm": 25.657743453979492, + "learning_rate": 8.445811221999983e-06, + "loss": 6.897252082824707, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.2986472360789776, + "loss_total_v6": 0.4310782700777054, + "step": 1267 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29243963963963965, + "grad_norm": 24.45419692993164, + "learning_rate": 8.44305826953265e-06, + "loss": 7.1155314445495605, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.309558380395174, + "loss_total_v6": 0.44472073018550873, + "step": 1268 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29267027027027026, + "grad_norm": 24.032106399536133, + "learning_rate": 8.440303330590626e-06, + "loss": 7.001578330993652, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3031687494367361, + "loss_total_v6": 0.43759866431355476, + "step": 1269 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2929009009009009, + "grad_norm": 22.850479125976562, + "learning_rate": 8.43754640676337e-06, + "loss": 7.096916198730469, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.3075251542031765, + "loss_total_v6": 0.4435572661459446, + "step": 1270 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29313153153153154, + "grad_norm": 24.065610885620117, + "learning_rate": 8.434787499641492e-06, + "loss": 7.308736324310303, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.3232053145766258, + "loss_total_v6": 0.4567960128188133, + "step": 1271 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29336216216216215, + "grad_norm": 26.73378562927246, + "learning_rate": 8.432026610816745e-06, + "loss": 7.164787292480469, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.31283524818718433, + "loss_total_v6": 0.4477992318570614, + "step": 1272 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29359279279279277, + "grad_norm": 23.07343292236328, + "learning_rate": 8.429263741882026e-06, + "loss": 7.046832084655762, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.3073550797998905, + "loss_total_v6": 0.4404270015656948, + "step": 1273 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29382342342342344, + "grad_norm": 24.534982681274414, + "learning_rate": 8.426498894431375e-06, + "loss": 7.353806018829346, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3255033865571022, + "loss_total_v6": 0.4596128650009632, + "step": 1274 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29405405405405405, + "grad_norm": 23.49444580078125, + "learning_rate": 8.423732070059972e-06, + "loss": 7.266429901123047, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.31976768746972084, + "loss_total_v6": 0.45415184646844864, + "step": 1275 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29428468468468466, + "grad_norm": 22.71343421936035, + "learning_rate": 8.420963270364138e-06, + "loss": 6.641339302062988, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.2811268325895071, + "loss_total_v6": 0.41508372500538826, + "step": 1276 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29451531531531533, + "grad_norm": 23.828535079956055, + "learning_rate": 8.418192496941331e-06, + "loss": 7.270625591278076, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.31916020810604095, + "loss_total_v6": 0.45441409945487976, + "step": 1277 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29474594594594594, + "grad_norm": 24.943891525268555, + "learning_rate": 8.415419751390155e-06, + "loss": 6.8363752365112305, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.296856589615345, + "loss_total_v6": 0.4272734709084034, + "step": 1278 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29497657657657655, + "grad_norm": 22.279802322387695, + "learning_rate": 8.412645035310345e-06, + "loss": 6.748289108276367, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.28715504705905914, + "loss_total_v6": 0.42176808416843414, + "step": 1279 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2952072072072072, + "grad_norm": 23.071748733520508, + "learning_rate": 8.409868350302774e-06, + "loss": 6.7752485275268555, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.28732934780418873, + "loss_total_v6": 0.4234530180692673, + "step": 1280 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29543783783783784, + "grad_norm": 22.772136688232422, + "learning_rate": 8.407089697969458e-06, + "loss": 7.388060569763184, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3260726146399975, + "loss_total_v6": 0.4617537595331669, + "step": 1281 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29566846846846845, + "grad_norm": 25.076078414916992, + "learning_rate": 8.404309079913533e-06, + "loss": 7.114843845367432, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3132537826895714, + "loss_total_v6": 0.4446777403354645, + "step": 1282 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2958990990990991, + "grad_norm": 24.02098846435547, + "learning_rate": 8.401526497739288e-06, + "loss": 7.395047187805176, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3267229199409485, + "loss_total_v6": 0.4621904604136944, + "step": 1283 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29612972972972973, + "grad_norm": 25.416215896606445, + "learning_rate": 8.398741953052128e-06, + "loss": 7.076164245605469, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3100123666226864, + "loss_total_v6": 0.4422602839767933, + "step": 1284 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29636036036036034, + "grad_norm": 24.084138870239258, + "learning_rate": 8.395955447458601e-06, + "loss": 7.317990303039551, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.32038096711039543, + "loss_total_v6": 0.45737437531352043, + "step": 1285 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.296590990990991, + "grad_norm": 24.350738525390625, + "learning_rate": 8.393166982566384e-06, + "loss": 6.893924713134766, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.29581473022699356, + "loss_total_v6": 0.43087030202150345, + "step": 1286 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2968216216216216, + "grad_norm": 22.8326358795166, + "learning_rate": 8.390376559984282e-06, + "loss": 6.74772834777832, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.28805074840784073, + "loss_total_v6": 0.421733021736145, + "step": 1287 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29705225225225224, + "grad_norm": 26.636503219604492, + "learning_rate": 8.387584181322233e-06, + "loss": 6.775362968444824, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.2856122609227896, + "loss_total_v6": 0.4234601706266403, + "step": 1288 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2972828828828829, + "grad_norm": 26.45832061767578, + "learning_rate": 8.3847898481913e-06, + "loss": 7.053994655609131, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.30667364969849586, + "loss_total_v6": 0.4408746771514416, + "step": 1289 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2975135135135135, + "grad_norm": 27.000747680664062, + "learning_rate": 8.381993562203675e-06, + "loss": 7.237860679626465, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.32039302587509155, + "loss_total_v6": 0.45236629247665405, + "step": 1290 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.29774414414414413, + "grad_norm": 23.961193084716797, + "learning_rate": 8.37919532497268e-06, + "loss": 7.328529357910156, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.32363369688391685, + "loss_total_v6": 0.45803309604525566, + "step": 1291 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2979747747747748, + "grad_norm": 24.848201751708984, + "learning_rate": 8.376395138112755e-06, + "loss": 6.8563618659973145, + "loss_alignment": 0.2735595703125, + "loss_alignment_w": 0.13677978515625, + "loss_sft": 0.29134610295295715, + "loss_total_v6": 0.42852262407541275, + "step": 1292 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2982054054054054, + "grad_norm": 23.494359970092773, + "learning_rate": 8.373593003239474e-06, + "loss": 7.447225570678711, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.32946528121829033, + "loss_total_v6": 0.46545159444212914, + "step": 1293 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.298436036036036, + "grad_norm": 21.817829132080078, + "learning_rate": 8.370788921969529e-06, + "loss": 7.26419734954834, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.31813282147049904, + "loss_total_v6": 0.45401233434677124, + "step": 1294 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2986666666666667, + "grad_norm": 24.62862777709961, + "learning_rate": 8.367982895920734e-06, + "loss": 7.113400459289551, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3092115558683872, + "loss_total_v6": 0.4445875398814678, + "step": 1295 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2988972972972973, + "grad_norm": 21.543319702148438, + "learning_rate": 8.365174926712032e-06, + "loss": 7.539090156555176, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3343980945646763, + "loss_total_v6": 0.4711931422352791, + "step": 1296 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2991279279279279, + "grad_norm": 24.421306610107422, + "learning_rate": 8.362365015963478e-06, + "loss": 7.052982807159424, + "loss_alignment": 0.2689208984375, + "loss_alignment_w": 0.13446044921875, + "loss_sft": 0.3061831146478653, + "loss_total_v6": 0.4408114105463028, + "step": 1297 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2993585585585586, + "grad_norm": 24.339725494384766, + "learning_rate": 8.359553165296257e-06, + "loss": 7.408743858337402, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.32875388115644455, + "loss_total_v6": 0.46304647997021675, + "step": 1298 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2995891891891892, + "grad_norm": 23.636518478393555, + "learning_rate": 8.356739376332665e-06, + "loss": 7.1243391036987305, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.31413718312978745, + "loss_total_v6": 0.44527120888233185, + "step": 1299 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.2998198198198198, + "grad_norm": 22.442161560058594, + "learning_rate": 8.353923650696119e-06, + "loss": 7.541460037231445, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.33433255553245544, + "loss_total_v6": 0.47134123742580414, + "step": 1300 + }, + { + "epoch": 0.2998198198198198, + "eval_loss": 0.6026567220687866, + "eval_loss_alignment": 0.24321980236872145, + "eval_loss_alignment_w": 0.12160990118436073, + "eval_loss_sft": 0.4810468615819466, + "eval_loss_total_v6": 0.6026567624346034, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.4038, + "eval_samples_per_second": 19.091, + "eval_steps_per_second": 2.396, + "step": 1300 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3000504504504504, + "grad_norm": 21.2752685546875, + "learning_rate": 8.351105990011154e-06, + "loss": 6.87876033782959, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.29611818864941597, + "loss_total_v6": 0.42992250621318817, + "step": 1301 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3002810810810811, + "grad_norm": 26.83425521850586, + "learning_rate": 8.348286395903423e-06, + "loss": 7.403532981872559, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3263225071132183, + "loss_total_v6": 0.4627208150923252, + "step": 1302 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3005117117117117, + "grad_norm": 23.297460556030273, + "learning_rate": 8.345464869999688e-06, + "loss": 7.399998664855957, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3267424814403057, + "loss_total_v6": 0.4624999314546585, + "step": 1303 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3007423423423423, + "grad_norm": 24.97939109802246, + "learning_rate": 8.342641413927836e-06, + "loss": 7.039825439453125, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3101825639605522, + "loss_total_v6": 0.43998908624053, + "step": 1304 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.300972972972973, + "grad_norm": 23.943443298339844, + "learning_rate": 8.339816029316859e-06, + "loss": 7.400753021240234, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.32595038414001465, + "loss_total_v6": 0.46254705637693405, + "step": 1305 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3012036036036036, + "grad_norm": 23.54371452331543, + "learning_rate": 8.336988717796865e-06, + "loss": 7.297178745269775, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.31941595301032066, + "loss_total_v6": 0.45607366040349007, + "step": 1306 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3014342342342342, + "grad_norm": 26.094343185424805, + "learning_rate": 8.33415948099907e-06, + "loss": 7.060044288635254, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.30566318705677986, + "loss_total_v6": 0.44125277921557426, + "step": 1307 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3016648648648649, + "grad_norm": 25.63722801208496, + "learning_rate": 8.331328320555812e-06, + "loss": 6.775112152099609, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.290174275636673, + "loss_total_v6": 0.42344454303383827, + "step": 1308 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3018954954954955, + "grad_norm": 25.421939849853516, + "learning_rate": 8.328495238100522e-06, + "loss": 7.494784832000732, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.33466551825404167, + "loss_total_v6": 0.4684240557253361, + "step": 1309 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3021261261261261, + "grad_norm": 24.306381225585938, + "learning_rate": 8.325660235267753e-06, + "loss": 7.040044784545898, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.30401649698615074, + "loss_total_v6": 0.44000282138586044, + "step": 1310 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3023567567567568, + "grad_norm": 24.52485466003418, + "learning_rate": 8.322823313693162e-06, + "loss": 7.046346664428711, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.30624140053987503, + "loss_total_v6": 0.44039666280150414, + "step": 1311 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3025873873873874, + "grad_norm": 23.13258171081543, + "learning_rate": 8.319984475013512e-06, + "loss": 7.4290008544921875, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3305845521390438, + "loss_total_v6": 0.4643125869333744, + "step": 1312 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.302818018018018, + "grad_norm": 23.293209075927734, + "learning_rate": 8.317143720866671e-06, + "loss": 7.032027721405029, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3050565551966429, + "loss_total_v6": 0.4395017437636852, + "step": 1313 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30304864864864867, + "grad_norm": 25.100561141967773, + "learning_rate": 8.314301052891618e-06, + "loss": 7.003976821899414, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.30101451463997364, + "loss_total_v6": 0.4377485252916813, + "step": 1314 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3032792792792793, + "grad_norm": 23.656782150268555, + "learning_rate": 8.31145647272843e-06, + "loss": 6.714167594909668, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.28628889098763466, + "loss_total_v6": 0.41963545605540276, + "step": 1315 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3035099099099099, + "grad_norm": 22.176578521728516, + "learning_rate": 8.308609982018286e-06, + "loss": 6.7501420974731445, + "loss_alignment": 0.2672119140625, + "loss_alignment_w": 0.13360595703125, + "loss_sft": 0.28827795200049877, + "loss_total_v6": 0.421883899718523, + "step": 1316 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30374054054054056, + "grad_norm": 26.299440383911133, + "learning_rate": 8.305761582403477e-06, + "loss": 7.096827507019043, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.308938667178154, + "loss_total_v6": 0.4435517117381096, + "step": 1317 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3039711711711712, + "grad_norm": 24.03463363647461, + "learning_rate": 8.302911275527385e-06, + "loss": 7.44045352935791, + "loss_alignment": 0.284423828125, + "loss_alignment_w": 0.1422119140625, + "loss_sft": 0.322755366563797, + "loss_total_v6": 0.4650283195078373, + "step": 1318 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3042018018018018, + "grad_norm": 23.066579818725586, + "learning_rate": 8.300059063034498e-06, + "loss": 7.021427154541016, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3028681054711342, + "loss_total_v6": 0.4388391710817814, + "step": 1319 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30443243243243245, + "grad_norm": 23.37514305114746, + "learning_rate": 8.297204946570398e-06, + "loss": 6.965226173400879, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.30042366683483124, + "loss_total_v6": 0.43532662466168404, + "step": 1320 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30466306306306307, + "grad_norm": 27.52848243713379, + "learning_rate": 8.294348927781776e-06, + "loss": 6.962404251098633, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.29965222999453545, + "loss_total_v6": 0.43515027314424515, + "step": 1321 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3048936936936937, + "grad_norm": 23.232036590576172, + "learning_rate": 8.291491008316409e-06, + "loss": 7.157530784606934, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.311435641720891, + "loss_total_v6": 0.44734567403793335, + "step": 1322 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30512432432432435, + "grad_norm": 25.474973678588867, + "learning_rate": 8.288631189823177e-06, + "loss": 6.816933631896973, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.2919488362967968, + "loss_total_v6": 0.4260583333671093, + "step": 1323 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30535495495495496, + "grad_norm": 22.99012565612793, + "learning_rate": 8.285769473952052e-06, + "loss": 7.476566314697266, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.3334963209927082, + "loss_total_v6": 0.4672853909432888, + "step": 1324 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3055855855855856, + "grad_norm": 22.423194885253906, + "learning_rate": 8.282905862354107e-06, + "loss": 6.9871625900268555, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.30072660744190216, + "loss_total_v6": 0.43669767305254936, + "step": 1325 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30581621621621624, + "grad_norm": 26.53348731994629, + "learning_rate": 8.280040356681503e-06, + "loss": 7.52000093460083, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.3309772275388241, + "loss_total_v6": 0.4700000584125519, + "step": 1326 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30604684684684685, + "grad_norm": 25.398908615112305, + "learning_rate": 8.277172958587494e-06, + "loss": 6.854747295379639, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.2939764969050884, + "loss_total_v6": 0.4284216985106468, + "step": 1327 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30627747747747747, + "grad_norm": 26.18514633178711, + "learning_rate": 8.274303669726427e-06, + "loss": 7.361462593078613, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.32566146925091743, + "loss_total_v6": 0.46009141579270363, + "step": 1328 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30650810810810813, + "grad_norm": 26.571889877319336, + "learning_rate": 8.271432491753739e-06, + "loss": 7.406781196594238, + "loss_alignment": 0.27587890625, + "loss_alignment_w": 0.137939453125, + "loss_sft": 0.3250301294028759, + "loss_total_v6": 0.4629238098859787, + "step": 1329 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30673873873873875, + "grad_norm": 23.480525970458984, + "learning_rate": 8.268559426325964e-06, + "loss": 6.95412540435791, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3014541231095791, + "loss_total_v6": 0.4346328265964985, + "step": 1330 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30696936936936936, + "grad_norm": 23.492691040039062, + "learning_rate": 8.265684475100713e-06, + "loss": 7.5579023361206055, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.33687083423137665, + "loss_total_v6": 0.47236887738108635, + "step": 1331 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3072, + "grad_norm": 23.479385375976562, + "learning_rate": 8.262807639736693e-06, + "loss": 7.534477710723877, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3363681100308895, + "loss_total_v6": 0.4709048643708229, + "step": 1332 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30743063063063064, + "grad_norm": 25.545351028442383, + "learning_rate": 8.259928921893694e-06, + "loss": 7.322569847106934, + "loss_alignment": 0.2686767578125, + "loss_alignment_w": 0.13433837890625, + "loss_sft": 0.32316965609788895, + "loss_total_v6": 0.45766060799360275, + "step": 1333 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30766126126126125, + "grad_norm": 21.80638313293457, + "learning_rate": 8.257048323232598e-06, + "loss": 6.996212005615234, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.30182620510458946, + "loss_total_v6": 0.43726323917508125, + "step": 1334 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30789189189189187, + "grad_norm": 23.62477684020996, + "learning_rate": 8.254165845415366e-06, + "loss": 7.103011131286621, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.30892842449247837, + "loss_total_v6": 0.4439381994307041, + "step": 1335 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30812252252252254, + "grad_norm": 24.999296188354492, + "learning_rate": 8.251281490105045e-06, + "loss": 6.912834167480469, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.2988734282553196, + "loss_total_v6": 0.432052131742239, + "step": 1336 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30835315315315315, + "grad_norm": 22.20986557006836, + "learning_rate": 8.248395258965767e-06, + "loss": 7.643159866333008, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3408261425793171, + "loss_total_v6": 0.4776974804699421, + "step": 1337 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30858378378378376, + "grad_norm": 22.39478302001953, + "learning_rate": 8.245507153662744e-06, + "loss": 6.836678981781006, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.29220636561512947, + "loss_total_v6": 0.42729242891073227, + "step": 1338 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30881441441441443, + "grad_norm": 23.976335525512695, + "learning_rate": 8.242617175862271e-06, + "loss": 6.835709571838379, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.2905436232686043, + "loss_total_v6": 0.4272318482398987, + "step": 1339 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30904504504504504, + "grad_norm": 25.148555755615234, + "learning_rate": 8.239725327231721e-06, + "loss": 6.783169746398926, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.29417210444808006, + "loss_total_v6": 0.42394810914993286, + "step": 1340 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30927567567567565, + "grad_norm": 24.54561424255371, + "learning_rate": 8.23683160943955e-06, + "loss": 6.918654441833496, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.2998017445206642, + "loss_total_v6": 0.4324158802628517, + "step": 1341 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3095063063063063, + "grad_norm": 24.684494018554688, + "learning_rate": 8.233936024155289e-06, + "loss": 7.21291446685791, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3150497041642666, + "loss_total_v6": 0.4508071541786194, + "step": 1342 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30973693693693694, + "grad_norm": 23.75119972229004, + "learning_rate": 8.231038573049548e-06, + "loss": 7.092233180999756, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.30801064893603325, + "loss_total_v6": 0.44326456636190414, + "step": 1343 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.30996756756756755, + "grad_norm": 28.698827743530273, + "learning_rate": 8.228139257794012e-06, + "loss": 7.066964149475098, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3060193546116352, + "loss_total_v6": 0.4416852593421936, + "step": 1344 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3101981981981982, + "grad_norm": 23.357572555541992, + "learning_rate": 8.225238080061447e-06, + "loss": 7.412230968475342, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3268203493207693, + "loss_total_v6": 0.46326444298028946, + "step": 1345 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31042882882882883, + "grad_norm": 22.80537223815918, + "learning_rate": 8.222335041525683e-06, + "loss": 7.153717517852783, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.31301311776041985, + "loss_total_v6": 0.44710734859108925, + "step": 1346 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31065945945945944, + "grad_norm": 24.15171241760254, + "learning_rate": 8.219430143861638e-06, + "loss": 7.130491256713867, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3137281872332096, + "loss_total_v6": 0.4456557035446167, + "step": 1347 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3108900900900901, + "grad_norm": 25.85732078552246, + "learning_rate": 8.216523388745288e-06, + "loss": 6.664234161376953, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.28237462043762207, + "loss_total_v6": 0.41651463508605957, + "step": 1348 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3111207207207207, + "grad_norm": 23.468608856201172, + "learning_rate": 8.213614777853688e-06, + "loss": 7.047298431396484, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.30204368382692337, + "loss_total_v6": 0.4404561445116997, + "step": 1349 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31135135135135134, + "grad_norm": 24.561899185180664, + "learning_rate": 8.210704312864965e-06, + "loss": 6.916021347045898, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.2963260244578123, + "loss_total_v6": 0.43225132301449776, + "step": 1350 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.311581981981982, + "grad_norm": 28.433258056640625, + "learning_rate": 8.207791995458314e-06, + "loss": 6.930456161499023, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.29857100173830986, + "loss_total_v6": 0.43315352872014046, + "step": 1351 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3118126126126126, + "grad_norm": 28.23648452758789, + "learning_rate": 8.204877827313997e-06, + "loss": 6.6770477294921875, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.28341962583363056, + "loss_total_v6": 0.4173154830932617, + "step": 1352 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31204324324324323, + "grad_norm": 22.905174255371094, + "learning_rate": 8.201961810113346e-06, + "loss": 7.093888759613037, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.30922803096473217, + "loss_total_v6": 0.4433680474758148, + "step": 1353 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3122738738738739, + "grad_norm": 25.084993362426758, + "learning_rate": 8.199043945538759e-06, + "loss": 7.350786209106445, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.32522308081388474, + "loss_total_v6": 0.45942414179444313, + "step": 1354 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3125045045045045, + "grad_norm": 21.03199005126953, + "learning_rate": 8.196124235273698e-06, + "loss": 7.032106399536133, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.3056412748992443, + "loss_total_v6": 0.4395066313445568, + "step": 1355 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3127351351351351, + "grad_norm": 24.703351974487305, + "learning_rate": 8.193202681002692e-06, + "loss": 7.017290115356445, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.30315888300538063, + "loss_total_v6": 0.43858062848448753, + "step": 1356 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3129657657657658, + "grad_norm": 23.182384490966797, + "learning_rate": 8.190279284411337e-06, + "loss": 6.679207801818848, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.2820745129138231, + "loss_total_v6": 0.4174504950642586, + "step": 1357 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3131963963963964, + "grad_norm": 22.40826988220215, + "learning_rate": 8.187354047186286e-06, + "loss": 7.166680812835693, + "loss_alignment": 0.278076171875, + "loss_alignment_w": 0.1390380859375, + "loss_sft": 0.30887944996356964, + "loss_total_v6": 0.44791755080223083, + "step": 1358 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.313427027027027, + "grad_norm": 24.085798263549805, + "learning_rate": 8.184426971015257e-06, + "loss": 7.164600372314453, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.31346440128982067, + "loss_total_v6": 0.447787519544363, + "step": 1359 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3136576576576577, + "grad_norm": 27.202049255371094, + "learning_rate": 8.181498057587028e-06, + "loss": 6.867501258850098, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.29501780308783054, + "loss_total_v6": 0.4292188473045826, + "step": 1360 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3138882882882883, + "grad_norm": 25.484331130981445, + "learning_rate": 8.178567308591438e-06, + "loss": 7.411016464233398, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.3290332555770874, + "loss_total_v6": 0.463188536465168, + "step": 1361 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3141189189189189, + "grad_norm": 26.004074096679688, + "learning_rate": 8.175634725719384e-06, + "loss": 7.193127632141113, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3137977682054043, + "loss_total_v6": 0.4495704732835293, + "step": 1362 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3143495495495495, + "grad_norm": 26.4322452545166, + "learning_rate": 8.17270031066282e-06, + "loss": 7.165665626525879, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.31285957992076874, + "loss_total_v6": 0.44785410538315773, + "step": 1363 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3145801801801802, + "grad_norm": 26.10048484802246, + "learning_rate": 8.169764065114764e-06, + "loss": 7.080026149749756, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.30601176992058754, + "loss_total_v6": 0.44250163063406944, + "step": 1364 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3148108108108108, + "grad_norm": 22.25503158569336, + "learning_rate": 8.166825990769281e-06, + "loss": 7.508277893066406, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3374161869287491, + "loss_total_v6": 0.4692673608660698, + "step": 1365 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3150414414414414, + "grad_norm": 25.029808044433594, + "learning_rate": 8.163886089321493e-06, + "loss": 6.945608139038086, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.30353105068206787, + "loss_total_v6": 0.43410051986575127, + "step": 1366 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3152720720720721, + "grad_norm": 23.93115997314453, + "learning_rate": 8.160944362467582e-06, + "loss": 7.008689880371094, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3023009430617094, + "loss_total_v6": 0.43804312124848366, + "step": 1367 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3155027027027027, + "grad_norm": 21.47385597229004, + "learning_rate": 8.158000811904779e-06, + "loss": 7.075894355773926, + "loss_alignment": 0.2686767578125, + "loss_alignment_w": 0.13433837890625, + "loss_sft": 0.30778297781944275, + "loss_total_v6": 0.44224341958761215, + "step": 1368 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3157333333333333, + "grad_norm": 22.213985443115234, + "learning_rate": 8.155055439331363e-06, + "loss": 6.984752655029297, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.30385664477944374, + "loss_total_v6": 0.43654707074165344, + "step": 1369 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.315963963963964, + "grad_norm": 23.759931564331055, + "learning_rate": 8.152108246446673e-06, + "loss": 6.776278018951416, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.28696646355092525, + "loss_total_v6": 0.4235173724591732, + "step": 1370 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3161945945945946, + "grad_norm": 24.236698150634766, + "learning_rate": 8.149159234951091e-06, + "loss": 7.0305352210998535, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.30571093037724495, + "loss_total_v6": 0.43940845131874084, + "step": 1371 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3164252252252252, + "grad_norm": 23.207744598388672, + "learning_rate": 8.146208406546054e-06, + "loss": 7.513299465179443, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3361736163496971, + "loss_total_v6": 0.4695812091231346, + "step": 1372 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3166558558558559, + "grad_norm": 23.462669372558594, + "learning_rate": 8.143255762934042e-06, + "loss": 7.475886344909668, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.33056989312171936, + "loss_total_v6": 0.46724288165569305, + "step": 1373 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3168864864864865, + "grad_norm": 23.156681060791016, + "learning_rate": 8.140301305818586e-06, + "loss": 7.01124382019043, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.30476463586091995, + "loss_total_v6": 0.43820273876190186, + "step": 1374 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3171171171171171, + "grad_norm": 26.230144500732422, + "learning_rate": 8.13734503690426e-06, + "loss": 7.312555313110352, + "loss_alignment": 0.2711181640625, + "loss_alignment_w": 0.13555908203125, + "loss_sft": 0.321323037147522, + "loss_total_v6": 0.4570347033441067, + "step": 1375 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31734774774774777, + "grad_norm": 24.962629318237305, + "learning_rate": 8.134386957896688e-06, + "loss": 7.363590240478516, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.32541298493742943, + "loss_total_v6": 0.4602243900299072, + "step": 1376 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3175783783783784, + "grad_norm": 26.4075927734375, + "learning_rate": 8.131427070502535e-06, + "loss": 7.281425476074219, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.31821775808930397, + "loss_total_v6": 0.45508909597992897, + "step": 1377 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.317809009009009, + "grad_norm": 23.510953903198242, + "learning_rate": 8.128465376429507e-06, + "loss": 7.136317253112793, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.30997245758771896, + "loss_total_v6": 0.44601983204483986, + "step": 1378 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31803963963963966, + "grad_norm": 28.620525360107422, + "learning_rate": 8.125501877386359e-06, + "loss": 7.113382339477539, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3142000623047352, + "loss_total_v6": 0.4445863999426365, + "step": 1379 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3182702702702703, + "grad_norm": 25.616233825683594, + "learning_rate": 8.122536575082883e-06, + "loss": 7.214310169219971, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3160982374101877, + "loss_total_v6": 0.45089438930153847, + "step": 1380 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3185009009009009, + "grad_norm": 22.412797927856445, + "learning_rate": 8.11956947122991e-06, + "loss": 7.40327787399292, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3279850296676159, + "loss_total_v6": 0.4627048783004284, + "step": 1381 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31873153153153155, + "grad_norm": 25.56075668334961, + "learning_rate": 8.116600567539313e-06, + "loss": 6.774415016174316, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.28962715342640877, + "loss_total_v6": 0.42340094968676567, + "step": 1382 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31896216216216217, + "grad_norm": 30.004243850708008, + "learning_rate": 8.113629865724006e-06, + "loss": 7.237359046936035, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3155856616795063, + "loss_total_v6": 0.4523349404335022, + "step": 1383 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3191927927927928, + "grad_norm": 23.777631759643555, + "learning_rate": 8.110657367497934e-06, + "loss": 7.525062561035156, + "loss_alignment": 0.2745361328125, + "loss_alignment_w": 0.13726806640625, + "loss_sft": 0.3327278830111027, + "loss_total_v6": 0.47031639888882637, + "step": 1384 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31942342342342345, + "grad_norm": 24.68973731994629, + "learning_rate": 8.107683074576081e-06, + "loss": 6.925248146057129, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.29685697704553604, + "loss_total_v6": 0.43282803520560265, + "step": 1385 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.31965405405405406, + "grad_norm": 23.973522186279297, + "learning_rate": 8.104706988674471e-06, + "loss": 6.991024017333984, + "loss_alignment": 0.2701416015625, + "loss_alignment_w": 0.13507080078125, + "loss_sft": 0.301807202398777, + "loss_total_v6": 0.4369390271604061, + "step": 1386 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3198846846846847, + "grad_norm": 27.229557037353516, + "learning_rate": 8.101729111510158e-06, + "loss": 7.102800369262695, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.31152450293302536, + "loss_total_v6": 0.44392502307891846, + "step": 1387 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32011531531531534, + "grad_norm": 30.101354598999023, + "learning_rate": 8.098749444801226e-06, + "loss": 6.862565517425537, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.29545697942376137, + "loss_total_v6": 0.42891034856438637, + "step": 1388 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32034594594594595, + "grad_norm": 25.154664993286133, + "learning_rate": 8.095767990266799e-06, + "loss": 7.353652477264404, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.32343388348817825, + "loss_total_v6": 0.45960329473018646, + "step": 1389 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32057657657657657, + "grad_norm": 24.519987106323242, + "learning_rate": 8.09278474962703e-06, + "loss": 6.9966888427734375, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.30374813452363014, + "loss_total_v6": 0.43729307129979134, + "step": 1390 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32080720720720723, + "grad_norm": 21.800983428955078, + "learning_rate": 8.089799724603099e-06, + "loss": 7.29288911819458, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.32030753791332245, + "loss_total_v6": 0.45580558478832245, + "step": 1391 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32103783783783785, + "grad_norm": 23.147415161132812, + "learning_rate": 8.086812916917219e-06, + "loss": 6.906752586364746, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.298966359347105, + "loss_total_v6": 0.4316720478236675, + "step": 1392 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32126846846846846, + "grad_norm": 24.278200149536133, + "learning_rate": 8.083824328292631e-06, + "loss": 6.724304676055908, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.2849388178437948, + "loss_total_v6": 0.42026903480291367, + "step": 1393 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3214990990990991, + "grad_norm": 24.61444091796875, + "learning_rate": 8.080833960453604e-06, + "loss": 6.748900413513184, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.28548426181077957, + "loss_total_v6": 0.42180628702044487, + "step": 1394 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32172972972972974, + "grad_norm": 24.710378646850586, + "learning_rate": 8.077841815125431e-06, + "loss": 7.320981502532959, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.32011017203330994, + "loss_total_v6": 0.45756133273243904, + "step": 1395 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32196036036036035, + "grad_norm": 22.775470733642578, + "learning_rate": 8.074847894034434e-06, + "loss": 7.50501823425293, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.3352288156747818, + "loss_total_v6": 0.4690636545419693, + "step": 1396 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32219099099099097, + "grad_norm": 24.763355255126953, + "learning_rate": 8.071852198907956e-06, + "loss": 7.080883026123047, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.30786583200097084, + "loss_total_v6": 0.44255517423152924, + "step": 1397 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32242162162162163, + "grad_norm": 23.46663475036621, + "learning_rate": 8.068854731474366e-06, + "loss": 6.945823669433594, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.300324909389019, + "loss_total_v6": 0.4341139905154705, + "step": 1398 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32265225225225225, + "grad_norm": 25.222965240478516, + "learning_rate": 8.065855493463055e-06, + "loss": 6.978160858154297, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.30562663450837135, + "loss_total_v6": 0.43613504618406296, + "step": 1399 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32288288288288286, + "grad_norm": 26.094554901123047, + "learning_rate": 8.062854486604434e-06, + "loss": 6.988152503967285, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3036265932023525, + "loss_total_v6": 0.4367595240473747, + "step": 1400 + }, + { + "epoch": 0.32288288288288286, + "eval_loss": 0.6003024578094482, + "eval_loss_alignment": 0.2429656285673516, + "eval_loss_alignment_w": 0.1214828142836758, + "eval_loss_sft": 0.478819569002956, + "eval_loss_total_v6": 0.6003023835417887, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.1745, + "eval_samples_per_second": 19.139, + "eval_steps_per_second": 2.402, + "step": 1400 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32311351351351353, + "grad_norm": 30.31133270263672, + "learning_rate": 8.059851712629937e-06, + "loss": 7.148717880249023, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.31099165976047516, + "loss_total_v6": 0.44679487869143486, + "step": 1401 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32334414414414414, + "grad_norm": 26.996347427368164, + "learning_rate": 8.056847173272017e-06, + "loss": 7.238297462463379, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.31875709258019924, + "loss_total_v6": 0.4523935690522194, + "step": 1402 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32357477477477475, + "grad_norm": 24.737804412841797, + "learning_rate": 8.053840870264144e-06, + "loss": 7.429843425750732, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.3269903250038624, + "loss_total_v6": 0.4643652066588402, + "step": 1403 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3238054054054054, + "grad_norm": 48.55314636230469, + "learning_rate": 8.050832805340807e-06, + "loss": 7.5280914306640625, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3355875015258789, + "loss_total_v6": 0.4705057069659233, + "step": 1404 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32403603603603603, + "grad_norm": 22.921194076538086, + "learning_rate": 8.047822980237509e-06, + "loss": 7.33331823348999, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3217051848769188, + "loss_total_v6": 0.4583323933184147, + "step": 1405 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32426666666666665, + "grad_norm": 25.707134246826172, + "learning_rate": 8.044811396690772e-06, + "loss": 6.759302616119385, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.28762975707650185, + "loss_total_v6": 0.42245640978217125, + "step": 1406 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3244972972972973, + "grad_norm": 28.88498878479004, + "learning_rate": 8.041798056438133e-06, + "loss": 7.160155773162842, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.31047054566442966, + "loss_total_v6": 0.4475097432732582, + "step": 1407 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32472792792792793, + "grad_norm": 23.3710994720459, + "learning_rate": 8.038782961218136e-06, + "loss": 6.920056343078613, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.2959373351186514, + "loss_total_v6": 0.43250349909067154, + "step": 1408 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32495855855855854, + "grad_norm": 25.082130432128906, + "learning_rate": 8.035766112770349e-06, + "loss": 6.7453765869140625, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.2878122366964817, + "loss_total_v6": 0.4215860515832901, + "step": 1409 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3251891891891892, + "grad_norm": 26.091684341430664, + "learning_rate": 8.032747512835338e-06, + "loss": 7.430635929107666, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3292981646955013, + "loss_total_v6": 0.4644147530198097, + "step": 1410 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3254198198198198, + "grad_norm": 26.95563316345215, + "learning_rate": 8.029727163154691e-06, + "loss": 6.8447747230529785, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.29101862758398056, + "loss_total_v6": 0.42779841274023056, + "step": 1411 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32565045045045043, + "grad_norm": 22.688053131103516, + "learning_rate": 8.026705065470996e-06, + "loss": 7.3363447189331055, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.3219553977251053, + "loss_total_v6": 0.45852155610919, + "step": 1412 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3258810810810811, + "grad_norm": 26.65427589416504, + "learning_rate": 8.023681221527858e-06, + "loss": 6.690401554107666, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.28805367089807987, + "loss_total_v6": 0.4181501120328903, + "step": 1413 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3261117117117117, + "grad_norm": 24.031221389770508, + "learning_rate": 8.020655633069883e-06, + "loss": 6.833742618560791, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.2920533623546362, + "loss_total_v6": 0.42710892483592033, + "step": 1414 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.32634234234234233, + "grad_norm": 23.67163848876953, + "learning_rate": 8.017628301842689e-06, + "loss": 7.273850440979004, + "loss_alignment": 0.2763671875, + "loss_alignment_w": 0.13818359375, + "loss_sft": 0.31638630107045174, + "loss_total_v6": 0.45461566746234894, + "step": 1415 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.326572972972973, + "grad_norm": 22.83223533630371, + "learning_rate": 8.014599229592895e-06, + "loss": 7.339345932006836, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.32160888612270355, + "loss_total_v6": 0.45870909839868546, + "step": 1416 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3268036036036036, + "grad_norm": 23.195302963256836, + "learning_rate": 8.011568418068124e-06, + "loss": 7.119099140167236, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.30927780643105507, + "loss_total_v6": 0.44494369626045227, + "step": 1417 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3270342342342342, + "grad_norm": 24.86341667175293, + "learning_rate": 8.008535869017012e-06, + "loss": 7.0345563888549805, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.3053671829402447, + "loss_total_v6": 0.4396597743034363, + "step": 1418 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3272648648648649, + "grad_norm": 22.580312728881836, + "learning_rate": 8.005501584189184e-06, + "loss": 6.644675254821777, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.2793669290840626, + "loss_total_v6": 0.4152922183275223, + "step": 1419 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3274954954954955, + "grad_norm": 23.347187042236328, + "learning_rate": 8.002465565335272e-06, + "loss": 6.949535846710205, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.2987106330692768, + "loss_total_v6": 0.4343460015952587, + "step": 1420 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3277261261261261, + "grad_norm": 25.1131591796875, + "learning_rate": 7.999427814206911e-06, + "loss": 6.978561878204346, + "loss_alignment": 0.2681884765625, + "loss_alignment_w": 0.13409423828125, + "loss_sft": 0.30215744860470295, + "loss_total_v6": 0.4361601285636425, + "step": 1421 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3279567567567568, + "grad_norm": 25.4200439453125, + "learning_rate": 7.996388332556735e-06, + "loss": 6.863224983215332, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2971308622509241, + "loss_total_v6": 0.42895153909921646, + "step": 1422 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3281873873873874, + "grad_norm": 25.24857521057129, + "learning_rate": 7.993347122138374e-06, + "loss": 7.408096790313721, + "loss_alignment": 0.27783203125, + "loss_alignment_w": 0.138916015625, + "loss_sft": 0.3241510409861803, + "loss_total_v6": 0.46300602704286575, + "step": 1423 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.328418018018018, + "grad_norm": 28.91551399230957, + "learning_rate": 7.990304184706454e-06, + "loss": 6.662647247314453, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.28351139836013317, + "loss_total_v6": 0.4164154455065727, + "step": 1424 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3286486486486486, + "grad_norm": 22.42705726623535, + "learning_rate": 7.987259522016602e-06, + "loss": 7.211906433105469, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.31787061132490635, + "loss_total_v6": 0.4507441520690918, + "step": 1425 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3288792792792793, + "grad_norm": 22.76435089111328, + "learning_rate": 7.984213135825437e-06, + "loss": 7.170019626617432, + "loss_alignment": 0.2637939453125, + "loss_alignment_w": 0.13189697265625, + "loss_sft": 0.3159545809030533, + "loss_total_v6": 0.4481262154877186, + "step": 1426 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3291099099099099, + "grad_norm": 27.20754051208496, + "learning_rate": 7.981165027890576e-06, + "loss": 6.872189521789551, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.2925336975604296, + "loss_total_v6": 0.4295118451118469, + "step": 1427 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3293405405405405, + "grad_norm": 25.132671356201172, + "learning_rate": 7.978115199970621e-06, + "loss": 6.79611873626709, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.2906631864607334, + "loss_total_v6": 0.424757432192564, + "step": 1428 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3295711711711712, + "grad_norm": 24.23400115966797, + "learning_rate": 7.975063653825179e-06, + "loss": 6.79135799407959, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.29157106578350067, + "loss_total_v6": 0.42445986345410347, + "step": 1429 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3298018018018018, + "grad_norm": 25.679643630981445, + "learning_rate": 7.972010391214837e-06, + "loss": 6.791718482971191, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.2886486742645502, + "loss_total_v6": 0.42448239773511887, + "step": 1430 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3300324324324324, + "grad_norm": 24.194082260131836, + "learning_rate": 7.96895541390118e-06, + "loss": 6.998727798461914, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3015257213264704, + "loss_total_v6": 0.4374205023050308, + "step": 1431 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3302630630630631, + "grad_norm": 24.667152404785156, + "learning_rate": 7.965898723646777e-06, + "loss": 7.462456703186035, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.3286319524049759, + "loss_total_v6": 0.4664035551249981, + "step": 1432 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3304936936936937, + "grad_norm": 24.59953498840332, + "learning_rate": 7.962840322215188e-06, + "loss": 6.8153533935546875, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.29030894488096237, + "loss_total_v6": 0.42595959454774857, + "step": 1433 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3307243243243243, + "grad_norm": 29.374746322631836, + "learning_rate": 7.959780211370958e-06, + "loss": 6.943126678466797, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.2992102839052677, + "loss_total_v6": 0.4339453987777233, + "step": 1434 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33095495495495497, + "grad_norm": 27.316158294677734, + "learning_rate": 7.956718392879623e-06, + "loss": 7.23390531539917, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.3146068826317787, + "loss_total_v6": 0.4521190822124481, + "step": 1435 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3311855855855856, + "grad_norm": 24.5533447265625, + "learning_rate": 7.9536548685077e-06, + "loss": 7.456798076629639, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.3304908163845539, + "loss_total_v6": 0.4660498835146427, + "step": 1436 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3314162162162162, + "grad_norm": 27.242586135864258, + "learning_rate": 7.95058964002269e-06, + "loss": 6.9970383644104, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3034037686884403, + "loss_total_v6": 0.4373149052262306, + "step": 1437 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33164684684684687, + "grad_norm": 33.06241226196289, + "learning_rate": 7.94752270919308e-06, + "loss": 7.001232147216797, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3014685846865177, + "loss_total_v6": 0.4375769942998886, + "step": 1438 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3318774774774775, + "grad_norm": 26.73855972290039, + "learning_rate": 7.944454077788339e-06, + "loss": 7.145398139953613, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.31256942451000214, + "loss_total_v6": 0.44658737257122993, + "step": 1439 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3321081081081081, + "grad_norm": 23.709495544433594, + "learning_rate": 7.941383747578913e-06, + "loss": 7.207231521606445, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3119632191956043, + "loss_total_v6": 0.450451985001564, + "step": 1440 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33233873873873876, + "grad_norm": 24.104143142700195, + "learning_rate": 7.93831172033623e-06, + "loss": 6.6347784996032715, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.2804115954786539, + "loss_total_v6": 0.41467365622520447, + "step": 1441 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33256936936936937, + "grad_norm": 23.546552658081055, + "learning_rate": 7.935237997832704e-06, + "loss": 6.891650199890137, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.2956420723348856, + "loss_total_v6": 0.43072811514139175, + "step": 1442 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3328, + "grad_norm": 28.384613037109375, + "learning_rate": 7.932162581841715e-06, + "loss": 7.203347206115723, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.31382616609334946, + "loss_total_v6": 0.45020921900868416, + "step": 1443 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33303063063063065, + "grad_norm": 23.66608428955078, + "learning_rate": 7.92908547413763e-06, + "loss": 7.412539482116699, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.3304864689707756, + "loss_total_v6": 0.4632837027311325, + "step": 1444 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33326126126126127, + "grad_norm": 23.732507705688477, + "learning_rate": 7.926006676495787e-06, + "loss": 6.734094619750977, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.2856575436890125, + "loss_total_v6": 0.4208809323608875, + "step": 1445 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3334918918918919, + "grad_norm": 22.804452896118164, + "learning_rate": 7.9229261906925e-06, + "loss": 7.69174861907959, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.34680790826678276, + "loss_total_v6": 0.48073428869247437, + "step": 1446 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33372252252252255, + "grad_norm": 21.610523223876953, + "learning_rate": 7.919844018505058e-06, + "loss": 6.302606582641602, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.2595592774450779, + "loss_total_v6": 0.3939129114151001, + "step": 1447 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33395315315315316, + "grad_norm": 25.79914665222168, + "learning_rate": 7.916760161711721e-06, + "loss": 7.407763481140137, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.32718198001384735, + "loss_total_v6": 0.46298520267009735, + "step": 1448 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33418378378378377, + "grad_norm": 22.617950439453125, + "learning_rate": 7.913674622091722e-06, + "loss": 7.206002712249756, + "loss_alignment": 0.2684326171875, + "loss_alignment_w": 0.13421630859375, + "loss_sft": 0.31621992215514183, + "loss_total_v6": 0.45037518069148064, + "step": 1449 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33441441441441444, + "grad_norm": 24.0794677734375, + "learning_rate": 7.91058740142527e-06, + "loss": 7.16302490234375, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3136863596737385, + "loss_total_v6": 0.4476890303194523, + "step": 1450 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33464504504504505, + "grad_norm": 25.90068244934082, + "learning_rate": 7.907498501493532e-06, + "loss": 7.2944793701171875, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3237486034631729, + "loss_total_v6": 0.4559049755334854, + "step": 1451 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33487567567567567, + "grad_norm": 24.756898880004883, + "learning_rate": 7.904407924078653e-06, + "loss": 7.073845386505127, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.30776171013712883, + "loss_total_v6": 0.44211532548069954, + "step": 1452 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3351063063063063, + "grad_norm": 22.43535614013672, + "learning_rate": 7.901315670963747e-06, + "loss": 7.034054279327393, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.30234504118561745, + "loss_total_v6": 0.43962837010622025, + "step": 1453 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33533693693693695, + "grad_norm": 23.578170776367188, + "learning_rate": 7.898221743932887e-06, + "loss": 7.228985786437988, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.32131844758987427, + "loss_total_v6": 0.45181162655353546, + "step": 1454 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33556756756756756, + "grad_norm": 22.008102416992188, + "learning_rate": 7.89512614477112e-06, + "loss": 6.619969367980957, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.27991325221955776, + "loss_total_v6": 0.4137480780482292, + "step": 1455 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3357981981981982, + "grad_norm": 25.261369705200195, + "learning_rate": 7.892028875264452e-06, + "loss": 6.6655592918396, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.28114520758390427, + "loss_total_v6": 0.41659747809171677, + "step": 1456 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33602882882882884, + "grad_norm": 24.191749572753906, + "learning_rate": 7.888929937199853e-06, + "loss": 7.28729772567749, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.31931719556450844, + "loss_total_v6": 0.45545611158013344, + "step": 1457 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33625945945945945, + "grad_norm": 25.113096237182617, + "learning_rate": 7.88582933236526e-06, + "loss": 6.905590057373047, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.29880212247371674, + "loss_total_v6": 0.43159937486052513, + "step": 1458 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33649009009009007, + "grad_norm": 23.545886993408203, + "learning_rate": 7.882727062549569e-06, + "loss": 6.645877838134766, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.27964042499661446, + "loss_total_v6": 0.41536734625697136, + "step": 1459 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33672072072072073, + "grad_norm": 22.249658584594727, + "learning_rate": 7.879623129542633e-06, + "loss": 7.554465293884277, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.33641189709305763, + "loss_total_v6": 0.47215409576892853, + "step": 1460 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33695135135135135, + "grad_norm": 25.689424514770508, + "learning_rate": 7.876517535135273e-06, + "loss": 7.006911754608154, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3045090977102518, + "loss_total_v6": 0.43793196603655815, + "step": 1461 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33718198198198196, + "grad_norm": 23.322694778442383, + "learning_rate": 7.87341028111926e-06, + "loss": 7.0140228271484375, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.30650994926691055, + "loss_total_v6": 0.43837640434503555, + "step": 1462 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33741261261261263, + "grad_norm": 24.533639907836914, + "learning_rate": 7.870301369287328e-06, + "loss": 7.478399276733398, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3314746543765068, + "loss_total_v6": 0.4673999398946762, + "step": 1463 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33764324324324324, + "grad_norm": 24.843534469604492, + "learning_rate": 7.867190801433167e-06, + "loss": 6.799516677856445, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.2915622182190418, + "loss_total_v6": 0.4249698109924793, + "step": 1464 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33787387387387385, + "grad_norm": 23.109760284423828, + "learning_rate": 7.864078579351418e-06, + "loss": 6.56563663482666, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.27753976359963417, + "loss_total_v6": 0.41035226732492447, + "step": 1465 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3381045045045045, + "grad_norm": 25.02081871032715, + "learning_rate": 7.86096470483768e-06, + "loss": 6.9546589851379395, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.2981153018772602, + "loss_total_v6": 0.4346662051975727, + "step": 1466 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33833513513513513, + "grad_norm": 22.677839279174805, + "learning_rate": 7.857849179688506e-06, + "loss": 7.2243757247924805, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.31430117785930634, + "loss_total_v6": 0.45152348279953003, + "step": 1467 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33856576576576575, + "grad_norm": 24.722963333129883, + "learning_rate": 7.854732005701401e-06, + "loss": 7.2477192878723145, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.31502773612737656, + "loss_total_v6": 0.45298245549201965, + "step": 1468 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3387963963963964, + "grad_norm": 24.92571449279785, + "learning_rate": 7.851613184674821e-06, + "loss": 7.141932964324951, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.30988091602921486, + "loss_total_v6": 0.44637080654501915, + "step": 1469 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33902702702702703, + "grad_norm": 23.874370574951172, + "learning_rate": 7.84849271840817e-06, + "loss": 6.720829963684082, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.28522520512342453, + "loss_total_v6": 0.42005186527967453, + "step": 1470 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33925765765765764, + "grad_norm": 24.81334686279297, + "learning_rate": 7.845370608701804e-06, + "loss": 7.340767860412598, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.32554300129413605, + "loss_total_v6": 0.45879799127578735, + "step": 1471 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3394882882882883, + "grad_norm": 21.500768661499023, + "learning_rate": 7.842246857357024e-06, + "loss": 7.257551670074463, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.31788531318306923, + "loss_total_v6": 0.45359698310494423, + "step": 1472 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3397189189189189, + "grad_norm": 25.791179656982422, + "learning_rate": 7.839121466176082e-06, + "loss": 6.999098777770996, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.304600665345788, + "loss_total_v6": 0.43744368851184845, + "step": 1473 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.33994954954954953, + "grad_norm": 24.046051025390625, + "learning_rate": 7.835994436962176e-06, + "loss": 6.583662986755371, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.2793531082570553, + "loss_total_v6": 0.4114789590239525, + "step": 1474 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3401801801801802, + "grad_norm": 24.141016006469727, + "learning_rate": 7.832865771519445e-06, + "loss": 6.86705207824707, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.29503548331558704, + "loss_total_v6": 0.4291907511651516, + "step": 1475 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3404108108108108, + "grad_norm": 29.783544540405273, + "learning_rate": 7.829735471652978e-06, + "loss": 7.075464725494385, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.30854952707886696, + "loss_total_v6": 0.44221652671694756, + "step": 1476 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34064144144144143, + "grad_norm": 26.08761215209961, + "learning_rate": 7.826603539168798e-06, + "loss": 6.862212181091309, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.29343599267303944, + "loss_total_v6": 0.428888276219368, + "step": 1477 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3408720720720721, + "grad_norm": 23.551206588745117, + "learning_rate": 7.823469975873881e-06, + "loss": 7.01002311706543, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3060921113938093, + "loss_total_v6": 0.43812642619013786, + "step": 1478 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3411027027027027, + "grad_norm": 26.144527435302734, + "learning_rate": 7.820334783576134e-06, + "loss": 7.01641845703125, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.30350114218890667, + "loss_total_v6": 0.438526164740324, + "step": 1479 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3413333333333333, + "grad_norm": 30.028329849243164, + "learning_rate": 7.81719796408441e-06, + "loss": 7.464802265167236, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.33001450821757317, + "loss_total_v6": 0.46655014157295227, + "step": 1480 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.341563963963964, + "grad_norm": 26.97036361694336, + "learning_rate": 7.814059519208503e-06, + "loss": 7.034674644470215, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.30452005565166473, + "loss_total_v6": 0.43966716155409813, + "step": 1481 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3417945945945946, + "grad_norm": 25.499380111694336, + "learning_rate": 7.810919450759134e-06, + "loss": 7.094943523406982, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.30795117653906345, + "loss_total_v6": 0.4434339739382267, + "step": 1482 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3420252252252252, + "grad_norm": 26.099184036254883, + "learning_rate": 7.807777760547972e-06, + "loss": 7.470391273498535, + "loss_alignment": 0.276123046875, + "loss_alignment_w": 0.1380615234375, + "loss_sft": 0.32847169041633606, + "loss_total_v6": 0.46689945086836815, + "step": 1483 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34225585585585583, + "grad_norm": 27.60753059387207, + "learning_rate": 7.804634450387616e-06, + "loss": 7.653133869171143, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.3427770435810089, + "loss_total_v6": 0.4783208630979061, + "step": 1484 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3424864864864865, + "grad_norm": 23.98664093017578, + "learning_rate": 7.8014895220916e-06, + "loss": 6.902442455291748, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.2940125185996294, + "loss_total_v6": 0.43140264227986336, + "step": 1485 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3427171171171171, + "grad_norm": 21.261608123779297, + "learning_rate": 7.798342977474391e-06, + "loss": 7.187154769897461, + "loss_alignment": 0.2745361328125, + "loss_alignment_w": 0.13726806640625, + "loss_sft": 0.3124021142721176, + "loss_total_v6": 0.449197169393301, + "step": 1486 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3429477477477477, + "grad_norm": 26.5393123626709, + "learning_rate": 7.795194818351395e-06, + "loss": 7.00383186340332, + "loss_alignment": 0.27490234375, + "loss_alignment_w": 0.137451171875, + "loss_sft": 0.3005477301776409, + "loss_total_v6": 0.4377394951879978, + "step": 1487 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3431783783783784, + "grad_norm": 26.15106964111328, + "learning_rate": 7.79204504653894e-06, + "loss": 7.3750152587890625, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3262643814086914, + "loss_total_v6": 0.4609384424984455, + "step": 1488 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.343409009009009, + "grad_norm": 21.675188064575195, + "learning_rate": 7.788893663854292e-06, + "loss": 7.216614723205566, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.3157387264072895, + "loss_total_v6": 0.4510384052991867, + "step": 1489 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3436396396396396, + "grad_norm": 23.37644386291504, + "learning_rate": 7.785740672115641e-06, + "loss": 7.437471389770508, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.32656683027744293, + "loss_total_v6": 0.46484196186065674, + "step": 1490 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3438702702702703, + "grad_norm": 23.157291412353516, + "learning_rate": 7.782586073142107e-06, + "loss": 7.555671215057373, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3380131460726261, + "loss_total_v6": 0.4722294583916664, + "step": 1491 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3441009009009009, + "grad_norm": 25.4503116607666, + "learning_rate": 7.77942986875374e-06, + "loss": 6.476682186126709, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.2721784897148609, + "loss_total_v6": 0.4047926291823387, + "step": 1492 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3443315315315315, + "grad_norm": 25.941965103149414, + "learning_rate": 7.776272060771513e-06, + "loss": 6.837047576904297, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.2948234435170889, + "loss_total_v6": 0.42731550335884094, + "step": 1493 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3445621621621622, + "grad_norm": 21.740074157714844, + "learning_rate": 7.773112651017326e-06, + "loss": 6.971194267272949, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3011018745601177, + "loss_total_v6": 0.4356996491551399, + "step": 1494 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3447927927927928, + "grad_norm": 21.698253631591797, + "learning_rate": 7.769951641314002e-06, + "loss": 6.996541976928711, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3039373401552439, + "loss_total_v6": 0.4372838884592056, + "step": 1495 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3450234234234234, + "grad_norm": 26.373485565185547, + "learning_rate": 7.766789033485287e-06, + "loss": 6.933733940124512, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.3016902729868889, + "loss_total_v6": 0.4333583787083626, + "step": 1496 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34525405405405407, + "grad_norm": 26.59595489501953, + "learning_rate": 7.763624829355851e-06, + "loss": 7.413762092590332, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.32729751989245415, + "loss_total_v6": 0.46336015313863754, + "step": 1497 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3454846846846847, + "grad_norm": 23.25310707092285, + "learning_rate": 7.760459030751285e-06, + "loss": 7.332746505737305, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.32205095887184143, + "loss_total_v6": 0.45829667523503304, + "step": 1498 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3457153153153153, + "grad_norm": 24.159332275390625, + "learning_rate": 7.757291639498094e-06, + "loss": 7.007566928863525, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.30761709809303284, + "loss_total_v6": 0.43797293305397034, + "step": 1499 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34594594594594597, + "grad_norm": 24.61600112915039, + "learning_rate": 7.75412265742371e-06, + "loss": 7.008996963500977, + "loss_alignment": 0.2696533203125, + "loss_alignment_w": 0.13482666015625, + "loss_sft": 0.30341873317956924, + "loss_total_v6": 0.43806230649352074, + "step": 1500 + }, + { + "epoch": 0.34594594594594597, + "eval_loss": 0.597126305103302, + "eval_loss_alignment": 0.2424171482591324, + "eval_loss_alignment_w": 0.1212085741295662, + "eval_loss_sft": 0.4759177350562457, + "eval_loss_total_v6": 0.597126309781178, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.3046, + "eval_samples_per_second": 19.112, + "eval_steps_per_second": 2.399, + "step": 1500 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3461765765765766, + "grad_norm": 25.536584854125977, + "learning_rate": 7.750952086356477e-06, + "loss": 7.197003364562988, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.31342966109514236, + "loss_total_v6": 0.44981271028518677, + "step": 1501 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3464072072072072, + "grad_norm": 33.97829055786133, + "learning_rate": 7.74777992812566e-06, + "loss": 6.827275276184082, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.2953570168465376, + "loss_total_v6": 0.42670467868447304, + "step": 1502 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34663783783783786, + "grad_norm": 22.84373664855957, + "learning_rate": 7.744606184561439e-06, + "loss": 6.805293560028076, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.29044314101338387, + "loss_total_v6": 0.42533083260059357, + "step": 1503 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34686846846846847, + "grad_norm": 24.137924194335938, + "learning_rate": 7.741430857494904e-06, + "loss": 7.075072765350342, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3097304981201887, + "loss_total_v6": 0.44219205155968666, + "step": 1504 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3470990990990991, + "grad_norm": 24.947675704956055, + "learning_rate": 7.738253948758063e-06, + "loss": 6.699599742889404, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2856530640274286, + "loss_total_v6": 0.41872497648000717, + "step": 1505 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34732972972972975, + "grad_norm": 23.922733306884766, + "learning_rate": 7.735075460183835e-06, + "loss": 7.0381059646606445, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.30494814179837704, + "loss_total_v6": 0.439881619066, + "step": 1506 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34756036036036037, + "grad_norm": 23.00334930419922, + "learning_rate": 7.731895393606052e-06, + "loss": 6.7980804443359375, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.29014492221176624, + "loss_total_v6": 0.4248800165951252, + "step": 1507 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.347790990990991, + "grad_norm": 24.04009246826172, + "learning_rate": 7.728713750859458e-06, + "loss": 7.186995506286621, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3135518543422222, + "loss_total_v6": 0.4491872265934944, + "step": 1508 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34802162162162165, + "grad_norm": 23.3593807220459, + "learning_rate": 7.7255305337797e-06, + "loss": 7.238685607910156, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.32065822556614876, + "loss_total_v6": 0.45241787657141685, + "step": 1509 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34825225225225226, + "grad_norm": 24.585235595703125, + "learning_rate": 7.72234574420334e-06, + "loss": 7.04229211807251, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.30565228313207626, + "loss_total_v6": 0.44014324992895126, + "step": 1510 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34848288288288287, + "grad_norm": 22.83348274230957, + "learning_rate": 7.71915938396784e-06, + "loss": 6.830787658691406, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.29371500574052334, + "loss_total_v6": 0.4269242212176323, + "step": 1511 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34871351351351354, + "grad_norm": 25.269290924072266, + "learning_rate": 7.715971454911576e-06, + "loss": 6.907745838165283, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2993641309440136, + "loss_total_v6": 0.4317341186106205, + "step": 1512 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34894414414414415, + "grad_norm": 25.834749221801758, + "learning_rate": 7.712781958873827e-06, + "loss": 7.191954612731934, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.31583019345998764, + "loss_total_v6": 0.44949717447161674, + "step": 1513 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34917477477477477, + "grad_norm": 24.028766632080078, + "learning_rate": 7.709590897694771e-06, + "loss": 6.746341705322266, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.28743008337914944, + "loss_total_v6": 0.4216463640332222, + "step": 1514 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3494054054054054, + "grad_norm": 23.129121780395508, + "learning_rate": 7.706398273215497e-06, + "loss": 7.1119232177734375, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.31201836839318275, + "loss_total_v6": 0.44449519366025925, + "step": 1515 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34963603603603605, + "grad_norm": 22.526884078979492, + "learning_rate": 7.703204087277989e-06, + "loss": 6.869786262512207, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.29653386026620865, + "loss_total_v6": 0.42936163768172264, + "step": 1516 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.34986666666666666, + "grad_norm": 28.919775009155273, + "learning_rate": 7.700008341725134e-06, + "loss": 7.389809608459473, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.32721956819295883, + "loss_total_v6": 0.46186309307813644, + "step": 1517 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35009729729729727, + "grad_norm": 24.293001174926758, + "learning_rate": 7.696811038400722e-06, + "loss": 7.623594284057617, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3401831388473511, + "loss_total_v6": 0.47647465020418167, + "step": 1518 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35032792792792794, + "grad_norm": 21.776798248291016, + "learning_rate": 7.693612179149437e-06, + "loss": 6.990638732910156, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.30135586112737656, + "loss_total_v6": 0.43691493570804596, + "step": 1519 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35055855855855855, + "grad_norm": 23.352909088134766, + "learning_rate": 7.690411765816864e-06, + "loss": 7.499578475952148, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3347209803760052, + "loss_total_v6": 0.4687236547470093, + "step": 1520 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35078918918918917, + "grad_norm": 23.354097366333008, + "learning_rate": 7.687209800249484e-06, + "loss": 7.264172077178955, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3201606497168541, + "loss_total_v6": 0.4540107436478138, + "step": 1521 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35101981981981983, + "grad_norm": 24.464387893676758, + "learning_rate": 7.684006284294672e-06, + "loss": 6.912086486816406, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.2986130826175213, + "loss_total_v6": 0.4320054166018963, + "step": 1522 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35125045045045045, + "grad_norm": 25.90777587890625, + "learning_rate": 7.680801219800698e-06, + "loss": 6.576109886169434, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2810019701719284, + "loss_total_v6": 0.4110068492591381, + "step": 1523 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35148108108108106, + "grad_norm": 26.192291259765625, + "learning_rate": 7.67759460861673e-06, + "loss": 7.2321343421936035, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.3149692118167877, + "loss_total_v6": 0.4520083963871002, + "step": 1524 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3517117117117117, + "grad_norm": 25.967782974243164, + "learning_rate": 7.674386452592818e-06, + "loss": 6.788328170776367, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.2910307664424181, + "loss_total_v6": 0.42427050322294235, + "step": 1525 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35194234234234234, + "grad_norm": 23.398681640625, + "learning_rate": 7.671176753579914e-06, + "loss": 7.292651653289795, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.3189804293215275, + "loss_total_v6": 0.4557907320559025, + "step": 1526 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35217297297297295, + "grad_norm": 24.23615837097168, + "learning_rate": 7.667965513429856e-06, + "loss": 6.930919647216797, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.301651693880558, + "loss_total_v6": 0.4331824705004692, + "step": 1527 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3524036036036036, + "grad_norm": 27.866226196289062, + "learning_rate": 7.66475273399537e-06, + "loss": 6.90420389175415, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2985476478934288, + "loss_total_v6": 0.4315127357840538, + "step": 1528 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35263423423423423, + "grad_norm": 29.130149841308594, + "learning_rate": 7.661538417130072e-06, + "loss": 7.0470356941223145, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3073067869991064, + "loss_total_v6": 0.44043971225619316, + "step": 1529 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35286486486486485, + "grad_norm": 26.836376190185547, + "learning_rate": 7.658322564688463e-06, + "loss": 7.51796817779541, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.33484797552227974, + "loss_total_v6": 0.46987300366163254, + "step": 1530 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3530954954954955, + "grad_norm": 25.71156120300293, + "learning_rate": 7.655105178525932e-06, + "loss": 6.393678665161133, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.26929484494030476, + "loss_total_v6": 0.3996049128472805, + "step": 1531 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3533261261261261, + "grad_norm": 25.992502212524414, + "learning_rate": 7.651886260498751e-06, + "loss": 7.603147983551025, + "loss_alignment": 0.2689208984375, + "loss_alignment_w": 0.13446044921875, + "loss_sft": 0.3405379541218281, + "loss_total_v6": 0.4751967564225197, + "step": 1532 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35355675675675674, + "grad_norm": 24.847061157226562, + "learning_rate": 7.648665812464079e-06, + "loss": 7.291595458984375, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.32042503356933594, + "loss_total_v6": 0.45572471991181374, + "step": 1533 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3537873873873874, + "grad_norm": 26.547515869140625, + "learning_rate": 7.645443836279951e-06, + "loss": 7.24549674987793, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.3182305209338665, + "loss_total_v6": 0.4528435617685318, + "step": 1534 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.354018018018018, + "grad_norm": 25.922569274902344, + "learning_rate": 7.642220333805294e-06, + "loss": 7.1378068923950195, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.3121560253202915, + "loss_total_v6": 0.4461129307746887, + "step": 1535 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35424864864864863, + "grad_norm": 24.112119674682617, + "learning_rate": 7.638995306899907e-06, + "loss": 6.883272647857666, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.29688848182559013, + "loss_total_v6": 0.4302045367658138, + "step": 1536 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3544792792792793, + "grad_norm": 22.760040283203125, + "learning_rate": 7.635768757424473e-06, + "loss": 7.468358993530273, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.3357757180929184, + "loss_total_v6": 0.4667724296450615, + "step": 1537 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3547099099099099, + "grad_norm": 23.447511672973633, + "learning_rate": 7.632540687240547e-06, + "loss": 7.325991630554199, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.32391755655407906, + "loss_total_v6": 0.45787447690963745, + "step": 1538 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35494054054054053, + "grad_norm": 24.673542022705078, + "learning_rate": 7.62931109821057e-06, + "loss": 7.440149307250977, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.32949599623680115, + "loss_total_v6": 0.46500930562615395, + "step": 1539 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3551711711711712, + "grad_norm": 26.103595733642578, + "learning_rate": 7.626079992197857e-06, + "loss": 7.240694999694824, + "loss_alignment": 0.274658203125, + "loss_alignment_w": 0.1373291015625, + "loss_sft": 0.31515330262482166, + "loss_total_v6": 0.4525434337556362, + "step": 1540 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3554018018018018, + "grad_norm": 21.840686798095703, + "learning_rate": 7.6228473710665955e-06, + "loss": 6.6118855476379395, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.2817883975803852, + "loss_total_v6": 0.4132428541779518, + "step": 1541 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3556324324324324, + "grad_norm": 27.552919387817383, + "learning_rate": 7.619613236681845e-06, + "loss": 6.799681663513184, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.29251856356859207, + "loss_total_v6": 0.42498011887073517, + "step": 1542 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3558630630630631, + "grad_norm": 23.275653839111328, + "learning_rate": 7.616377590909545e-06, + "loss": 6.932024002075195, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3015528842806816, + "loss_total_v6": 0.4332514889538288, + "step": 1543 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3560936936936937, + "grad_norm": 23.767398834228516, + "learning_rate": 7.613140435616503e-06, + "loss": 6.657710075378418, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2837521359324455, + "loss_total_v6": 0.4161068797111511, + "step": 1544 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3563243243243243, + "grad_norm": 27.40993309020996, + "learning_rate": 7.609901772670396e-06, + "loss": 6.845271110534668, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.2946201879531145, + "loss_total_v6": 0.42782942578196526, + "step": 1545 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35655495495495493, + "grad_norm": 27.17775535583496, + "learning_rate": 7.6066616039397755e-06, + "loss": 6.742316722869873, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.28537795692682266, + "loss_total_v6": 0.42139479890465736, + "step": 1546 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3567855855855856, + "grad_norm": 24.838912963867188, + "learning_rate": 7.603419931294056e-06, + "loss": 7.463446617126465, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.33173028007149696, + "loss_total_v6": 0.46646540611982346, + "step": 1547 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3570162162162162, + "grad_norm": 24.071712493896484, + "learning_rate": 7.600176756603524e-06, + "loss": 7.497928619384766, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.33298516273498535, + "loss_total_v6": 0.46862053126096725, + "step": 1548 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3572468468468468, + "grad_norm": 26.665496826171875, + "learning_rate": 7.5969320817393326e-06, + "loss": 7.236353397369385, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.31681980565190315, + "loss_total_v6": 0.45227208361029625, + "step": 1549 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3574774774774775, + "grad_norm": 28.124582290649414, + "learning_rate": 7.593685908573497e-06, + "loss": 6.985260963439941, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.30518538504838943, + "loss_total_v6": 0.43657881394028664, + "step": 1550 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3577081081081081, + "grad_norm": 23.527389526367188, + "learning_rate": 7.590438238978899e-06, + "loss": 7.257225036621094, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.31705617159605026, + "loss_total_v6": 0.45357657968997955, + "step": 1551 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3579387387387387, + "grad_norm": 24.499889373779297, + "learning_rate": 7.587189074829284e-06, + "loss": 7.061783313751221, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3085336834192276, + "loss_total_v6": 0.4413614422082901, + "step": 1552 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3581693693693694, + "grad_norm": 26.252538681030273, + "learning_rate": 7.583938417999261e-06, + "loss": 7.33220100402832, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3241072744131088, + "loss_total_v6": 0.4582625553011894, + "step": 1553 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3584, + "grad_norm": 26.08668327331543, + "learning_rate": 7.580686270364298e-06, + "loss": 7.386902809143066, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.32760243862867355, + "loss_total_v6": 0.46168141067028046, + "step": 1554 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3586306306306306, + "grad_norm": 25.708200454711914, + "learning_rate": 7.577432633800721e-06, + "loss": 6.97161865234375, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3059348873794079, + "loss_total_v6": 0.4357261508703232, + "step": 1555 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3588612612612613, + "grad_norm": 22.177967071533203, + "learning_rate": 7.574177510185721e-06, + "loss": 6.889369010925293, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.2995430901646614, + "loss_total_v6": 0.4305855669081211, + "step": 1556 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3590918918918919, + "grad_norm": 23.483837127685547, + "learning_rate": 7.570920901397342e-06, + "loss": 7.078640937805176, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3063371740281582, + "loss_total_v6": 0.4424150548875332, + "step": 1557 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3593225225225225, + "grad_norm": 22.00381088256836, + "learning_rate": 7.567662809314485e-06, + "loss": 7.287054061889648, + "loss_alignment": 0.2657470703125, + "loss_alignment_w": 0.13287353515625, + "loss_sft": 0.3226283937692642, + "loss_total_v6": 0.45544087514281273, + "step": 1558 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.35955315315315317, + "grad_norm": 25.315303802490234, + "learning_rate": 7.564403235816913e-06, + "loss": 6.908418655395508, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.29890259355306625, + "loss_total_v6": 0.43177614361047745, + "step": 1559 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3597837837837838, + "grad_norm": 24.064790725708008, + "learning_rate": 7.561142182785234e-06, + "loss": 7.359644412994385, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.32368626818060875, + "loss_total_v6": 0.45997776836156845, + "step": 1560 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3600144144144144, + "grad_norm": 23.391983032226562, + "learning_rate": 7.557879652100917e-06, + "loss": 7.460684776306152, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.33105412125587463, + "loss_total_v6": 0.46629277244210243, + "step": 1561 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36024504504504506, + "grad_norm": 23.69780731201172, + "learning_rate": 7.554615645646282e-06, + "loss": 7.038196563720703, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.30657123029232025, + "loss_total_v6": 0.43988727033138275, + "step": 1562 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3604756756756757, + "grad_norm": 24.299896240234375, + "learning_rate": 7.5513501653045e-06, + "loss": 6.8975324630737305, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.29776449128985405, + "loss_total_v6": 0.43109579384326935, + "step": 1563 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3607063063063063, + "grad_norm": 24.700037002563477, + "learning_rate": 7.548083212959588e-06, + "loss": 7.1564130783081055, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.31351728551089764, + "loss_total_v6": 0.4472758322954178, + "step": 1564 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36093693693693696, + "grad_norm": 26.456724166870117, + "learning_rate": 7.5448147904964196e-06, + "loss": 7.044134140014648, + "loss_alignment": 0.2652587890625, + "loss_alignment_w": 0.13262939453125, + "loss_sft": 0.307461142539978, + "loss_total_v6": 0.44025838002562523, + "step": 1565 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36116756756756757, + "grad_norm": 27.585613250732422, + "learning_rate": 7.541544899800712e-06, + "loss": 7.167885780334473, + "loss_alignment": 0.2669677734375, + "loss_alignment_w": 0.13348388671875, + "loss_sft": 0.31434111669659615, + "loss_total_v6": 0.44799283891916275, + "step": 1566 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3613981981981982, + "grad_norm": 22.705455780029297, + "learning_rate": 7.538273542759033e-06, + "loss": 6.637203693389893, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.28186013363301754, + "loss_total_v6": 0.4148252308368683, + "step": 1567 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36162882882882885, + "grad_norm": 23.660053253173828, + "learning_rate": 7.535000721258792e-06, + "loss": 7.127774238586426, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.31300910003483295, + "loss_total_v6": 0.4454858973622322, + "step": 1568 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36185945945945946, + "grad_norm": 24.991561889648438, + "learning_rate": 7.531726437188247e-06, + "loss": 7.222721099853516, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3160593621432781, + "loss_total_v6": 0.4514200873672962, + "step": 1569 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3620900900900901, + "grad_norm": 24.121477127075195, + "learning_rate": 7.528450692436495e-06, + "loss": 6.847520351409912, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.2928686812520027, + "loss_total_v6": 0.4279700145125389, + "step": 1570 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36232072072072075, + "grad_norm": 22.687227249145508, + "learning_rate": 7.525173488893485e-06, + "loss": 7.047562122344971, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3092012479901314, + "loss_total_v6": 0.4404726214706898, + "step": 1571 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36255135135135136, + "grad_norm": 22.889286041259766, + "learning_rate": 7.521894828449994e-06, + "loss": 7.28511905670166, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3198371194303036, + "loss_total_v6": 0.45531991869211197, + "step": 1572 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36278198198198197, + "grad_norm": 26.76798439025879, + "learning_rate": 7.518614712997652e-06, + "loss": 6.774439811706543, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2933518309146166, + "loss_total_v6": 0.42340249195694923, + "step": 1573 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36301261261261264, + "grad_norm": 26.73969841003418, + "learning_rate": 7.515333144428925e-06, + "loss": 7.0179266929626465, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.30448037944734097, + "loss_total_v6": 0.4386203996837139, + "step": 1574 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36324324324324325, + "grad_norm": 23.511991500854492, + "learning_rate": 7.512050124637114e-06, + "loss": 6.8742356300354, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.2956523075699806, + "loss_total_v6": 0.4296397343277931, + "step": 1575 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36347387387387387, + "grad_norm": 26.725414276123047, + "learning_rate": 7.508765655516358e-06, + "loss": 7.070833206176758, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.30601707100868225, + "loss_total_v6": 0.44192709773778915, + "step": 1576 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3637045045045045, + "grad_norm": 25.057573318481445, + "learning_rate": 7.505479738961635e-06, + "loss": 6.890650272369385, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.2939621414989233, + "loss_total_v6": 0.43066563829779625, + "step": 1577 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36393513513513515, + "grad_norm": 24.07012367248535, + "learning_rate": 7.502192376868758e-06, + "loss": 6.80596923828125, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.2914314176887274, + "loss_total_v6": 0.4253730736672878, + "step": 1578 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36416576576576576, + "grad_norm": 22.764190673828125, + "learning_rate": 7.498903571134371e-06, + "loss": 7.017232894897461, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.3042233996093273, + "loss_total_v6": 0.4385770484805107, + "step": 1579 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36439639639639637, + "grad_norm": 26.499406814575195, + "learning_rate": 7.4956133236559525e-06, + "loss": 6.882020950317383, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.2960778381675482, + "loss_total_v6": 0.4301263056695461, + "step": 1580 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36462702702702704, + "grad_norm": 25.831472396850586, + "learning_rate": 7.492321636331814e-06, + "loss": 6.801468372344971, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.29002100601792336, + "loss_total_v6": 0.42509179562330246, + "step": 1581 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36485765765765765, + "grad_norm": 26.76854133605957, + "learning_rate": 7.489028511061097e-06, + "loss": 7.060447692871094, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3055663350969553, + "loss_total_v6": 0.44127798452973366, + "step": 1582 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36508828828828827, + "grad_norm": 25.182899475097656, + "learning_rate": 7.485733949743771e-06, + "loss": 6.728475093841553, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.28718313947319984, + "loss_total_v6": 0.42052969336509705, + "step": 1583 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36531891891891893, + "grad_norm": 23.59071922302246, + "learning_rate": 7.4824379542806355e-06, + "loss": 7.370427131652832, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3255045935511589, + "loss_total_v6": 0.4606516808271408, + "step": 1584 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36554954954954955, + "grad_norm": 22.531545639038086, + "learning_rate": 7.479140526573317e-06, + "loss": 7.192407131195068, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.31688082218170166, + "loss_total_v6": 0.44952545315027237, + "step": 1585 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36578018018018016, + "grad_norm": 23.50129508972168, + "learning_rate": 7.475841668524268e-06, + "loss": 7.167594909667969, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.31479597836732864, + "loss_total_v6": 0.44797468930482864, + "step": 1586 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3660108108108108, + "grad_norm": 25.646936416625977, + "learning_rate": 7.47254138203677e-06, + "loss": 7.140399932861328, + "loss_alignment": 0.2677001953125, + "loss_alignment_w": 0.13385009765625, + "loss_sft": 0.3125317133963108, + "loss_total_v6": 0.4462749846279621, + "step": 1587 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36624144144144144, + "grad_norm": 24.57822608947754, + "learning_rate": 7.469239669014922e-06, + "loss": 7.2894062995910645, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.320059334859252, + "loss_total_v6": 0.45558788999915123, + "step": 1588 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36647207207207205, + "grad_norm": 21.768648147583008, + "learning_rate": 7.465936531363651e-06, + "loss": 6.977721214294434, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3018607459962368, + "loss_total_v6": 0.4361075721681118, + "step": 1589 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3667027027027027, + "grad_norm": 21.376741409301758, + "learning_rate": 7.462631970988703e-06, + "loss": 7.495367050170898, + "loss_alignment": 0.276611328125, + "loss_alignment_w": 0.1383056640625, + "loss_sft": 0.33026161417365074, + "loss_total_v6": 0.46846046298742294, + "step": 1590 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36693333333333333, + "grad_norm": 26.320314407348633, + "learning_rate": 7.459325989796647e-06, + "loss": 6.880151748657227, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.2965713646262884, + "loss_total_v6": 0.43000947311520576, + "step": 1591 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36716396396396395, + "grad_norm": 39.90589141845703, + "learning_rate": 7.4560185896948735e-06, + "loss": 6.866608619689941, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.29310040920972824, + "loss_total_v6": 0.42916303500533104, + "step": 1592 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3673945945945946, + "grad_norm": 29.211647033691406, + "learning_rate": 7.452709772591586e-06, + "loss": 6.676787376403809, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.2839221339672804, + "loss_total_v6": 0.41729920729994774, + "step": 1593 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3676252252252252, + "grad_norm": 25.55448341369629, + "learning_rate": 7.449399540395807e-06, + "loss": 6.921975135803223, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.300299234688282, + "loss_total_v6": 0.4326234422624111, + "step": 1594 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36785585585585584, + "grad_norm": 27.32575035095215, + "learning_rate": 7.446087895017382e-06, + "loss": 7.532894134521484, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.33683374151587486, + "loss_total_v6": 0.47080590203404427, + "step": 1595 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3680864864864865, + "grad_norm": 26.847837448120117, + "learning_rate": 7.442774838366965e-06, + "loss": 7.245822906494141, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3173201195895672, + "loss_total_v6": 0.4528639391064644, + "step": 1596 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3683171171171171, + "grad_norm": 23.23828887939453, + "learning_rate": 7.439460372356025e-06, + "loss": 6.802006244659424, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.2928621992468834, + "loss_total_v6": 0.4251253791153431, + "step": 1597 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.36854774774774773, + "grad_norm": 24.268301010131836, + "learning_rate": 7.436144498896845e-06, + "loss": 7.011897087097168, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.30219620652496815, + "loss_total_v6": 0.4382435642182827, + "step": 1598 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3687783783783784, + "grad_norm": 21.20546531677246, + "learning_rate": 7.432827219902522e-06, + "loss": 7.076081275939941, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3073521628975868, + "loss_total_v6": 0.442255113273859, + "step": 1599 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.369009009009009, + "grad_norm": 25.22710418701172, + "learning_rate": 7.429508537286963e-06, + "loss": 6.841528415679932, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.29318084567785263, + "loss_total_v6": 0.42759552225470543, + "step": 1600 + }, + { + "epoch": 0.369009009009009, + "eval_loss": 0.5984223484992981, + "eval_loss_alignment": 0.2408073808504566, + "eval_loss_alignment_w": 0.1204036904252283, + "eval_loss_sft": 0.47801861385538424, + "eval_loss_total_v6": 0.5984223045612851, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.0266, + "eval_samples_per_second": 19.17, + "eval_steps_per_second": 2.406, + "step": 1600 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3692396396396396, + "grad_norm": 24.44710922241211, + "learning_rate": 7.426188452964882e-06, + "loss": 6.994662761688232, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3054983504116535, + "loss_total_v6": 0.4371664263308048, + "step": 1601 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3694702702702703, + "grad_norm": 25.330665588378906, + "learning_rate": 7.422866968851805e-06, + "loss": 7.402541637420654, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.32801528647542, + "loss_total_v6": 0.4626588597893715, + "step": 1602 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3697009009009009, + "grad_norm": 22.996944427490234, + "learning_rate": 7.4195440868640624e-06, + "loss": 6.936790466308594, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.30066060088574886, + "loss_total_v6": 0.4335494004189968, + "step": 1603 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3699315315315315, + "grad_norm": 21.7581729888916, + "learning_rate": 7.4162198089187945e-06, + "loss": 7.436119556427002, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.330342810600996, + "loss_total_v6": 0.4647574685513973, + "step": 1604 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3701621621621622, + "grad_norm": 24.964797973632812, + "learning_rate": 7.412894136933947e-06, + "loss": 6.882411003112793, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.29460686817765236, + "loss_total_v6": 0.43015068769454956, + "step": 1605 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3703927927927928, + "grad_norm": 24.557342529296875, + "learning_rate": 7.409567072828268e-06, + "loss": 6.7921037673950195, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.294959357008338, + "loss_total_v6": 0.4245064966380596, + "step": 1606 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3706234234234234, + "grad_norm": 27.392345428466797, + "learning_rate": 7.406238618521306e-06, + "loss": 7.195457935333252, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.31864311918616295, + "loss_total_v6": 0.44971611723303795, + "step": 1607 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.370854054054054, + "grad_norm": 26.167884826660156, + "learning_rate": 7.402908775933419e-06, + "loss": 6.692355632781982, + "loss_alignment": 0.2667236328125, + "loss_alignment_w": 0.13336181640625, + "loss_sft": 0.28497146628797054, + "loss_total_v6": 0.4182722382247448, + "step": 1608 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3710846846846847, + "grad_norm": 24.6630802154541, + "learning_rate": 7.399577546985761e-06, + "loss": 7.397016525268555, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.32657136023044586, + "loss_total_v6": 0.46231354027986526, + "step": 1609 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3713153153153153, + "grad_norm": 24.60287094116211, + "learning_rate": 7.396244933600285e-06, + "loss": 7.414942741394043, + "loss_alignment": 0.27685546875, + "loss_alignment_w": 0.138427734375, + "loss_sft": 0.3249604068696499, + "loss_total_v6": 0.463433925062418, + "step": 1610 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3715459459459459, + "grad_norm": 23.66234016418457, + "learning_rate": 7.392910937699745e-06, + "loss": 7.581403732299805, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3418644592165947, + "loss_total_v6": 0.4738377146422863, + "step": 1611 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3717765765765766, + "grad_norm": 24.04705810546875, + "learning_rate": 7.389575561207693e-06, + "loss": 7.255360126495361, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.31960989721119404, + "loss_total_v6": 0.4534599930047989, + "step": 1612 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3720072072072072, + "grad_norm": 25.415616989135742, + "learning_rate": 7.386238806048476e-06, + "loss": 6.71225643157959, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.28447575867176056, + "loss_total_v6": 0.41951602697372437, + "step": 1613 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3722378378378378, + "grad_norm": 24.384681701660156, + "learning_rate": 7.382900674147238e-06, + "loss": 6.849094390869141, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.2966596968472004, + "loss_total_v6": 0.4280683882534504, + "step": 1614 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3724684684684685, + "grad_norm": 24.839962005615234, + "learning_rate": 7.379561167429915e-06, + "loss": 7.114594459533691, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3102779649198055, + "loss_total_v6": 0.4446621313691139, + "step": 1615 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3726990990990991, + "grad_norm": 30.1224308013916, + "learning_rate": 7.376220287823237e-06, + "loss": 7.0881028175354, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.31167399510741234, + "loss_total_v6": 0.44300641119480133, + "step": 1616 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3729297297297297, + "grad_norm": 22.487058639526367, + "learning_rate": 7.372878037254728e-06, + "loss": 7.247948169708252, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.31777336820960045, + "loss_total_v6": 0.45299674570560455, + "step": 1617 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3731603603603604, + "grad_norm": 24.38092803955078, + "learning_rate": 7.369534417652702e-06, + "loss": 7.1446027755737305, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.31277911737561226, + "loss_total_v6": 0.44653765484690666, + "step": 1618 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.373390990990991, + "grad_norm": 25.77460289001465, + "learning_rate": 7.366189430946262e-06, + "loss": 7.227968215942383, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3211022801697254, + "loss_total_v6": 0.4517480283975601, + "step": 1619 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3736216216216216, + "grad_norm": 24.753341674804688, + "learning_rate": 7.3628430790653e-06, + "loss": 7.118670463562012, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.31082264706492424, + "loss_total_v6": 0.44491690397262573, + "step": 1620 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37385225225225227, + "grad_norm": 23.003623962402344, + "learning_rate": 7.359495363940495e-06, + "loss": 6.736838340759277, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.29031508043408394, + "loss_total_v6": 0.42105237767100334, + "step": 1621 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3740828828828829, + "grad_norm": 24.352985382080078, + "learning_rate": 7.356146287503316e-06, + "loss": 6.962278842926025, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3028182052075863, + "loss_total_v6": 0.4351424239575863, + "step": 1622 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3743135135135135, + "grad_norm": 23.98801612854004, + "learning_rate": 7.3527958516860124e-06, + "loss": 6.558675765991211, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.2810109592974186, + "loss_total_v6": 0.4099172204732895, + "step": 1623 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37454414414414416, + "grad_norm": 22.895023345947266, + "learning_rate": 7.34944405842162e-06, + "loss": 7.220295429229736, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3170369006693363, + "loss_total_v6": 0.4512684680521488, + "step": 1624 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3747747747747748, + "grad_norm": 23.501829147338867, + "learning_rate": 7.346090909643959e-06, + "loss": 7.560110092163086, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.33839739486575127, + "loss_total_v6": 0.47250688821077347, + "step": 1625 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3750054054054054, + "grad_norm": 21.40238380432129, + "learning_rate": 7.3427364072876325e-06, + "loss": 6.918224334716797, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.29916456528007984, + "loss_total_v6": 0.4323890432715416, + "step": 1626 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37523603603603606, + "grad_norm": 24.89432716369629, + "learning_rate": 7.339380553288021e-06, + "loss": 7.257161617279053, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3212483935058117, + "loss_total_v6": 0.4535726010799408, + "step": 1627 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37546666666666667, + "grad_norm": 24.39481544494629, + "learning_rate": 7.336023349581287e-06, + "loss": 7.209973335266113, + "loss_alignment": 0.2672119140625, + "loss_alignment_w": 0.13360595703125, + "loss_sft": 0.31645284965634346, + "loss_total_v6": 0.450623344630003, + "step": 1628 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3756972972972973, + "grad_norm": 22.003332138061523, + "learning_rate": 7.332664798104371e-06, + "loss": 7.152076721191406, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3135209120810032, + "loss_total_v6": 0.4470047913491726, + "step": 1629 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37592792792792795, + "grad_norm": 23.611196517944336, + "learning_rate": 7.329304900794991e-06, + "loss": 7.082262992858887, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3073875196278095, + "loss_total_v6": 0.44264141470193863, + "step": 1630 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37615855855855856, + "grad_norm": 22.983816146850586, + "learning_rate": 7.325943659591644e-06, + "loss": 7.270209789276123, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3222012184560299, + "loss_total_v6": 0.4543880969285965, + "step": 1631 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3763891891891892, + "grad_norm": 24.32892417907715, + "learning_rate": 7.3225810764335965e-06, + "loss": 7.018994331359863, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.30500489473342896, + "loss_total_v6": 0.43868714943528175, + "step": 1632 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37661981981981985, + "grad_norm": 24.555221557617188, + "learning_rate": 7.319217153260897e-06, + "loss": 6.8768510818481445, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.2964413855224848, + "loss_total_v6": 0.42980320379137993, + "step": 1633 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37685045045045046, + "grad_norm": 23.214609146118164, + "learning_rate": 7.3158518920143615e-06, + "loss": 6.581235885620117, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.2772177215665579, + "loss_total_v6": 0.41132723167538643, + "step": 1634 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37708108108108107, + "grad_norm": 24.719396591186523, + "learning_rate": 7.312485294635581e-06, + "loss": 7.369567394256592, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3241538666188717, + "loss_total_v6": 0.4605979584157467, + "step": 1635 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37731171171171174, + "grad_norm": 24.985002517700195, + "learning_rate": 7.3091173630669135e-06, + "loss": 7.253429412841797, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.318985678255558, + "loss_total_v6": 0.4533393308520317, + "step": 1636 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37754234234234235, + "grad_norm": 23.811677932739258, + "learning_rate": 7.3057480992514895e-06, + "loss": 6.794306755065918, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.29317446425557137, + "loss_total_v6": 0.42464418709278107, + "step": 1637 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37777297297297296, + "grad_norm": 21.806493759155273, + "learning_rate": 7.302377505133209e-06, + "loss": 6.88950777053833, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.2958286181092262, + "loss_total_v6": 0.43059423938393593, + "step": 1638 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3780036036036036, + "grad_norm": 24.09798812866211, + "learning_rate": 7.299005582656739e-06, + "loss": 7.635630130767822, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3420187532901764, + "loss_total_v6": 0.477226871997118, + "step": 1639 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37823423423423425, + "grad_norm": 23.788063049316406, + "learning_rate": 7.295632333767513e-06, + "loss": 6.785043716430664, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.29044403694570065, + "loss_total_v6": 0.4240652322769165, + "step": 1640 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37846486486486486, + "grad_norm": 25.925100326538086, + "learning_rate": 7.292257760411726e-06, + "loss": 6.777517318725586, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.28855456970632076, + "loss_total_v6": 0.42359484359622, + "step": 1641 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37869549549549547, + "grad_norm": 31.519392013549805, + "learning_rate": 7.288881864536344e-06, + "loss": 7.196380615234375, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.31696129217743874, + "loss_total_v6": 0.44977377727627754, + "step": 1642 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37892612612612614, + "grad_norm": 23.692386627197266, + "learning_rate": 7.28550464808909e-06, + "loss": 7.083457946777344, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3083777502179146, + "loss_total_v6": 0.4427161104977131, + "step": 1643 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37915675675675675, + "grad_norm": 23.89344596862793, + "learning_rate": 7.28212611301845e-06, + "loss": 7.052088737487793, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.3052422497421503, + "loss_total_v6": 0.44075554236769676, + "step": 1644 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37938738738738736, + "grad_norm": 23.236038208007812, + "learning_rate": 7.278746261273676e-06, + "loss": 6.553256988525391, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.27484345249831676, + "loss_total_v6": 0.4095785468816757, + "step": 1645 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37961801801801803, + "grad_norm": 24.124799728393555, + "learning_rate": 7.2753650948047716e-06, + "loss": 6.656579971313477, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.28046192042529583, + "loss_total_v6": 0.4160362556576729, + "step": 1646 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.37984864864864865, + "grad_norm": 25.880733489990234, + "learning_rate": 7.271982615562506e-06, + "loss": 6.965943336486816, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.30204015225172043, + "loss_total_v6": 0.43537144735455513, + "step": 1647 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38007927927927926, + "grad_norm": 23.297290802001953, + "learning_rate": 7.2685988254984006e-06, + "loss": 7.217016220092773, + "loss_alignment": 0.2645263671875, + "loss_alignment_w": 0.13226318359375, + "loss_sft": 0.31863248348236084, + "loss_total_v6": 0.45106352120637894, + "step": 1648 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3803099099099099, + "grad_norm": 25.414777755737305, + "learning_rate": 7.265213726564737e-06, + "loss": 7.308586597442627, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3233027718961239, + "loss_total_v6": 0.4567866660654545, + "step": 1649 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38054054054054054, + "grad_norm": 24.632003784179688, + "learning_rate": 7.26182732071455e-06, + "loss": 7.306707382202148, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.3239482566714287, + "loss_total_v6": 0.456669207662344, + "step": 1650 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38077117117117115, + "grad_norm": 22.14801597595215, + "learning_rate": 7.258439609901625e-06, + "loss": 6.828631401062012, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.29330558702349663, + "loss_total_v6": 0.42678947001695633, + "step": 1651 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3810018018018018, + "grad_norm": 25.687585830688477, + "learning_rate": 7.25505059608051e-06, + "loss": 7.258368492126465, + "loss_alignment": 0.275390625, + "loss_alignment_w": 0.1376953125, + "loss_sft": 0.3159069400280714, + "loss_total_v6": 0.45364802330732346, + "step": 1652 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38123243243243243, + "grad_norm": 27.908855438232422, + "learning_rate": 7.251660281206495e-06, + "loss": 7.222290992736816, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.3166885878890753, + "loss_total_v6": 0.45139317959547043, + "step": 1653 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38146306306306305, + "grad_norm": 22.19097900390625, + "learning_rate": 7.248268667235624e-06, + "loss": 6.724129676818848, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.28735403530299664, + "loss_total_v6": 0.4202580861747265, + "step": 1654 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3816936936936937, + "grad_norm": 26.110164642333984, + "learning_rate": 7.244875756124693e-06, + "loss": 7.072266578674316, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.30865482054650784, + "loss_total_v6": 0.4420166425406933, + "step": 1655 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3819243243243243, + "grad_norm": 25.331321716308594, + "learning_rate": 7.241481549831244e-06, + "loss": 7.032713890075684, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3064574636518955, + "loss_total_v6": 0.4395446181297302, + "step": 1656 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38215495495495494, + "grad_norm": 23.62612533569336, + "learning_rate": 7.238086050313563e-06, + "loss": 6.587470054626465, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.27980465441942215, + "loss_total_v6": 0.41171689331531525, + "step": 1657 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3823855855855856, + "grad_norm": 24.34870147705078, + "learning_rate": 7.23468925953069e-06, + "loss": 7.165963172912598, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.31344279274344444, + "loss_total_v6": 0.44787271320819855, + "step": 1658 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3826162162162162, + "grad_norm": 22.704734802246094, + "learning_rate": 7.231291179442404e-06, + "loss": 6.535958766937256, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.27751599438488483, + "loss_total_v6": 0.40849744156003, + "step": 1659 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38284684684684683, + "grad_norm": 23.924026489257812, + "learning_rate": 7.22789181200923e-06, + "loss": 7.060744762420654, + "loss_alignment": 0.2657470703125, + "loss_alignment_w": 0.13287353515625, + "loss_sft": 0.3083619624376297, + "loss_total_v6": 0.4412965551018715, + "step": 1660 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3830774774774775, + "grad_norm": 25.19426155090332, + "learning_rate": 7.224491159192434e-06, + "loss": 7.446375846862793, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.33112113177776337, + "loss_total_v6": 0.46539847925305367, + "step": 1661 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3833081081081081, + "grad_norm": 27.30877113342285, + "learning_rate": 7.221089222954026e-06, + "loss": 7.001918792724609, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3048684485256672, + "loss_total_v6": 0.4376199170947075, + "step": 1662 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3835387387387387, + "grad_norm": 25.777969360351562, + "learning_rate": 7.217686005256755e-06, + "loss": 7.118655204772949, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3099519480019808, + "loss_total_v6": 0.44491593539714813, + "step": 1663 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3837693693693694, + "grad_norm": 21.72647476196289, + "learning_rate": 7.214281508064107e-06, + "loss": 7.203842639923096, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3161764293909073, + "loss_total_v6": 0.4502401612699032, + "step": 1664 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.384, + "grad_norm": 24.71710205078125, + "learning_rate": 7.210875733340315e-06, + "loss": 6.92830228805542, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3033497165888548, + "loss_total_v6": 0.43301890790462494, + "step": 1665 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3842306306306306, + "grad_norm": 23.789350509643555, + "learning_rate": 7.207468683050338e-06, + "loss": 7.305902481079102, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.32534755393862724, + "loss_total_v6": 0.45661891251802444, + "step": 1666 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3844612612612613, + "grad_norm": 25.033092498779297, + "learning_rate": 7.204060359159877e-06, + "loss": 6.888477325439453, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2993347570300102, + "loss_total_v6": 0.4305298216640949, + "step": 1667 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3846918918918919, + "grad_norm": 23.17169761657715, + "learning_rate": 7.200650763635366e-06, + "loss": 7.157124042510986, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3120968732982874, + "loss_total_v6": 0.44732025265693665, + "step": 1668 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3849225225225225, + "grad_norm": 27.179718017578125, + "learning_rate": 7.1972398984439765e-06, + "loss": 7.172853946685791, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.31283584609627724, + "loss_total_v6": 0.44830336794257164, + "step": 1669 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3851531531531531, + "grad_norm": 22.722623825073242, + "learning_rate": 7.193827765553605e-06, + "loss": 7.274972438812256, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3201490305364132, + "loss_total_v6": 0.4546857699751854, + "step": 1670 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3853837837837838, + "grad_norm": 22.094600677490234, + "learning_rate": 7.190414366932886e-06, + "loss": 6.9654340744018555, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.29843781143426895, + "loss_total_v6": 0.43533965572714806, + "step": 1671 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3856144144144144, + "grad_norm": 22.912424087524414, + "learning_rate": 7.186999704551182e-06, + "loss": 7.038025379180908, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3078117538243532, + "loss_total_v6": 0.43987657874822617, + "step": 1672 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.385845045045045, + "grad_norm": 23.40143394470215, + "learning_rate": 7.183583780378585e-06, + "loss": 7.099645614624023, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3120902646332979, + "loss_total_v6": 0.4437278285622597, + "step": 1673 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3860756756756757, + "grad_norm": 25.74488639831543, + "learning_rate": 7.180166596385915e-06, + "loss": 7.203403949737549, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3144553005695343, + "loss_total_v6": 0.4502127431333065, + "step": 1674 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3863063063063063, + "grad_norm": 29.117341995239258, + "learning_rate": 7.176748154544717e-06, + "loss": 7.097412586212158, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3105316609144211, + "loss_total_v6": 0.4435882978141308, + "step": 1675 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3865369369369369, + "grad_norm": 22.900110244750977, + "learning_rate": 7.1733284568272635e-06, + "loss": 6.836758613586426, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.2960260082036257, + "loss_total_v6": 0.4272973909974098, + "step": 1676 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3867675675675676, + "grad_norm": 22.790483474731445, + "learning_rate": 7.169907505206552e-06, + "loss": 7.415060997009277, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.33044572174549103, + "loss_total_v6": 0.46344131603837013, + "step": 1677 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3869981981981982, + "grad_norm": 23.657264709472656, + "learning_rate": 7.166485301656302e-06, + "loss": 7.292551040649414, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.32416211627423763, + "loss_total_v6": 0.4557844251394272, + "step": 1678 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3872288288288288, + "grad_norm": 23.303791046142578, + "learning_rate": 7.163061848150959e-06, + "loss": 7.534794330596924, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.33788327872753143, + "loss_total_v6": 0.47092464938759804, + "step": 1679 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3874594594594595, + "grad_norm": 25.620651245117188, + "learning_rate": 7.159637146665681e-06, + "loss": 6.855310440063477, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2957054190337658, + "loss_total_v6": 0.4284568913280964, + "step": 1680 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3876900900900901, + "grad_norm": 21.87583351135254, + "learning_rate": 7.156211199176357e-06, + "loss": 6.929945945739746, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.29728787019848824, + "loss_total_v6": 0.43312160670757294, + "step": 1681 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3879207207207207, + "grad_norm": 23.962182998657227, + "learning_rate": 7.1527840076595876e-06, + "loss": 6.985899925231934, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.30617133155465126, + "loss_total_v6": 0.43661872297525406, + "step": 1682 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38815135135135137, + "grad_norm": 24.506662368774414, + "learning_rate": 7.149355574092691e-06, + "loss": 7.022454261779785, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.30538900941610336, + "loss_total_v6": 0.43890340626239777, + "step": 1683 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.388381981981982, + "grad_norm": 24.10947036743164, + "learning_rate": 7.14592590045371e-06, + "loss": 7.045566558837891, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.30704711005091667, + "loss_total_v6": 0.44034790247678757, + "step": 1684 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3886126126126126, + "grad_norm": 24.462194442749023, + "learning_rate": 7.14249498872139e-06, + "loss": 6.90922737121582, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.2970000244677067, + "loss_total_v6": 0.4318266846239567, + "step": 1685 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38884324324324326, + "grad_norm": 27.027809143066406, + "learning_rate": 7.139062840875204e-06, + "loss": 6.785226821899414, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.2906690984964371, + "loss_total_v6": 0.4240766912698746, + "step": 1686 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3890738738738739, + "grad_norm": 23.604530334472656, + "learning_rate": 7.135629458895328e-06, + "loss": 7.349968910217285, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3264537528157234, + "loss_total_v6": 0.4593730755150318, + "step": 1687 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3893045045045045, + "grad_norm": 22.999404907226562, + "learning_rate": 7.1321948447626545e-06, + "loss": 7.102720737457275, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.3079795092344284, + "loss_total_v6": 0.44392004981637, + "step": 1688 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38953513513513516, + "grad_norm": 26.436092376708984, + "learning_rate": 7.128759000458785e-06, + "loss": 7.256513595581055, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.31777462363243103, + "loss_total_v6": 0.453532088547945, + "step": 1689 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.38976576576576577, + "grad_norm": 26.14457893371582, + "learning_rate": 7.125321927966037e-06, + "loss": 6.998250961303711, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.30407464876770973, + "loss_total_v6": 0.43739067390561104, + "step": 1690 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3899963963963964, + "grad_norm": 33.475154876708984, + "learning_rate": 7.121883629267428e-06, + "loss": 6.583884239196777, + "loss_alignment": 0.2667236328125, + "loss_alignment_w": 0.13336181640625, + "loss_sft": 0.2784208618104458, + "loss_total_v6": 0.4114927686750889, + "step": 1691 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39022702702702705, + "grad_norm": 24.662471771240234, + "learning_rate": 7.118444106346687e-06, + "loss": 7.215078353881836, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.31838930398225784, + "loss_total_v6": 0.45094239711761475, + "step": 1692 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39045765765765766, + "grad_norm": 22.5034122467041, + "learning_rate": 7.115003361188248e-06, + "loss": 7.020051956176758, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3031941819936037, + "loss_total_v6": 0.43875324726104736, + "step": 1693 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3906882882882883, + "grad_norm": 24.0106143951416, + "learning_rate": 7.111561395777255e-06, + "loss": 7.540870189666748, + "loss_alignment": 0.277587890625, + "loss_alignment_w": 0.1387939453125, + "loss_sft": 0.33266303315758705, + "loss_total_v6": 0.47130439057946205, + "step": 1694 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39091891891891895, + "grad_norm": 25.94040870666504, + "learning_rate": 7.1081182120995486e-06, + "loss": 7.139913558959961, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3126233760267496, + "loss_total_v6": 0.44624459370970726, + "step": 1695 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39114954954954956, + "grad_norm": 25.534353256225586, + "learning_rate": 7.104673812141676e-06, + "loss": 7.1833953857421875, + "loss_alignment": 0.2696533203125, + "loss_alignment_w": 0.13482666015625, + "loss_sft": 0.314089747145772, + "loss_total_v6": 0.44896218925714493, + "step": 1696 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39138018018018017, + "grad_norm": 24.12320327758789, + "learning_rate": 7.101228197890887e-06, + "loss": 7.294053077697754, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3227454051375389, + "loss_total_v6": 0.4558783359825611, + "step": 1697 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3916108108108108, + "grad_norm": 23.977083206176758, + "learning_rate": 7.097781371335131e-06, + "loss": 6.738323211669922, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.29114027321338654, + "loss_total_v6": 0.4211451783776283, + "step": 1698 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39184144144144145, + "grad_norm": 26.070892333984375, + "learning_rate": 7.094333334463056e-06, + "loss": 7.234188556671143, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3173254132270813, + "loss_total_v6": 0.4521368071436882, + "step": 1699 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39207207207207206, + "grad_norm": 24.274581909179688, + "learning_rate": 7.090884089264012e-06, + "loss": 6.886474609375, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.2946624755859375, + "loss_total_v6": 0.4304046593606472, + "step": 1700 + }, + { + "epoch": 0.39207207207207206, + "eval_loss": 0.5947983264923096, + "eval_loss_alignment": 0.24087426869292236, + "eval_loss_alignment_w": 0.12043713434646118, + "eval_loss_sft": 0.47436123860140916, + "eval_loss_total_v6": 0.5947983736112782, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.3035, + "eval_samples_per_second": 19.112, + "eval_steps_per_second": 2.399, + "step": 1700 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3923027027027027, + "grad_norm": 22.072107315063477, + "learning_rate": 7.087433637728038e-06, + "loss": 7.231792449951172, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.31539032980799675, + "loss_total_v6": 0.45198700949549675, + "step": 1701 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39253333333333335, + "grad_norm": 23.98543357849121, + "learning_rate": 7.083981981845878e-06, + "loss": 7.451864242553711, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3319677412509918, + "loss_total_v6": 0.4657415300607681, + "step": 1702 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39276396396396396, + "grad_norm": 24.05841636657715, + "learning_rate": 7.080529123608967e-06, + "loss": 7.419858932495117, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3273276127874851, + "loss_total_v6": 0.4637411907315254, + "step": 1703 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39299459459459457, + "grad_norm": 29.439851760864258, + "learning_rate": 7.0770750650094335e-06, + "loss": 7.404600143432617, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.32876959815621376, + "loss_total_v6": 0.46278753131628036, + "step": 1704 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39322522522522524, + "grad_norm": 26.821186065673828, + "learning_rate": 7.073619808040098e-06, + "loss": 7.077065944671631, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3088632579892874, + "loss_total_v6": 0.4423166364431381, + "step": 1705 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39345585585585585, + "grad_norm": 25.48453140258789, + "learning_rate": 7.0701633546944725e-06, + "loss": 7.009154796600342, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.30503078550100327, + "loss_total_v6": 0.43807217106223106, + "step": 1706 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39368648648648646, + "grad_norm": 26.322250366210938, + "learning_rate": 7.0667057069667625e-06, + "loss": 6.711942195892334, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.28558527305722237, + "loss_total_v6": 0.41949639841914177, + "step": 1707 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39391711711711713, + "grad_norm": 22.63037109375, + "learning_rate": 7.0632468668518584e-06, + "loss": 7.496121883392334, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3342760466039181, + "loss_total_v6": 0.4685076177120209, + "step": 1708 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39414774774774775, + "grad_norm": 23.10558319091797, + "learning_rate": 7.059786836345339e-06, + "loss": 6.7886199951171875, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.2896299436688423, + "loss_total_v6": 0.4242887496948242, + "step": 1709 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39437837837837836, + "grad_norm": 24.07335090637207, + "learning_rate": 7.056325617443475e-06, + "loss": 7.271568298339844, + "loss_alignment": 0.2645263671875, + "loss_alignment_w": 0.13226318359375, + "loss_sft": 0.32230141945183277, + "loss_total_v6": 0.4544730521738529, + "step": 1710 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.394609009009009, + "grad_norm": 24.32813835144043, + "learning_rate": 7.052863212143215e-06, + "loss": 7.551877975463867, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.3381727859377861, + "loss_total_v6": 0.4719923809170723, + "step": 1711 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39483963963963964, + "grad_norm": 24.285160064697266, + "learning_rate": 7.049399622442199e-06, + "loss": 6.765775680541992, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.28968227095901966, + "loss_total_v6": 0.4228609874844551, + "step": 1712 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39507027027027025, + "grad_norm": 25.785478591918945, + "learning_rate": 7.045934850338745e-06, + "loss": 7.0166521072387695, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3061249665915966, + "loss_total_v6": 0.4385407567024231, + "step": 1713 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3953009009009009, + "grad_norm": 25.130098342895508, + "learning_rate": 7.0424688978318576e-06, + "loss": 6.741524696350098, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.2894483134150505, + "loss_total_v6": 0.4213452897965908, + "step": 1714 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39553153153153153, + "grad_norm": 24.27228355407715, + "learning_rate": 7.0390017669212175e-06, + "loss": 6.7981858253479, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.2946833558380604, + "loss_total_v6": 0.4248866066336632, + "step": 1715 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39576216216216215, + "grad_norm": 28.391515731811523, + "learning_rate": 7.035533459607189e-06, + "loss": 7.1878180503845215, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.313771091401577, + "loss_total_v6": 0.4492386318743229, + "step": 1716 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3959927927927928, + "grad_norm": 21.681045532226562, + "learning_rate": 7.032063977890817e-06, + "loss": 6.870307922363281, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2993130497634411, + "loss_total_v6": 0.4293942376971245, + "step": 1717 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3962234234234234, + "grad_norm": 25.918054580688477, + "learning_rate": 7.028593323773819e-06, + "loss": 7.0449371337890625, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.30516148172318935, + "loss_total_v6": 0.4403085634112358, + "step": 1718 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39645405405405404, + "grad_norm": 22.19954490661621, + "learning_rate": 7.025121499258591e-06, + "loss": 7.19187068939209, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3175949230790138, + "loss_total_v6": 0.449491910636425, + "step": 1719 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3966846846846847, + "grad_norm": 22.748023986816406, + "learning_rate": 7.0216485063482045e-06, + "loss": 7.576654434204102, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.33894312754273415, + "loss_total_v6": 0.47354088723659515, + "step": 1720 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3969153153153153, + "grad_norm": 24.12191390991211, + "learning_rate": 7.018174347046406e-06, + "loss": 6.865686416625977, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.29646072164177895, + "loss_total_v6": 0.42910538241267204, + "step": 1721 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39714594594594593, + "grad_norm": 23.023361206054688, + "learning_rate": 7.014699023357612e-06, + "loss": 7.179342746734619, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3148283138871193, + "loss_total_v6": 0.4487089328467846, + "step": 1722 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3973765765765766, + "grad_norm": 22.0933895111084, + "learning_rate": 7.011222537286914e-06, + "loss": 6.8123321533203125, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.29454516246914864, + "loss_total_v6": 0.42577076330780983, + "step": 1723 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3976072072072072, + "grad_norm": 23.94240951538086, + "learning_rate": 7.0077448908400736e-06, + "loss": 6.940882682800293, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.30093161948025227, + "loss_total_v6": 0.4338051490485668, + "step": 1724 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3978378378378378, + "grad_norm": 22.112590789794922, + "learning_rate": 7.004266086023521e-06, + "loss": 7.280457973480225, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3205376584082842, + "loss_total_v6": 0.45502861589193344, + "step": 1725 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3980684684684685, + "grad_norm": 23.640592575073242, + "learning_rate": 7.000786124844355e-06, + "loss": 6.913600444793701, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.30165265686810017, + "loss_total_v6": 0.432100024074316, + "step": 1726 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3982990990990991, + "grad_norm": 22.22998046875, + "learning_rate": 6.997305009310338e-06, + "loss": 7.094301700592041, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.31013885512948036, + "loss_total_v6": 0.44339384883642197, + "step": 1727 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3985297297297297, + "grad_norm": 25.49310874938965, + "learning_rate": 6.9938227414299066e-06, + "loss": 7.050789833068848, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3103795666247606, + "loss_total_v6": 0.4406743608415127, + "step": 1728 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.39876036036036033, + "grad_norm": 23.529720306396484, + "learning_rate": 6.990339323212154e-06, + "loss": 7.169214248657227, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.31193696334958076, + "loss_total_v6": 0.44807588309049606, + "step": 1729 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.398990990990991, + "grad_norm": 23.671977996826172, + "learning_rate": 6.986854756666843e-06, + "loss": 6.6976847648620605, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.28382440842688084, + "loss_total_v6": 0.4186052940785885, + "step": 1730 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3992216216216216, + "grad_norm": 25.011428833007812, + "learning_rate": 6.983369043804396e-06, + "loss": 7.190242290496826, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3139531407505274, + "loss_total_v6": 0.44939015433192253, + "step": 1731 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3994522522522522, + "grad_norm": 23.39865493774414, + "learning_rate": 6.979882186635897e-06, + "loss": 7.1915202140808105, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.31738990545272827, + "loss_total_v6": 0.44947000592947006, + "step": 1732 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3996828828828829, + "grad_norm": 23.46658706665039, + "learning_rate": 6.97639418717309e-06, + "loss": 6.916979789733887, + "loss_alignment": 0.2686767578125, + "loss_alignment_w": 0.13433837890625, + "loss_sft": 0.2980643957853317, + "loss_total_v6": 0.4323112219572067, + "step": 1733 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.3999135135135135, + "grad_norm": 22.147666931152344, + "learning_rate": 6.97290504742838e-06, + "loss": 6.660827159881592, + "loss_alignment": 0.2689208984375, + "loss_alignment_w": 0.13446044921875, + "loss_sft": 0.28197857551276684, + "loss_total_v6": 0.4163016974925995, + "step": 1734 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4001441441441441, + "grad_norm": 22.709781646728516, + "learning_rate": 6.969414769414826e-06, + "loss": 6.882769584655762, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.2979709170758724, + "loss_total_v6": 0.4301730766892433, + "step": 1735 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4003747747747748, + "grad_norm": 23.252483367919922, + "learning_rate": 6.965923355146148e-06, + "loss": 7.312272071838379, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.32263289391994476, + "loss_total_v6": 0.45701703429222107, + "step": 1736 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4006054054054054, + "grad_norm": 23.959766387939453, + "learning_rate": 6.962430806636719e-06, + "loss": 6.761720657348633, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.29037486016750336, + "loss_total_v6": 0.42260752618312836, + "step": 1737 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.400836036036036, + "grad_norm": 25.11557388305664, + "learning_rate": 6.958937125901568e-06, + "loss": 7.065372467041016, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3060724791139364, + "loss_total_v6": 0.4415857791900635, + "step": 1738 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4010666666666667, + "grad_norm": 28.377796173095703, + "learning_rate": 6.955442314956375e-06, + "loss": 7.797832012176514, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.3563067764043808, + "loss_total_v6": 0.4873645007610321, + "step": 1739 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4012972972972973, + "grad_norm": 23.433956146240234, + "learning_rate": 6.9519463758174745e-06, + "loss": 6.911541938781738, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.29818231984972954, + "loss_total_v6": 0.43197138980031013, + "step": 1740 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4015279279279279, + "grad_norm": 25.12382698059082, + "learning_rate": 6.9484493105018515e-06, + "loss": 7.136404991149902, + "loss_alignment": 0.2681884765625, + "loss_alignment_w": 0.13409423828125, + "loss_sft": 0.3118242621421814, + "loss_total_v6": 0.4460253044962883, + "step": 1741 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4017585585585586, + "grad_norm": 25.215723037719727, + "learning_rate": 6.944951121027137e-06, + "loss": 7.107004165649414, + "loss_alignment": 0.2708740234375, + "loss_alignment_w": 0.13543701171875, + "loss_sft": 0.30894909985363483, + "loss_total_v6": 0.4441877566277981, + "step": 1742 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4019891891891892, + "grad_norm": 24.324737548828125, + "learning_rate": 6.941451809411618e-06, + "loss": 7.076571464538574, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3109685927629471, + "loss_total_v6": 0.4422857128083706, + "step": 1743 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4022198198198198, + "grad_norm": 23.291345596313477, + "learning_rate": 6.937951377674222e-06, + "loss": 7.251629829406738, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.31893428042531013, + "loss_total_v6": 0.45322687178850174, + "step": 1744 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40245045045045047, + "grad_norm": 23.06678009033203, + "learning_rate": 6.9344498278345244e-06, + "loss": 7.035035610198975, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.3087387979030609, + "loss_total_v6": 0.4396897330880165, + "step": 1745 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4026810810810811, + "grad_norm": 23.43814468383789, + "learning_rate": 6.930947161912749e-06, + "loss": 6.867583274841309, + "loss_alignment": 0.2686767578125, + "loss_alignment_w": 0.13433837890625, + "loss_sft": 0.29499236308038235, + "loss_total_v6": 0.4292239286005497, + "step": 1746 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4029117117117117, + "grad_norm": 21.549623489379883, + "learning_rate": 6.927443381929761e-06, + "loss": 6.548892021179199, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.27699675038456917, + "loss_total_v6": 0.40930572897195816, + "step": 1747 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40314234234234236, + "grad_norm": 24.642494201660156, + "learning_rate": 6.9239384899070665e-06, + "loss": 7.26820182800293, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3211296945810318, + "loss_total_v6": 0.4542626291513443, + "step": 1748 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.403372972972973, + "grad_norm": 25.01157569885254, + "learning_rate": 6.920432487866816e-06, + "loss": 6.639026641845703, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.28247762843966484, + "loss_total_v6": 0.41493918001651764, + "step": 1749 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4036036036036036, + "grad_norm": 25.676986694335938, + "learning_rate": 6.916925377831801e-06, + "loss": 7.247930526733398, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3194965161383152, + "loss_total_v6": 0.4529956690967083, + "step": 1750 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40383423423423426, + "grad_norm": 23.637802124023438, + "learning_rate": 6.913417161825449e-06, + "loss": 7.199637413024902, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3161882758140564, + "loss_total_v6": 0.4499773494899273, + "step": 1751 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40406486486486487, + "grad_norm": 28.054569244384766, + "learning_rate": 6.90990784187183e-06, + "loss": 6.988208293914795, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3012191690504551, + "loss_total_v6": 0.4367630109190941, + "step": 1752 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4042954954954955, + "grad_norm": 26.005613327026367, + "learning_rate": 6.9063974199956454e-06, + "loss": 6.994873523712158, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3046569861471653, + "loss_total_v6": 0.4371795915067196, + "step": 1753 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40452612612612615, + "grad_norm": 22.23686408996582, + "learning_rate": 6.902885898222238e-06, + "loss": 6.906289100646973, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.29794556088745594, + "loss_total_v6": 0.4316430613398552, + "step": 1754 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40475675675675676, + "grad_norm": 23.4333438873291, + "learning_rate": 6.89937327857758e-06, + "loss": 6.586069107055664, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.28008332662284374, + "loss_total_v6": 0.4116293489933014, + "step": 1755 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4049873873873874, + "grad_norm": 28.16299057006836, + "learning_rate": 6.895859563088283e-06, + "loss": 7.1887407302856445, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.3152325935661793, + "loss_total_v6": 0.4492963030934334, + "step": 1756 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40521801801801804, + "grad_norm": 24.006561279296875, + "learning_rate": 6.8923447537815845e-06, + "loss": 7.225373268127441, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.3171406574547291, + "loss_total_v6": 0.4515858441591263, + "step": 1757 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40544864864864866, + "grad_norm": 23.847116470336914, + "learning_rate": 6.8888288526853564e-06, + "loss": 7.4077019691467285, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.32862773165106773, + "loss_total_v6": 0.46298137307167053, + "step": 1758 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40567927927927927, + "grad_norm": 26.32425880432129, + "learning_rate": 6.8853118618281014e-06, + "loss": 6.886918067932129, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.29800136014819145, + "loss_total_v6": 0.43043239042162895, + "step": 1759 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4059099099099099, + "grad_norm": 23.44271469116211, + "learning_rate": 6.881793783238948e-06, + "loss": 7.124020576477051, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31399519369006157, + "loss_total_v6": 0.4452512860298157, + "step": 1760 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40614054054054055, + "grad_norm": 24.491640090942383, + "learning_rate": 6.8782746189476544e-06, + "loss": 6.980618000030518, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.30310992524027824, + "loss_total_v6": 0.43628861010074615, + "step": 1761 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40637117117117116, + "grad_norm": 26.044822692871094, + "learning_rate": 6.8747543709846064e-06, + "loss": 6.896728038787842, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.2977294586598873, + "loss_total_v6": 0.4310455061495304, + "step": 1762 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4066018018018018, + "grad_norm": 23.92768096923828, + "learning_rate": 6.871233041380812e-06, + "loss": 6.944746017456055, + "loss_alignment": 0.274169921875, + "loss_alignment_w": 0.1370849609375, + "loss_sft": 0.2970989868044853, + "loss_total_v6": 0.4340466186404228, + "step": 1763 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40683243243243244, + "grad_norm": 24.65842628479004, + "learning_rate": 6.867710632167903e-06, + "loss": 6.895452499389648, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.2981227859854698, + "loss_total_v6": 0.4309657849371433, + "step": 1764 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40706306306306306, + "grad_norm": 28.616729736328125, + "learning_rate": 6.864187145378138e-06, + "loss": 7.456714630126953, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.33066869899630547, + "loss_total_v6": 0.46604468300938606, + "step": 1765 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40729369369369367, + "grad_norm": 26.825101852416992, + "learning_rate": 6.860662583044395e-06, + "loss": 7.101797580718994, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3092188388109207, + "loss_total_v6": 0.4438623636960983, + "step": 1766 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40752432432432434, + "grad_norm": 23.470773696899414, + "learning_rate": 6.857136947200171e-06, + "loss": 7.379549980163574, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.327768512070179, + "loss_total_v6": 0.4612218849360943, + "step": 1767 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40775495495495495, + "grad_norm": 23.71815299987793, + "learning_rate": 6.853610239879586e-06, + "loss": 7.548844337463379, + "loss_alignment": 0.279541015625, + "loss_alignment_w": 0.1397705078125, + "loss_sft": 0.3318338766694069, + "loss_total_v6": 0.4718027599155903, + "step": 1768 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40798558558558556, + "grad_norm": 27.064537048339844, + "learning_rate": 6.850082463117374e-06, + "loss": 6.766752243041992, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.28721033222973347, + "loss_total_v6": 0.4229220002889633, + "step": 1769 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40821621621621623, + "grad_norm": 24.63895034790039, + "learning_rate": 6.846553618948889e-06, + "loss": 7.081521034240723, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.30905015021562576, + "loss_total_v6": 0.44259506091475487, + "step": 1770 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40844684684684684, + "grad_norm": 24.31847381591797, + "learning_rate": 6.8430237094100985e-06, + "loss": 7.816771507263184, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.3521956466138363, + "loss_total_v6": 0.4885482043027878, + "step": 1771 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40867747747747746, + "grad_norm": 21.581233978271484, + "learning_rate": 6.839492736537588e-06, + "loss": 7.3116302490234375, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.32428644225001335, + "loss_total_v6": 0.45697688311338425, + "step": 1772 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4089081081081081, + "grad_norm": 23.768531799316406, + "learning_rate": 6.8359607023685544e-06, + "loss": 7.480464935302734, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3328855149447918, + "loss_total_v6": 0.4675290733575821, + "step": 1773 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40913873873873874, + "grad_norm": 23.57581329345703, + "learning_rate": 6.832427608940806e-06, + "loss": 6.851899147033691, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.29514127038419247, + "loss_total_v6": 0.4282436743378639, + "step": 1774 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40936936936936935, + "grad_norm": 24.432382583618164, + "learning_rate": 6.828893458292765e-06, + "loss": 7.579558849334717, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3411235474050045, + "loss_total_v6": 0.4737224318087101, + "step": 1775 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4096, + "grad_norm": 22.924699783325195, + "learning_rate": 6.82535825246346e-06, + "loss": 6.946450233459473, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.30091341957449913, + "loss_total_v6": 0.43415316566824913, + "step": 1776 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.40983063063063063, + "grad_norm": 23.8988037109375, + "learning_rate": 6.821821993492532e-06, + "loss": 6.937662124633789, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.30080664344131947, + "loss_total_v6": 0.4336038753390312, + "step": 1777 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41006126126126125, + "grad_norm": 27.14107894897461, + "learning_rate": 6.818284683420227e-06, + "loss": 7.26401424407959, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3190673999488354, + "loss_total_v6": 0.45400087535381317, + "step": 1778 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4102918918918919, + "grad_norm": 26.95549774169922, + "learning_rate": 6.814746324287397e-06, + "loss": 7.635558128356934, + "loss_alignment": 0.272216796875, + "loss_alignment_w": 0.1361083984375, + "loss_sft": 0.34100719541311264, + "loss_total_v6": 0.47722239419817924, + "step": 1779 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4105225225225225, + "grad_norm": 24.87458038330078, + "learning_rate": 6.811206918135502e-06, + "loss": 7.458937644958496, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.3350800946354866, + "loss_total_v6": 0.4661836139857769, + "step": 1780 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41075315315315314, + "grad_norm": 26.052053451538086, + "learning_rate": 6.807666467006605e-06, + "loss": 7.259003639221191, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.31985291838645935, + "loss_total_v6": 0.45368774235248566, + "step": 1781 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4109837837837838, + "grad_norm": 23.324832916259766, + "learning_rate": 6.804124972943372e-06, + "loss": 7.308723449707031, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.32732437923550606, + "loss_total_v6": 0.45679520070552826, + "step": 1782 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4112144144144144, + "grad_norm": 24.70726776123047, + "learning_rate": 6.80058243798907e-06, + "loss": 7.121578216552734, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3117520771920681, + "loss_total_v6": 0.4450986310839653, + "step": 1783 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41144504504504503, + "grad_norm": 22.264591217041016, + "learning_rate": 6.797038864187564e-06, + "loss": 7.090551376342773, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.30863798037171364, + "loss_total_v6": 0.44315946847200394, + "step": 1784 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4116756756756757, + "grad_norm": 24.221168518066406, + "learning_rate": 6.793494253583326e-06, + "loss": 6.759077072143555, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.29173554107546806, + "loss_total_v6": 0.42244232445955276, + "step": 1785 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4119063063063063, + "grad_norm": 22.123252868652344, + "learning_rate": 6.789948608221418e-06, + "loss": 6.845025062561035, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.297320906072855, + "loss_total_v6": 0.4278140589594841, + "step": 1786 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4121369369369369, + "grad_norm": 26.008848190307617, + "learning_rate": 6.786401930147502e-06, + "loss": 7.278994560241699, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3215295784175396, + "loss_total_v6": 0.4549371711909771, + "step": 1787 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4123675675675676, + "grad_norm": 23.467906951904297, + "learning_rate": 6.7828542214078385e-06, + "loss": 7.2738213539123535, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.319451492279768, + "loss_total_v6": 0.4546138420701027, + "step": 1788 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4125981981981982, + "grad_norm": 26.253549575805664, + "learning_rate": 6.77930548404928e-06, + "loss": 6.84505558013916, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.29347760044038296, + "loss_total_v6": 0.4278159625828266, + "step": 1789 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4128288288288288, + "grad_norm": 24.349613189697266, + "learning_rate": 6.775755720119272e-06, + "loss": 7.221066474914551, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.31960279494524, + "loss_total_v6": 0.45131663978099823, + "step": 1790 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41305945945945943, + "grad_norm": 23.08489990234375, + "learning_rate": 6.772204931665853e-06, + "loss": 6.980788230895996, + "loss_alignment": 0.2652587890625, + "loss_alignment_w": 0.13262939453125, + "loss_sft": 0.3036698829382658, + "loss_total_v6": 0.43629928678274155, + "step": 1791 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4132900900900901, + "grad_norm": 23.096067428588867, + "learning_rate": 6.7686531207376525e-06, + "loss": 7.037659645080566, + "loss_alignment": 0.2724609375, + "loss_alignment_w": 0.13623046875, + "loss_sft": 0.30360800214111805, + "loss_total_v6": 0.4398537576198578, + "step": 1792 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4135207207207207, + "grad_norm": 24.231672286987305, + "learning_rate": 6.765100289383891e-06, + "loss": 6.783980369567871, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.29323096573352814, + "loss_total_v6": 0.42399879172444344, + "step": 1793 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4137513513513513, + "grad_norm": 24.021482467651367, + "learning_rate": 6.761546439654376e-06, + "loss": 6.769906044006348, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.28899434953927994, + "loss_total_v6": 0.42311910167336464, + "step": 1794 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.413981981981982, + "grad_norm": 24.792375564575195, + "learning_rate": 6.757991573599504e-06, + "loss": 6.653813362121582, + "loss_alignment": 0.2657470703125, + "loss_alignment_w": 0.13287353515625, + "loss_sft": 0.2830660920590162, + "loss_total_v6": 0.41586335375905037, + "step": 1795 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4142126126126126, + "grad_norm": 23.740530014038086, + "learning_rate": 6.754435693270258e-06, + "loss": 7.151544094085693, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.311534496024251, + "loss_total_v6": 0.44697150960564613, + "step": 1796 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4144432432432432, + "grad_norm": 24.362417221069336, + "learning_rate": 6.750878800718203e-06, + "loss": 7.012139320373535, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.30628542602062225, + "loss_total_v6": 0.43825870379805565, + "step": 1797 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4146738738738739, + "grad_norm": 24.541915893554688, + "learning_rate": 6.747320897995493e-06, + "loss": 7.3417649269104, + "loss_alignment": 0.273193359375, + "loss_alignment_w": 0.1365966796875, + "loss_sft": 0.32262982428073883, + "loss_total_v6": 0.4588603079319, + "step": 1798 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4149045045045045, + "grad_norm": 23.41994857788086, + "learning_rate": 6.743761987154861e-06, + "loss": 6.668687343597412, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.2837820854038, + "loss_total_v6": 0.41679295152425766, + "step": 1799 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4151351351351351, + "grad_norm": 23.712053298950195, + "learning_rate": 6.740202070249623e-06, + "loss": 7.217047691345215, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.31814616546034813, + "loss_total_v6": 0.4510654844343662, + "step": 1800 + }, + { + "epoch": 0.4151351351351351, + "eval_loss": 0.5926827788352966, + "eval_loss_alignment": 0.23993115011415525, + "eval_loss_alignment_w": 0.11996557505707763, + "eval_loss_sft": 0.4727171967614051, + "eval_loss_total_v6": 0.5926827716738938, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.1591, + "eval_samples_per_second": 19.142, + "eval_steps_per_second": 2.402, + "step": 1800 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4153657657657658, + "grad_norm": 24.204572677612305, + "learning_rate": 6.7366411493336735e-06, + "loss": 6.950666904449463, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.30447282269597054, + "loss_total_v6": 0.43441667407751083, + "step": 1801 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4155963963963964, + "grad_norm": 26.05511474609375, + "learning_rate": 6.733079226461493e-06, + "loss": 7.2179789543151855, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.31712101213634014, + "loss_total_v6": 0.4511236883699894, + "step": 1802 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.415827027027027, + "grad_norm": 25.689592361450195, + "learning_rate": 6.729516303688133e-06, + "loss": 6.784222602844238, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.2916439287364483, + "loss_total_v6": 0.4240139238536358, + "step": 1803 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4160576576576577, + "grad_norm": 25.091230392456055, + "learning_rate": 6.725952383069222e-06, + "loss": 6.627202033996582, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.28350857831537724, + "loss_total_v6": 0.4142001010477543, + "step": 1804 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4162882882882883, + "grad_norm": 25.2420711517334, + "learning_rate": 6.722387466660966e-06, + "loss": 6.8825531005859375, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.2988576628267765, + "loss_total_v6": 0.430159542709589, + "step": 1805 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4165189189189189, + "grad_norm": 22.260534286499023, + "learning_rate": 6.718821556520151e-06, + "loss": 7.499554634094238, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3339107595384121, + "loss_total_v6": 0.4687221646308899, + "step": 1806 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41674954954954957, + "grad_norm": 24.92996597290039, + "learning_rate": 6.715254654704126e-06, + "loss": 7.084064960479736, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3099568020552397, + "loss_total_v6": 0.4427540525794029, + "step": 1807 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4169801801801802, + "grad_norm": 23.378313064575195, + "learning_rate": 6.7116867632708184e-06, + "loss": 7.1519269943237305, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.31360308453440666, + "loss_total_v6": 0.44699542596936226, + "step": 1808 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4172108108108108, + "grad_norm": 25.125444412231445, + "learning_rate": 6.7081178842787275e-06, + "loss": 7.338326454162598, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.32307106256484985, + "loss_total_v6": 0.45864540711045265, + "step": 1809 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41744144144144146, + "grad_norm": 28.58230209350586, + "learning_rate": 6.704548019786917e-06, + "loss": 6.700751781463623, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.28874633461236954, + "loss_total_v6": 0.41879698634147644, + "step": 1810 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4176720720720721, + "grad_norm": 25.05723762512207, + "learning_rate": 6.700977171855027e-06, + "loss": 6.903478145599365, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.2992347292602062, + "loss_total_v6": 0.4314673915505409, + "step": 1811 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4179027027027027, + "grad_norm": 23.290321350097656, + "learning_rate": 6.697405342543257e-06, + "loss": 6.787879467010498, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.2886986546218395, + "loss_total_v6": 0.4242424741387367, + "step": 1812 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41813333333333336, + "grad_norm": 23.124094009399414, + "learning_rate": 6.693832533912378e-06, + "loss": 7.077995300292969, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.30998946726322174, + "loss_total_v6": 0.44237470999360085, + "step": 1813 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41836396396396397, + "grad_norm": 23.443828582763672, + "learning_rate": 6.69025874802372e-06, + "loss": 6.9696044921875, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3032913412898779, + "loss_total_v6": 0.43560029938817024, + "step": 1814 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4185945945945946, + "grad_norm": 25.388992309570312, + "learning_rate": 6.6866839869391884e-06, + "loss": 6.893629550933838, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.29463662952184677, + "loss_total_v6": 0.43085185438394547, + "step": 1815 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41882522522522525, + "grad_norm": 23.92854881286621, + "learning_rate": 6.683108252721239e-06, + "loss": 6.786142349243164, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.29391541332006454, + "loss_total_v6": 0.42413391545414925, + "step": 1816 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41905585585585586, + "grad_norm": 30.25241470336914, + "learning_rate": 6.679531547432896e-06, + "loss": 6.493494987487793, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2744347285479307, + "loss_total_v6": 0.40584342554211617, + "step": 1817 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4192864864864865, + "grad_norm": 23.095373153686523, + "learning_rate": 6.6759538731377385e-06, + "loss": 6.902578830718994, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.301741985604167, + "loss_total_v6": 0.43141117691993713, + "step": 1818 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41951711711711714, + "grad_norm": 24.250377655029297, + "learning_rate": 6.672375231899913e-06, + "loss": 6.866866111755371, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.2981671616435051, + "loss_total_v6": 0.4291791208088398, + "step": 1819 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41974774774774776, + "grad_norm": 25.534957885742188, + "learning_rate": 6.668795625784114e-06, + "loss": 7.16403865814209, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3172745071351528, + "loss_total_v6": 0.4477524086833, + "step": 1820 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.41997837837837837, + "grad_norm": 21.161375045776367, + "learning_rate": 6.6652150568556e-06, + "loss": 7.117252349853516, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3134043216705322, + "loss_total_v6": 0.4448282793164253, + "step": 1821 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.420209009009009, + "grad_norm": 26.818029403686523, + "learning_rate": 6.661633527180184e-06, + "loss": 7.331514358520508, + "loss_alignment": 0.275146484375, + "loss_alignment_w": 0.1375732421875, + "loss_sft": 0.32058537751436234, + "loss_total_v6": 0.45821966603398323, + "step": 1822 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42043963963963965, + "grad_norm": 23.85163116455078, + "learning_rate": 6.6580510388242294e-06, + "loss": 6.92271614074707, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.30307688005268574, + "loss_total_v6": 0.4326697587966919, + "step": 1823 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42067027027027026, + "grad_norm": 25.236818313598633, + "learning_rate": 6.654467593854656e-06, + "loss": 6.844860076904297, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.29634926468133926, + "loss_total_v6": 0.42780375108122826, + "step": 1824 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4209009009009009, + "grad_norm": 22.225305557250977, + "learning_rate": 6.650883194338935e-06, + "loss": 7.435695171356201, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.3316438104957342, + "loss_total_v6": 0.46473095938563347, + "step": 1825 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42113153153153154, + "grad_norm": 23.801328659057617, + "learning_rate": 6.6472978423450874e-06, + "loss": 6.955099582672119, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.29776136949658394, + "loss_total_v6": 0.43469373881816864, + "step": 1826 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42136216216216216, + "grad_norm": 25.160139083862305, + "learning_rate": 6.643711539941684e-06, + "loss": 7.291599273681641, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.32375168055295944, + "loss_total_v6": 0.45572494715452194, + "step": 1827 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42159279279279277, + "grad_norm": 23.81449317932129, + "learning_rate": 6.640124289197845e-06, + "loss": 7.117920875549316, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3104706443846226, + "loss_total_v6": 0.44487006589770317, + "step": 1828 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42182342342342344, + "grad_norm": 25.667118072509766, + "learning_rate": 6.636536092183239e-06, + "loss": 6.938725471496582, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.2998202331364155, + "loss_total_v6": 0.4336703270673752, + "step": 1829 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42205405405405405, + "grad_norm": 27.28817367553711, + "learning_rate": 6.6329469509680756e-06, + "loss": 7.520132541656494, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.3362955302000046, + "loss_total_v6": 0.4700082950294018, + "step": 1830 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42228468468468466, + "grad_norm": 23.293384552001953, + "learning_rate": 6.629356867623112e-06, + "loss": 6.775583744049072, + "loss_alignment": 0.2684326171875, + "loss_alignment_w": 0.13421630859375, + "loss_sft": 0.289440780878067, + "loss_total_v6": 0.4234739802777767, + "step": 1831 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42251531531531533, + "grad_norm": 25.131879806518555, + "learning_rate": 6.625765844219651e-06, + "loss": 6.448996067047119, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.27050914987921715, + "loss_total_v6": 0.40306225046515465, + "step": 1832 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42274594594594594, + "grad_norm": 25.34829330444336, + "learning_rate": 6.622173882829536e-06, + "loss": 7.409156799316406, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.33103797771036625, + "loss_total_v6": 0.4630722962319851, + "step": 1833 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42297657657657656, + "grad_norm": 23.481115341186523, + "learning_rate": 6.618580985525149e-06, + "loss": 7.148917198181152, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3135217949748039, + "loss_total_v6": 0.4468073137104511, + "step": 1834 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4232072072072072, + "grad_norm": 25.611059188842773, + "learning_rate": 6.614987154379417e-06, + "loss": 7.162467002868652, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.31204934790730476, + "loss_total_v6": 0.44765419512987137, + "step": 1835 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42343783783783784, + "grad_norm": 24.26041603088379, + "learning_rate": 6.611392391465802e-06, + "loss": 6.935613632202148, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3023876063525677, + "loss_total_v6": 0.4334758594632149, + "step": 1836 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42366846846846845, + "grad_norm": 24.185222625732422, + "learning_rate": 6.6077966988583054e-06, + "loss": 6.717281341552734, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.2871091477572918, + "loss_total_v6": 0.4198300950229168, + "step": 1837 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4238990990990991, + "grad_norm": 25.59442901611328, + "learning_rate": 6.604200078631464e-06, + "loss": 6.956956386566162, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3013411555439234, + "loss_total_v6": 0.434809785336256, + "step": 1838 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42412972972972973, + "grad_norm": 23.053003311157227, + "learning_rate": 6.600602532860349e-06, + "loss": 6.6992621421813965, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.28647121973335743, + "loss_total_v6": 0.4187038764357567, + "step": 1839 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42436036036036034, + "grad_norm": 25.284955978393555, + "learning_rate": 6.597004063620568e-06, + "loss": 6.746916770935059, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.29047200083732605, + "loss_total_v6": 0.42168232053518295, + "step": 1840 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.424590990990991, + "grad_norm": 23.176023483276367, + "learning_rate": 6.593404672988261e-06, + "loss": 7.092006683349609, + "loss_alignment": 0.27294921875, + "loss_alignment_w": 0.136474609375, + "loss_sft": 0.3065927177667618, + "loss_total_v6": 0.4432504363358021, + "step": 1841 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4248216216216216, + "grad_norm": 24.26901626586914, + "learning_rate": 6.5898043630400975e-06, + "loss": 6.97260856628418, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.30032048374414444, + "loss_total_v6": 0.43578801304101944, + "step": 1842 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42505225225225224, + "grad_norm": 25.163976669311523, + "learning_rate": 6.58620313585328e-06, + "loss": 7.103283882141113, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3090522922575474, + "loss_total_v6": 0.4439552426338196, + "step": 1843 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4252828828828829, + "grad_norm": 25.67691993713379, + "learning_rate": 6.582600993505535e-06, + "loss": 7.018746376037598, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.30688149854540825, + "loss_total_v6": 0.43867165967822075, + "step": 1844 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4255135135135135, + "grad_norm": 26.973163604736328, + "learning_rate": 6.578997938075126e-06, + "loss": 6.703486919403076, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.28859684243798256, + "loss_total_v6": 0.41896793991327286, + "step": 1845 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42574414414414413, + "grad_norm": 24.818708419799805, + "learning_rate": 6.575393971640833e-06, + "loss": 6.8494954109191895, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.29862265288829803, + "loss_total_v6": 0.42809346318244934, + "step": 1846 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4259747747747748, + "grad_norm": 26.528976440429688, + "learning_rate": 6.571789096281973e-06, + "loss": 7.036110877990723, + "loss_alignment": 0.2657470703125, + "loss_alignment_w": 0.13287353515625, + "loss_sft": 0.30699020624160767, + "loss_total_v6": 0.43975692242383957, + "step": 1847 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4262054054054054, + "grad_norm": 23.64221954345703, + "learning_rate": 6.568183314078377e-06, + "loss": 6.764074325561523, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.2881111018359661, + "loss_total_v6": 0.4227546565234661, + "step": 1848 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.426436036036036, + "grad_norm": 28.144983291625977, + "learning_rate": 6.564576627110401e-06, + "loss": 6.954736232757568, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.30451352894306183, + "loss_total_v6": 0.4346710033714771, + "step": 1849 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4266666666666667, + "grad_norm": 24.693429946899414, + "learning_rate": 6.560969037458933e-06, + "loss": 7.0886430740356445, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3079693876206875, + "loss_total_v6": 0.4430401958525181, + "step": 1850 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4268972972972973, + "grad_norm": 22.72039222717285, + "learning_rate": 6.557360547205367e-06, + "loss": 7.28737211227417, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3210918679833412, + "loss_total_v6": 0.4554607719182968, + "step": 1851 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4271279279279279, + "grad_norm": 24.560253143310547, + "learning_rate": 6.553751158431627e-06, + "loss": 7.0483551025390625, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.30827426724135876, + "loss_total_v6": 0.4405221939086914, + "step": 1852 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.42735855855855853, + "grad_norm": 24.695556640625, + "learning_rate": 6.550140873220151e-06, + "loss": 7.01654577255249, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.30715592950582504, + "loss_total_v6": 0.43853410705924034, + "step": 1853 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4275891891891892, + "grad_norm": 23.55740737915039, + "learning_rate": 6.546529693653897e-06, + "loss": 7.302486896514893, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.325088307261467, + "loss_total_v6": 0.4564054310321808, + "step": 1854 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4278198198198198, + "grad_norm": 26.371450424194336, + "learning_rate": 6.5429176218163356e-06, + "loss": 6.96109676361084, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3025612309575081, + "loss_total_v6": 0.4350685514509678, + "step": 1855 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4280504504504504, + "grad_norm": 24.243499755859375, + "learning_rate": 6.539304659791457e-06, + "loss": 6.80364465713501, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.2904621586203575, + "loss_total_v6": 0.4252277798950672, + "step": 1856 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4282810810810811, + "grad_norm": 22.828880310058594, + "learning_rate": 6.535690809663757e-06, + "loss": 7.034340858459473, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.30684906616806984, + "loss_total_v6": 0.43964632228016853, + "step": 1857 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4285117117117117, + "grad_norm": 25.201871871948242, + "learning_rate": 6.532076073518253e-06, + "loss": 7.005693435668945, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3064013682305813, + "loss_total_v6": 0.43785585090518, + "step": 1858 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4287423423423423, + "grad_norm": 23.34696388244629, + "learning_rate": 6.528460453440466e-06, + "loss": 6.86505126953125, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.2974128667265177, + "loss_total_v6": 0.4290657006204128, + "step": 1859 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.428972972972973, + "grad_norm": 24.35470962524414, + "learning_rate": 6.524843951516436e-06, + "loss": 7.161704063415527, + "loss_alignment": 0.2744140625, + "loss_alignment_w": 0.13720703125, + "loss_sft": 0.31053679436445236, + "loss_total_v6": 0.44760649278759956, + "step": 1860 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4292036036036036, + "grad_norm": 26.002573013305664, + "learning_rate": 6.521226569832699e-06, + "loss": 6.9568772315979, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.30078686960041523, + "loss_total_v6": 0.4348048157989979, + "step": 1861 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4294342342342342, + "grad_norm": 23.724342346191406, + "learning_rate": 6.51760831047631e-06, + "loss": 7.023612022399902, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.30314200185239315, + "loss_total_v6": 0.4389757476747036, + "step": 1862 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4296648648648649, + "grad_norm": 24.81387710571289, + "learning_rate": 6.513989175534825e-06, + "loss": 6.9481353759765625, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.29966070130467415, + "loss_total_v6": 0.43425847962498665, + "step": 1863 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4298954954954955, + "grad_norm": 22.11371612548828, + "learning_rate": 6.510369167096308e-06, + "loss": 6.801024436950684, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.29220573604106903, + "loss_total_v6": 0.42506401613354683, + "step": 1864 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4301261261261261, + "grad_norm": 24.74908447265625, + "learning_rate": 6.50674828724932e-06, + "loss": 6.71868896484375, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2894859127700329, + "loss_total_v6": 0.4199180454015732, + "step": 1865 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4303567567567568, + "grad_norm": 23.501708984375, + "learning_rate": 6.5031265380829355e-06, + "loss": 7.464362621307373, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3344426043331623, + "loss_total_v6": 0.4665226601064205, + "step": 1866 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4305873873873874, + "grad_norm": 21.978628158569336, + "learning_rate": 6.499503921686725e-06, + "loss": 6.769733428955078, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.2872898541390896, + "loss_total_v6": 0.4231083281338215, + "step": 1867 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.430818018018018, + "grad_norm": 23.109113693237305, + "learning_rate": 6.495880440150756e-06, + "loss": 6.803309440612793, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.29189079627394676, + "loss_total_v6": 0.42520684748888016, + "step": 1868 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43104864864864867, + "grad_norm": 30.467403411865234, + "learning_rate": 6.4922560955656e-06, + "loss": 6.79038143157959, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.29154054820537567, + "loss_total_v6": 0.4243988171219826, + "step": 1869 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4312792792792793, + "grad_norm": 24.42734146118164, + "learning_rate": 6.488630890022324e-06, + "loss": 6.759491920471191, + "loss_alignment": 0.2706298828125, + "loss_alignment_w": 0.13531494140625, + "loss_sft": 0.287015987560153, + "loss_total_v6": 0.42246825620532036, + "step": 1870 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4315099099099099, + "grad_norm": 24.29322052001953, + "learning_rate": 6.485004825612495e-06, + "loss": 7.256925582885742, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3202418237924576, + "loss_total_v6": 0.4535578489303589, + "step": 1871 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43174054054054056, + "grad_norm": 25.884592056274414, + "learning_rate": 6.481377904428171e-06, + "loss": 7.035457611083984, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3056371137499809, + "loss_total_v6": 0.43971608206629753, + "step": 1872 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4319711711711712, + "grad_norm": 21.910078048706055, + "learning_rate": 6.477750128561907e-06, + "loss": 7.030409336090088, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.3066185973584652, + "loss_total_v6": 0.4394005760550499, + "step": 1873 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4322018018018018, + "grad_norm": 22.2122802734375, + "learning_rate": 6.474121500106752e-06, + "loss": 7.118762493133545, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.31368179991841316, + "loss_total_v6": 0.44492265954613686, + "step": 1874 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43243243243243246, + "grad_norm": 24.25014877319336, + "learning_rate": 6.470492021156243e-06, + "loss": 6.515196800231934, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.27791208773851395, + "loss_total_v6": 0.40719980746507645, + "step": 1875 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43266306306306307, + "grad_norm": 23.402132034301758, + "learning_rate": 6.466861693804413e-06, + "loss": 7.045675277709961, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.3093732539564371, + "loss_total_v6": 0.44035468623042107, + "step": 1876 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4328936936936937, + "grad_norm": 24.541540145874023, + "learning_rate": 6.46323052014578e-06, + "loss": 7.1796441078186035, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3174106292426586, + "loss_total_v6": 0.4487277641892433, + "step": 1877 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43312432432432435, + "grad_norm": 25.899375915527344, + "learning_rate": 6.4595985022753524e-06, + "loss": 7.256582736968994, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.32029667869210243, + "loss_total_v6": 0.45353642851114273, + "step": 1878 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43335495495495496, + "grad_norm": 31.521718978881836, + "learning_rate": 6.455965642288624e-06, + "loss": 6.963566303253174, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.3008997868746519, + "loss_total_v6": 0.43522290512919426, + "step": 1879 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4335855855855856, + "grad_norm": 23.408628463745117, + "learning_rate": 6.45233194228158e-06, + "loss": 7.057803153991699, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3058740273118019, + "loss_total_v6": 0.4411126784980297, + "step": 1880 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43381621621621624, + "grad_norm": 23.367351531982422, + "learning_rate": 6.448697404350681e-06, + "loss": 6.734011650085449, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.28858201019465923, + "loss_total_v6": 0.42087572067976, + "step": 1881 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43404684684684686, + "grad_norm": 25.56936264038086, + "learning_rate": 6.445062030592879e-06, + "loss": 7.006515979766846, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3048047963529825, + "loss_total_v6": 0.43790723010897636, + "step": 1882 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43427747747747747, + "grad_norm": 23.353626251220703, + "learning_rate": 6.441425823105603e-06, + "loss": 7.146674633026123, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.31350371055305004, + "loss_total_v6": 0.4466671645641327, + "step": 1883 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4345081081081081, + "grad_norm": 26.395727157592773, + "learning_rate": 6.437788783986767e-06, + "loss": 7.067471981048584, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3056848980486393, + "loss_total_v6": 0.4417169950902462, + "step": 1884 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43473873873873875, + "grad_norm": 24.876394271850586, + "learning_rate": 6.434150915334763e-06, + "loss": 7.343109130859375, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.32616232708096504, + "loss_total_v6": 0.45894430577754974, + "step": 1885 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43496936936936936, + "grad_norm": 23.018184661865234, + "learning_rate": 6.43051221924846e-06, + "loss": 7.2162370681762695, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.31748514622449875, + "loss_total_v6": 0.45101481303572655, + "step": 1886 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4352, + "grad_norm": 25.352190017700195, + "learning_rate": 6.426872697827208e-06, + "loss": 6.9867377281188965, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.30124934762716293, + "loss_total_v6": 0.4366711154580116, + "step": 1887 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43543063063063064, + "grad_norm": 22.232446670532227, + "learning_rate": 6.423232353170831e-06, + "loss": 6.751187324523926, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.28962499275803566, + "loss_total_v6": 0.42194921523332596, + "step": 1888 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43566126126126126, + "grad_norm": 23.9093074798584, + "learning_rate": 6.4195911873796264e-06, + "loss": 7.110700607299805, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.3089512549340725, + "loss_total_v6": 0.4444187991321087, + "step": 1889 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43589189189189187, + "grad_norm": 23.959060668945312, + "learning_rate": 6.41594920255437e-06, + "loss": 6.822198390960693, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.29395637661218643, + "loss_total_v6": 0.42638740688562393, + "step": 1890 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43612252252252254, + "grad_norm": 23.425321578979492, + "learning_rate": 6.412306400796306e-06, + "loss": 7.346062660217285, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3252940773963928, + "loss_total_v6": 0.4591289237141609, + "step": 1891 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43635315315315315, + "grad_norm": 25.199037551879883, + "learning_rate": 6.408662784207149e-06, + "loss": 7.293313026428223, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.32256180979311466, + "loss_total_v6": 0.4558320827782154, + "step": 1892 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43658378378378376, + "grad_norm": 23.55555534362793, + "learning_rate": 6.405018354889089e-06, + "loss": 7.174638271331787, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3152209296822548, + "loss_total_v6": 0.4484148994088173, + "step": 1893 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43681441441441443, + "grad_norm": 23.293182373046875, + "learning_rate": 6.401373114944781e-06, + "loss": 7.165703296661377, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.31331971660256386, + "loss_total_v6": 0.44785645976662636, + "step": 1894 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43704504504504504, + "grad_norm": 24.272926330566406, + "learning_rate": 6.397727066477348e-06, + "loss": 6.8069071769714355, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.2936110608279705, + "loss_total_v6": 0.4254317134618759, + "step": 1895 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43727567567567566, + "grad_norm": 23.504697799682617, + "learning_rate": 6.394080211590381e-06, + "loss": 6.854557037353516, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.29527687653899193, + "loss_total_v6": 0.42840979993343353, + "step": 1896 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4375063063063063, + "grad_norm": 22.455781936645508, + "learning_rate": 6.390432552387934e-06, + "loss": 7.151905059814453, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.31311342865228653, + "loss_total_v6": 0.44699404761195183, + "step": 1897 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43773693693693694, + "grad_norm": 29.51924705505371, + "learning_rate": 6.3867840909745284e-06, + "loss": 6.954897880554199, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.29971713572740555, + "loss_total_v6": 0.43468113243579865, + "step": 1898 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43796756756756755, + "grad_norm": 25.308013916015625, + "learning_rate": 6.383134829455142e-06, + "loss": 6.590997219085693, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2800098452717066, + "loss_total_v6": 0.41193732619285583, + "step": 1899 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4381981981981982, + "grad_norm": 24.784215927124023, + "learning_rate": 6.379484769935223e-06, + "loss": 6.922664642333984, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3040349557995796, + "loss_total_v6": 0.4326665364205837, + "step": 1900 + }, + { + "epoch": 0.4381981981981982, + "eval_loss": 0.5918627381324768, + "eval_loss_alignment": 0.2394673944063927, + "eval_loss_alignment_w": 0.11973369720319635, + "eval_loss_sft": 0.4721289865893724, + "eval_loss_total_v6": 0.5918626832227184, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.9953, + "eval_samples_per_second": 19.177, + "eval_steps_per_second": 2.407, + "step": 1900 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43842882882882883, + "grad_norm": 27.393295288085938, + "learning_rate": 6.3758339145206745e-06, + "loss": 7.014212131500244, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.30716267228126526, + "loss_total_v6": 0.43838825449347496, + "step": 1901 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43865945945945944, + "grad_norm": 24.47123908996582, + "learning_rate": 6.372182265317861e-06, + "loss": 7.257224082946777, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.31911603547632694, + "loss_total_v6": 0.4535764902830124, + "step": 1902 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4388900900900901, + "grad_norm": 24.782875061035156, + "learning_rate": 6.3685298244336e-06, + "loss": 7.257795333862305, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.32079968228936195, + "loss_total_v6": 0.45361217856407166, + "step": 1903 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4391207207207207, + "grad_norm": 26.37970733642578, + "learning_rate": 6.364876593975174e-06, + "loss": 6.689393043518066, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.28548818826675415, + "loss_total_v6": 0.41808704286813736, + "step": 1904 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.43935135135135134, + "grad_norm": 22.692411422729492, + "learning_rate": 6.361222576050312e-06, + "loss": 6.92606258392334, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3001884985715151, + "loss_total_v6": 0.43287893757224083, + "step": 1905 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.439581981981982, + "grad_norm": 24.511737823486328, + "learning_rate": 6.357567772767207e-06, + "loss": 6.984111785888672, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3030383586883545, + "loss_total_v6": 0.4365069828927517, + "step": 1906 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4398126126126126, + "grad_norm": 27.638954162597656, + "learning_rate": 6.353912186234498e-06, + "loss": 6.9141740798950195, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.2970803566277027, + "loss_total_v6": 0.4321358948945999, + "step": 1907 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44004324324324323, + "grad_norm": 25.139860153198242, + "learning_rate": 6.3502558185612775e-06, + "loss": 7.04292106628418, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.30741582810878754, + "loss_total_v6": 0.44018255546689034, + "step": 1908 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4402738738738739, + "grad_norm": 23.457212448120117, + "learning_rate": 6.346598671857088e-06, + "loss": 6.812457084655762, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.2908756136894226, + "loss_total_v6": 0.4257785677909851, + "step": 1909 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4405045045045045, + "grad_norm": 24.028846740722656, + "learning_rate": 6.342940748231925e-06, + "loss": 7.115517616271973, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3126702979207039, + "loss_total_v6": 0.4447198621928692, + "step": 1910 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4407351351351351, + "grad_norm": 27.540315628051758, + "learning_rate": 6.3392820497962305e-06, + "loss": 6.861906051635742, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.2956293672323227, + "loss_total_v6": 0.4288691133260727, + "step": 1911 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44096576576576574, + "grad_norm": 23.361425399780273, + "learning_rate": 6.335622578660889e-06, + "loss": 6.888416767120361, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.299330972135067, + "loss_total_v6": 0.4305260442197323, + "step": 1912 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4411963963963964, + "grad_norm": 24.842737197875977, + "learning_rate": 6.331962336937239e-06, + "loss": 7.1763529777526855, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3152517732232809, + "loss_total_v6": 0.44852205738425255, + "step": 1913 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.441427027027027, + "grad_norm": 24.216960906982422, + "learning_rate": 6.3283013267370576e-06, + "loss": 6.926823139190674, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3019144553691149, + "loss_total_v6": 0.4329264387488365, + "step": 1914 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44165765765765763, + "grad_norm": 24.450439453125, + "learning_rate": 6.324639550172567e-06, + "loss": 6.9142303466796875, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.2974958308041096, + "loss_total_v6": 0.43213939666748047, + "step": 1915 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4418882882882883, + "grad_norm": 23.063642501831055, + "learning_rate": 6.3209770093564315e-06, + "loss": 7.308589458465576, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.3202817142009735, + "loss_total_v6": 0.4567868337035179, + "step": 1916 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4421189189189189, + "grad_norm": 22.810754776000977, + "learning_rate": 6.317313706401755e-06, + "loss": 7.123170852661133, + "loss_alignment": 0.2686767578125, + "loss_alignment_w": 0.13433837890625, + "loss_sft": 0.3108903169631958, + "loss_total_v6": 0.4451981820166111, + "step": 1917 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4423495495495495, + "grad_norm": 24.469497680664062, + "learning_rate": 6.313649643422085e-06, + "loss": 7.166446685791016, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3144342564046383, + "loss_total_v6": 0.4479028917849064, + "step": 1918 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4425801801801802, + "grad_norm": 26.94784927368164, + "learning_rate": 6.309984822531404e-06, + "loss": 7.118433952331543, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3091752156615257, + "loss_total_v6": 0.44490212947130203, + "step": 1919 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4428108108108108, + "grad_norm": 22.174964904785156, + "learning_rate": 6.306319245844134e-06, + "loss": 7.311246395111084, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3240335676819086, + "loss_total_v6": 0.45695288851857185, + "step": 1920 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4430414414414414, + "grad_norm": 25.357166290283203, + "learning_rate": 6.302652915475129e-06, + "loss": 6.905013084411621, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3011617287993431, + "loss_total_v6": 0.4315633289515972, + "step": 1921 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4432720720720721, + "grad_norm": 22.009973526000977, + "learning_rate": 6.298985833539681e-06, + "loss": 6.649735450744629, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.28438290022313595, + "loss_total_v6": 0.4156084768474102, + "step": 1922 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4435027027027027, + "grad_norm": 22.68419647216797, + "learning_rate": 6.29531800215352e-06, + "loss": 6.723941802978516, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.2876475006341934, + "loss_total_v6": 0.4202463701367378, + "step": 1923 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4437333333333333, + "grad_norm": 24.86988067626953, + "learning_rate": 6.2916494234327985e-06, + "loss": 7.238757133483887, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3196250796318054, + "loss_total_v6": 0.4524223245680332, + "step": 1924 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.443963963963964, + "grad_norm": 24.226993560791016, + "learning_rate": 6.287980099494106e-06, + "loss": 6.982356071472168, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3056599572300911, + "loss_total_v6": 0.4363972619175911, + "step": 1925 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4441945945945946, + "grad_norm": 25.033166885375977, + "learning_rate": 6.284310032454463e-06, + "loss": 6.688390731811523, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.2874702289700508, + "loss_total_v6": 0.4180244207382202, + "step": 1926 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4444252252252252, + "grad_norm": 22.448200225830078, + "learning_rate": 6.280639224431317e-06, + "loss": 7.204029083251953, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.31702734529972076, + "loss_total_v6": 0.45025182887911797, + "step": 1927 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4446558558558559, + "grad_norm": 29.657960891723633, + "learning_rate": 6.276967677542542e-06, + "loss": 6.756890773773193, + "loss_alignment": 0.2667236328125, + "loss_alignment_w": 0.13336181640625, + "loss_sft": 0.2891727313399315, + "loss_total_v6": 0.4223056733608246, + "step": 1928 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4448864864864865, + "grad_norm": 22.23516845703125, + "learning_rate": 6.2732953939064416e-06, + "loss": 6.75520133972168, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.291554344817996, + "loss_total_v6": 0.4222000800073147, + "step": 1929 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4451171171171171, + "grad_norm": 25.599063873291016, + "learning_rate": 6.269622375641739e-06, + "loss": 7.326275825500488, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.3262546472251415, + "loss_total_v6": 0.4578922316431999, + "step": 1930 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44534774774774777, + "grad_norm": 23.442073822021484, + "learning_rate": 6.265948624867585e-06, + "loss": 7.434418678283691, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.32948880456387997, + "loss_total_v6": 0.4646511599421501, + "step": 1931 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4455783783783784, + "grad_norm": 28.524869918823242, + "learning_rate": 6.262274143703555e-06, + "loss": 7.384803771972656, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3290123827755451, + "loss_total_v6": 0.461550235748291, + "step": 1932 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.445809009009009, + "grad_norm": 26.652332305908203, + "learning_rate": 6.258598934269642e-06, + "loss": 7.210655689239502, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.31893687322735786, + "loss_total_v6": 0.45066599547863007, + "step": 1933 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44603963963963966, + "grad_norm": 24.554866790771484, + "learning_rate": 6.254922998686259e-06, + "loss": 6.884471893310547, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.2957732602953911, + "loss_total_v6": 0.4302794896066189, + "step": 1934 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4462702702702703, + "grad_norm": 26.73832893371582, + "learning_rate": 6.251246339074241e-06, + "loss": 7.311498641967773, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.32317959889769554, + "loss_total_v6": 0.45696864649653435, + "step": 1935 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4465009009009009, + "grad_norm": 27.76730728149414, + "learning_rate": 6.24756895755484e-06, + "loss": 6.738776206970215, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.2881626542657614, + "loss_total_v6": 0.421173520386219, + "step": 1936 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44673153153153156, + "grad_norm": 22.81801986694336, + "learning_rate": 6.24389085624972e-06, + "loss": 7.301653861999512, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.32502099499106407, + "loss_total_v6": 0.4563533812761307, + "step": 1937 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44696216216216217, + "grad_norm": 26.253833770751953, + "learning_rate": 6.240212037280967e-06, + "loss": 7.121905326843262, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.31320685520768166, + "loss_total_v6": 0.44511908665299416, + "step": 1938 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4471927927927928, + "grad_norm": 23.371315002441406, + "learning_rate": 6.236532502771078e-06, + "loss": 6.990998268127441, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3051319755613804, + "loss_total_v6": 0.4369373992085457, + "step": 1939 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44742342342342345, + "grad_norm": 24.301891326904297, + "learning_rate": 6.232852254842961e-06, + "loss": 7.09238338470459, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.31162112206220627, + "loss_total_v6": 0.44327395781874657, + "step": 1940 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44765405405405406, + "grad_norm": 24.007158279418945, + "learning_rate": 6.22917129561994e-06, + "loss": 6.750639915466309, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.2897433526813984, + "loss_total_v6": 0.421914990991354, + "step": 1941 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4478846846846847, + "grad_norm": 23.071489334106445, + "learning_rate": 6.225489627225746e-06, + "loss": 7.123595237731934, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.3096961621195078, + "loss_total_v6": 0.44522472098469734, + "step": 1942 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4481153153153153, + "grad_norm": 21.905818939208984, + "learning_rate": 6.221807251784517e-06, + "loss": 6.90367317199707, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.2961035966873169, + "loss_total_v6": 0.4314795695245266, + "step": 1943 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44834594594594596, + "grad_norm": 24.720754623413086, + "learning_rate": 6.218124171420806e-06, + "loss": 6.799471378326416, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.29471792466938496, + "loss_total_v6": 0.4249669574201107, + "step": 1944 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44857657657657657, + "grad_norm": 24.938310623168945, + "learning_rate": 6.214440388259565e-06, + "loss": 7.095900535583496, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3093843013048172, + "loss_total_v6": 0.4434937983751297, + "step": 1945 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4488072072072072, + "grad_norm": 26.573808670043945, + "learning_rate": 6.210755904426159e-06, + "loss": 6.937331199645996, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.29890912026166916, + "loss_total_v6": 0.43358318135142326, + "step": 1946 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44903783783783785, + "grad_norm": 23.33753776550293, + "learning_rate": 6.20707072204635e-06, + "loss": 7.146629333496094, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.31351612135767937, + "loss_total_v6": 0.44666432216763496, + "step": 1947 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44926846846846846, + "grad_norm": 23.995948791503906, + "learning_rate": 6.203384843246307e-06, + "loss": 7.062253952026367, + "loss_alignment": 0.2669677734375, + "loss_alignment_w": 0.13348388671875, + "loss_sft": 0.30769334360957146, + "loss_total_v6": 0.44139086455106735, + "step": 1948 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4494990990990991, + "grad_norm": 22.905324935913086, + "learning_rate": 6.199698270152602e-06, + "loss": 7.111231803894043, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.312814399600029, + "loss_total_v6": 0.444451991468668, + "step": 1949 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44972972972972974, + "grad_norm": 26.213708877563477, + "learning_rate": 6.196011004892204e-06, + "loss": 7.38568115234375, + "loss_alignment": 0.2691650390625, + "loss_alignment_w": 0.13458251953125, + "loss_sft": 0.3269615415483713, + "loss_total_v6": 0.4616050757467747, + "step": 1950 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.44996036036036036, + "grad_norm": 26.067468643188477, + "learning_rate": 6.192323049592483e-06, + "loss": 7.105403900146484, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3093831390142441, + "loss_total_v6": 0.4440877325832844, + "step": 1951 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45019099099099097, + "grad_norm": 25.718673706054688, + "learning_rate": 6.188634406381206e-06, + "loss": 6.8559699058532715, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.29399189725518227, + "loss_total_v6": 0.42849811911582947, + "step": 1952 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45042162162162164, + "grad_norm": 25.672025680541992, + "learning_rate": 6.1849450773865424e-06, + "loss": 7.107141971588135, + "loss_alignment": 0.2652587890625, + "loss_alignment_w": 0.13262939453125, + "loss_sft": 0.3115212209522724, + "loss_total_v6": 0.444196380674839, + "step": 1953 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45065225225225225, + "grad_norm": 24.398456573486328, + "learning_rate": 6.181255064737051e-06, + "loss": 7.052391052246094, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.30828237906098366, + "loss_total_v6": 0.44077444076538086, + "step": 1954 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45088288288288286, + "grad_norm": 23.56407928466797, + "learning_rate": 6.177564370561685e-06, + "loss": 6.77142333984375, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.2902946416288614, + "loss_total_v6": 0.42321396619081497, + "step": 1955 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45111351351351353, + "grad_norm": 23.164127349853516, + "learning_rate": 6.173872996989794e-06, + "loss": 7.057768821716309, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.31217378564178944, + "loss_total_v6": 0.4411105550825596, + "step": 1956 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45134414414414414, + "grad_norm": 24.615848541259766, + "learning_rate": 6.170180946151118e-06, + "loss": 7.026640892028809, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.30661197006702423, + "loss_total_v6": 0.43916506692767143, + "step": 1957 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45157477477477476, + "grad_norm": 25.17417335510254, + "learning_rate": 6.166488220175789e-06, + "loss": 7.004122734069824, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.3057233728468418, + "loss_total_v6": 0.437757670879364, + "step": 1958 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4518054054054054, + "grad_norm": 24.851398468017578, + "learning_rate": 6.162794821194327e-06, + "loss": 6.889999866485596, + "loss_alignment": 0.2684326171875, + "loss_alignment_w": 0.13421630859375, + "loss_sft": 0.29671385139226913, + "loss_total_v6": 0.43062498793005943, + "step": 1959 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45203603603603604, + "grad_norm": 26.284866333007812, + "learning_rate": 6.1591007513376425e-06, + "loss": 6.629759788513184, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.28144068270921707, + "loss_total_v6": 0.41435999795794487, + "step": 1960 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45226666666666665, + "grad_norm": 24.522733688354492, + "learning_rate": 6.1554060127370285e-06, + "loss": 7.516592025756836, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.33474669605493546, + "loss_total_v6": 0.46978697553277016, + "step": 1961 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4524972972972973, + "grad_norm": 27.341388702392578, + "learning_rate": 6.151710607524169e-06, + "loss": 7.049208164215088, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3088769093155861, + "loss_total_v6": 0.4405755139887333, + "step": 1962 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45272792792792793, + "grad_norm": 23.719711303710938, + "learning_rate": 6.148014537831131e-06, + "loss": 7.082198143005371, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3087720051407814, + "loss_total_v6": 0.4426373690366745, + "step": 1963 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45295855855855854, + "grad_norm": 25.059247970581055, + "learning_rate": 6.144317805790362e-06, + "loss": 7.206650733947754, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.3168402314186096, + "loss_total_v6": 0.450415663421154, + "step": 1964 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4531891891891892, + "grad_norm": 25.911972045898438, + "learning_rate": 6.140620413534695e-06, + "loss": 6.858379364013672, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.2980182357132435, + "loss_total_v6": 0.4286487251520157, + "step": 1965 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4534198198198198, + "grad_norm": 25.483606338500977, + "learning_rate": 6.136922363197344e-06, + "loss": 6.9317779541015625, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.30304815620183945, + "loss_total_v6": 0.43323614448308945, + "step": 1966 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45365045045045044, + "grad_norm": 25.08753776550293, + "learning_rate": 6.1332236569119e-06, + "loss": 7.139645576477051, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.31623823568224907, + "loss_total_v6": 0.44622786343097687, + "step": 1967 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4538810810810811, + "grad_norm": 23.046228408813477, + "learning_rate": 6.129524296812335e-06, + "loss": 6.296773910522461, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.2634061798453331, + "loss_total_v6": 0.3935483768582344, + "step": 1968 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4541117117117117, + "grad_norm": 28.044301986694336, + "learning_rate": 6.125824285032995e-06, + "loss": 6.711017608642578, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.28824351355433464, + "loss_total_v6": 0.41943858191370964, + "step": 1969 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45434234234234233, + "grad_norm": 23.379474639892578, + "learning_rate": 6.122123623708608e-06, + "loss": 7.295567035675049, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.31955934315919876, + "loss_total_v6": 0.45597292482852936, + "step": 1970 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.454572972972973, + "grad_norm": 24.378141403198242, + "learning_rate": 6.118422314974269e-06, + "loss": 6.9995527267456055, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.30537667870521545, + "loss_total_v6": 0.43747202679514885, + "step": 1971 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4548036036036036, + "grad_norm": 23.584186553955078, + "learning_rate": 6.1147203609654535e-06, + "loss": 7.368782043457031, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.32988787442445755, + "loss_total_v6": 0.46054888889193535, + "step": 1972 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4550342342342342, + "grad_norm": 24.908411026000977, + "learning_rate": 6.111017763818005e-06, + "loss": 7.160673141479492, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.31797968968749046, + "loss_total_v6": 0.44754207506775856, + "step": 1973 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45526486486486484, + "grad_norm": 23.46230125427246, + "learning_rate": 6.107314525668139e-06, + "loss": 7.0685133934021, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.30895430222153664, + "loss_total_v6": 0.44178207218647003, + "step": 1974 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4554954954954955, + "grad_norm": 26.357128143310547, + "learning_rate": 6.103610648652442e-06, + "loss": 7.498683929443359, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.33481762930750847, + "loss_total_v6": 0.46866773813962936, + "step": 1975 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4557261261261261, + "grad_norm": 22.263702392578125, + "learning_rate": 6.099906134907868e-06, + "loss": 7.214626312255859, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.3165299817919731, + "loss_total_v6": 0.4509141333401203, + "step": 1976 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45595675675675673, + "grad_norm": 23.683120727539062, + "learning_rate": 6.096200986571737e-06, + "loss": 6.919312477111816, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.30191807076334953, + "loss_total_v6": 0.43245701491832733, + "step": 1977 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4561873873873874, + "grad_norm": 22.836915969848633, + "learning_rate": 6.0924952057817385e-06, + "loss": 6.863499641418457, + "loss_alignment": 0.2652587890625, + "loss_alignment_w": 0.13262939453125, + "loss_sft": 0.29630882292985916, + "loss_total_v6": 0.42896873131394386, + "step": 1978 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.456418018018018, + "grad_norm": 26.213451385498047, + "learning_rate": 6.088788794675926e-06, + "loss": 7.117647171020508, + "loss_alignment": 0.2698974609375, + "loss_alignment_w": 0.13494873046875, + "loss_sft": 0.30978211387991905, + "loss_total_v6": 0.44485292211174965, + "step": 1979 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4566486486486486, + "grad_norm": 23.342018127441406, + "learning_rate": 6.0850817553927145e-06, + "loss": 6.573819160461426, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2788599245250225, + "loss_total_v6": 0.4108637049794197, + "step": 1980 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4568792792792793, + "grad_norm": 22.65999412536621, + "learning_rate": 6.081374090070882e-06, + "loss": 6.7553791999816895, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.290543120354414, + "loss_total_v6": 0.4222111999988556, + "step": 1981 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4571099099099099, + "grad_norm": 24.89232635498047, + "learning_rate": 6.077665800849568e-06, + "loss": 6.768157958984375, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.29099082574248314, + "loss_total_v6": 0.42300987616181374, + "step": 1982 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4573405405405405, + "grad_norm": 26.14310646057129, + "learning_rate": 6.073956889868275e-06, + "loss": 6.8234477043151855, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.29960391111671925, + "loss_total_v6": 0.4264654740691185, + "step": 1983 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4575711711711712, + "grad_norm": 22.323583602905273, + "learning_rate": 6.07024735926686e-06, + "loss": 6.484692096710205, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.27327422611415386, + "loss_total_v6": 0.4052932634949684, + "step": 1984 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4578018018018018, + "grad_norm": 23.245731353759766, + "learning_rate": 6.066537211185539e-06, + "loss": 7.305527687072754, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.32105163112282753, + "loss_total_v6": 0.45659545809030533, + "step": 1985 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4580324324324324, + "grad_norm": 28.053569793701172, + "learning_rate": 6.062826447764883e-06, + "loss": 6.531734943389893, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.27925090305507183, + "loss_total_v6": 0.4082334376871586, + "step": 1986 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4582630630630631, + "grad_norm": 28.543941497802734, + "learning_rate": 6.059115071145824e-06, + "loss": 7.110048770904541, + "loss_alignment": 0.275634765625, + "loss_alignment_w": 0.1378173828125, + "loss_sft": 0.30660643987357616, + "loss_total_v6": 0.4443780519068241, + "step": 1987 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4584936936936937, + "grad_norm": 27.00665855407715, + "learning_rate": 6.055403083469637e-06, + "loss": 6.9459662437438965, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.3000744357705116, + "loss_total_v6": 0.43412288278341293, + "step": 1988 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4587243243243243, + "grad_norm": 24.707077026367188, + "learning_rate": 6.051690486877961e-06, + "loss": 7.2390666007995605, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.3173098359256983, + "loss_total_v6": 0.45244167372584343, + "step": 1989 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.458954954954955, + "grad_norm": 24.54672622680664, + "learning_rate": 6.047977283512777e-06, + "loss": 7.0736589431762695, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.31165628880262375, + "loss_total_v6": 0.44210368022322655, + "step": 1990 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4591855855855856, + "grad_norm": 26.243606567382812, + "learning_rate": 6.044263475516421e-06, + "loss": 6.807218551635742, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.2954005151987076, + "loss_total_v6": 0.4254511632025242, + "step": 1991 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4594162162162162, + "grad_norm": 27.375473022460938, + "learning_rate": 6.04054906503158e-06, + "loss": 6.956539154052734, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3031156025826931, + "loss_total_v6": 0.4347836785018444, + "step": 1992 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.45964684684684687, + "grad_norm": 27.85314178466797, + "learning_rate": 6.036834054201283e-06, + "loss": 6.882051467895508, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.2981854844838381, + "loss_total_v6": 0.43012822419404984, + "step": 1993 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4598774774774775, + "grad_norm": 27.129762649536133, + "learning_rate": 6.0331184451689075e-06, + "loss": 7.822778701782227, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.35444797948002815, + "loss_total_v6": 0.48892366886138916, + "step": 1994 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4601081081081081, + "grad_norm": 23.855369567871094, + "learning_rate": 6.029402240078176e-06, + "loss": 6.9490861892700195, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.30510647408664227, + "loss_total_v6": 0.4343178831040859, + "step": 1995 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46033873873873876, + "grad_norm": 31.871644973754883, + "learning_rate": 6.025685441073157e-06, + "loss": 7.222748756408691, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3184567242860794, + "loss_total_v6": 0.4514218159019947, + "step": 1996 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4605693693693694, + "grad_norm": 25.539167404174805, + "learning_rate": 6.021968050298261e-06, + "loss": 7.195592880249023, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.31601179018616676, + "loss_total_v6": 0.44972456619143486, + "step": 1997 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4608, + "grad_norm": 22.71480369567871, + "learning_rate": 6.018250069898236e-06, + "loss": 6.841350078582764, + "loss_alignment": 0.2637939453125, + "loss_alignment_w": 0.13189697265625, + "loss_sft": 0.29570265486836433, + "loss_total_v6": 0.42758436501026154, + "step": 1998 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46103063063063066, + "grad_norm": 24.64288330078125, + "learning_rate": 6.014531502018175e-06, + "loss": 6.938502311706543, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.30327002704143524, + "loss_total_v6": 0.43365639820694923, + "step": 1999 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46126126126126127, + "grad_norm": 23.11226463317871, + "learning_rate": 6.0108123488035094e-06, + "loss": 7.140623569488525, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3114318083971739, + "loss_total_v6": 0.44628897681832314, + "step": 2000 + }, + { + "epoch": 0.46126126126126127, + "eval_loss": 0.5869500041007996, + "eval_loss_alignment": 0.23871825057077625, + "eval_loss_alignment_w": 0.11935912528538813, + "eval_loss_sft": 0.4675908713223047, + "eval_loss_total_v6": 0.5869499966332085, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.689, + "eval_samples_per_second": 19.242, + "eval_steps_per_second": 2.415, + "step": 2000 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4614918918918919, + "grad_norm": 27.152414321899414, + "learning_rate": 6.007092612400005e-06, + "loss": 7.38040018081665, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.32808104157447815, + "loss_total_v6": 0.46127500757575035, + "step": 2001 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46172252252252255, + "grad_norm": 21.54594612121582, + "learning_rate": 6.003372294953767e-06, + "loss": 6.72127103805542, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.29102060571312904, + "loss_total_v6": 0.42007944360375404, + "step": 2002 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46195315315315316, + "grad_norm": 25.185441970825195, + "learning_rate": 5.999651398611235e-06, + "loss": 6.998166084289551, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3044660519808531, + "loss_total_v6": 0.43738536536693573, + "step": 2003 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4621837837837838, + "grad_norm": 24.046749114990234, + "learning_rate": 5.995929925519181e-06, + "loss": 6.904596328735352, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.298587441444397, + "loss_total_v6": 0.43153727799654007, + "step": 2004 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4624144144144144, + "grad_norm": 28.29087257385254, + "learning_rate": 5.992207877824714e-06, + "loss": 7.095261096954346, + "loss_alignment": 0.272705078125, + "loss_alignment_w": 0.1363525390625, + "loss_sft": 0.30708601512014866, + "loss_total_v6": 0.4434538148343563, + "step": 2005 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46264504504504506, + "grad_norm": 23.648479461669922, + "learning_rate": 5.98848525767527e-06, + "loss": 7.2795586585998535, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.32618824765086174, + "loss_total_v6": 0.45497242361307144, + "step": 2006 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46287567567567567, + "grad_norm": 25.567581176757812, + "learning_rate": 5.984762067218617e-06, + "loss": 6.683795928955078, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.28611493296921253, + "loss_total_v6": 0.4177372567355633, + "step": 2007 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4631063063063063, + "grad_norm": 23.78334617614746, + "learning_rate": 5.981038308602853e-06, + "loss": 6.6786956787109375, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.28657436184585094, + "loss_total_v6": 0.4174184873700142, + "step": 2008 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46333693693693695, + "grad_norm": 24.070613861083984, + "learning_rate": 5.977313983976403e-06, + "loss": 6.966923713684082, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3018878139555454, + "loss_total_v6": 0.4354327321052551, + "step": 2009 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46356756756756756, + "grad_norm": 24.684091567993164, + "learning_rate": 5.973589095488017e-06, + "loss": 6.3753275871276855, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.26877352595329285, + "loss_total_v6": 0.39845798164606094, + "step": 2010 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4637981981981982, + "grad_norm": 23.6990966796875, + "learning_rate": 5.969863645286774e-06, + "loss": 6.520694732666016, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2751734238117933, + "loss_total_v6": 0.407543420791626, + "step": 2011 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46402882882882884, + "grad_norm": 24.9177303314209, + "learning_rate": 5.966137635522074e-06, + "loss": 7.0862507820129395, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.30833867006003857, + "loss_total_v6": 0.442890677601099, + "step": 2012 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46425945945945946, + "grad_norm": 22.71620750427246, + "learning_rate": 5.96241106834364e-06, + "loss": 6.830596923828125, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.2950306087732315, + "loss_total_v6": 0.4269123189151287, + "step": 2013 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46449009009009007, + "grad_norm": 24.659467697143555, + "learning_rate": 5.958683945901517e-06, + "loss": 6.813092231750488, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.29196817614138126, + "loss_total_v6": 0.4258182570338249, + "step": 2014 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46472072072072074, + "grad_norm": 24.165843963623047, + "learning_rate": 5.954956270346074e-06, + "loss": 7.090524196624756, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.309292396530509, + "loss_total_v6": 0.44315775856375694, + "step": 2015 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46495135135135135, + "grad_norm": 28.763551712036133, + "learning_rate": 5.951228043827991e-06, + "loss": 6.860273838043213, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.2961529828608036, + "loss_total_v6": 0.4287671111524105, + "step": 2016 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46518198198198196, + "grad_norm": 25.728944778442383, + "learning_rate": 5.947499268498274e-06, + "loss": 6.992938995361328, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3052074871957302, + "loss_total_v6": 0.4370586797595024, + "step": 2017 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46541261261261263, + "grad_norm": 26.974781036376953, + "learning_rate": 5.943769946508242e-06, + "loss": 7.00714635848999, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3082469515502453, + "loss_total_v6": 0.4379466511309147, + "step": 2018 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46564324324324324, + "grad_norm": 25.31604766845703, + "learning_rate": 5.9400400800095295e-06, + "loss": 7.130237579345703, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31415487080812454, + "loss_total_v6": 0.44563983753323555, + "step": 2019 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46587387387387386, + "grad_norm": 22.11109733581543, + "learning_rate": 5.936309671154085e-06, + "loss": 6.701456546783447, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.28662362694740295, + "loss_total_v6": 0.41884103044867516, + "step": 2020 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4661045045045045, + "grad_norm": 25.817901611328125, + "learning_rate": 5.932578722094169e-06, + "loss": 6.5780158042907715, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.28226551599800587, + "loss_total_v6": 0.4111259989440441, + "step": 2021 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46633513513513514, + "grad_norm": 25.63289451599121, + "learning_rate": 5.928847234982358e-06, + "loss": 7.472997188568115, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3320678398013115, + "loss_total_v6": 0.4670623317360878, + "step": 2022 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46656576576576575, + "grad_norm": 27.083351135253906, + "learning_rate": 5.925115211971533e-06, + "loss": 7.259912490844727, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3222442716360092, + "loss_total_v6": 0.4537445269525051, + "step": 2023 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4667963963963964, + "grad_norm": 31.479204177856445, + "learning_rate": 5.921382655214889e-06, + "loss": 7.211997032165527, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.31705233082175255, + "loss_total_v6": 0.45074982941150665, + "step": 2024 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46702702702702703, + "grad_norm": 23.805553436279297, + "learning_rate": 5.917649566865926e-06, + "loss": 6.80968713760376, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.29035154543817043, + "loss_total_v6": 0.4256054498255253, + "step": 2025 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46725765765765764, + "grad_norm": 23.091472625732422, + "learning_rate": 5.913915949078453e-06, + "loss": 7.043946743011475, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.307998713105917, + "loss_total_v6": 0.44024666026234627, + "step": 2026 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4674882882882883, + "grad_norm": 25.531085968017578, + "learning_rate": 5.910181804006582e-06, + "loss": 6.785195350646973, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.29097230546176434, + "loss_total_v6": 0.4240747354924679, + "step": 2027 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4677189189189189, + "grad_norm": 25.184120178222656, + "learning_rate": 5.906447133804731e-06, + "loss": 6.945244789123535, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30279117822647095, + "loss_total_v6": 0.43407779932022095, + "step": 2028 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46794954954954954, + "grad_norm": 26.82723617553711, + "learning_rate": 5.902711940627619e-06, + "loss": 6.847970485687256, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.2928510494530201, + "loss_total_v6": 0.4279981628060341, + "step": 2029 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4681801801801802, + "grad_norm": 23.293746948242188, + "learning_rate": 5.8989762266302685e-06, + "loss": 6.730548858642578, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2903644945472479, + "loss_total_v6": 0.42065928876399994, + "step": 2030 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4684108108108108, + "grad_norm": 26.801841735839844, + "learning_rate": 5.895239993968005e-06, + "loss": 6.758800029754639, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.2911841459572315, + "loss_total_v6": 0.4224249832332134, + "step": 2031 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46864144144144143, + "grad_norm": 23.227012634277344, + "learning_rate": 5.891503244796448e-06, + "loss": 6.960421562194824, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3012983165681362, + "loss_total_v6": 0.435026329010725, + "step": 2032 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4688720720720721, + "grad_norm": 24.391433715820312, + "learning_rate": 5.887765981271518e-06, + "loss": 7.360490322113037, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.32672987319529057, + "loss_total_v6": 0.4600306563079357, + "step": 2033 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4691027027027027, + "grad_norm": 26.376596450805664, + "learning_rate": 5.884028205549429e-06, + "loss": 7.038463592529297, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.30724405869841576, + "loss_total_v6": 0.43990397080779076, + "step": 2034 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4693333333333333, + "grad_norm": 24.0494441986084, + "learning_rate": 5.880289919786697e-06, + "loss": 7.1265764236450195, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3148568496108055, + "loss_total_v6": 0.4454110525548458, + "step": 2035 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.46956396396396394, + "grad_norm": 24.393835067749023, + "learning_rate": 5.876551126140125e-06, + "loss": 7.1458740234375, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.31499479338526726, + "loss_total_v6": 0.44661712646484375, + "step": 2036 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4697945945945946, + "grad_norm": 23.35833168029785, + "learning_rate": 5.872811826766817e-06, + "loss": 6.965863227844238, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3023250922560692, + "loss_total_v6": 0.4353664629161358, + "step": 2037 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4700252252252252, + "grad_norm": 24.310396194458008, + "learning_rate": 5.869072023824158e-06, + "loss": 6.656691551208496, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.28251355700194836, + "loss_total_v6": 0.4160432182252407, + "step": 2038 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47025585585585583, + "grad_norm": 26.754074096679688, + "learning_rate": 5.865331719469836e-06, + "loss": 7.121747016906738, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.31434137374162674, + "loss_total_v6": 0.44510921463370323, + "step": 2039 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4704864864864865, + "grad_norm": 24.65299415588379, + "learning_rate": 5.861590915861816e-06, + "loss": 6.989166259765625, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3058261703699827, + "loss_total_v6": 0.43682287260890007, + "step": 2040 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4707171171171171, + "grad_norm": 22.133861541748047, + "learning_rate": 5.857849615158363e-06, + "loss": 7.026102066040039, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.30996572598814964, + "loss_total_v6": 0.43913137167692184, + "step": 2041 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4709477477477477, + "grad_norm": 24.618114471435547, + "learning_rate": 5.8541078195180175e-06, + "loss": 7.224668025970459, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.31975157558918, + "loss_total_v6": 0.4515417478978634, + "step": 2042 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4711783783783784, + "grad_norm": 25.041505813598633, + "learning_rate": 5.850365531099614e-06, + "loss": 6.855319976806641, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.29495833814144135, + "loss_total_v6": 0.42845748737454414, + "step": 2043 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.471409009009009, + "grad_norm": 26.940536499023438, + "learning_rate": 5.846622752062268e-06, + "loss": 7.641262054443359, + "loss_alignment": 0.2694091796875, + "loss_alignment_w": 0.13470458984375, + "loss_sft": 0.34278276190161705, + "loss_total_v6": 0.47757888957858086, + "step": 2044 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4716396396396396, + "grad_norm": 24.888593673706055, + "learning_rate": 5.842879484565379e-06, + "loss": 6.615157604217529, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.2833356652408838, + "loss_total_v6": 0.4134473390877247, + "step": 2045 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4718702702702703, + "grad_norm": 24.21483039855957, + "learning_rate": 5.839135730768626e-06, + "loss": 7.295230865478516, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.32402442395687103, + "loss_total_v6": 0.45595191791653633, + "step": 2046 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4721009009009009, + "grad_norm": 25.955509185791016, + "learning_rate": 5.835391492831969e-06, + "loss": 6.749453544616699, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.287532951682806, + "loss_total_v6": 0.4218408428132534, + "step": 2047 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4723315315315315, + "grad_norm": 23.29521369934082, + "learning_rate": 5.831646772915651e-06, + "loss": 6.911055564880371, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3010968454182148, + "loss_total_v6": 0.4319409616291523, + "step": 2048 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4725621621621622, + "grad_norm": 25.862722396850586, + "learning_rate": 5.827901573180189e-06, + "loss": 7.076732158660889, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.30759118869900703, + "loss_total_v6": 0.44229576736688614, + "step": 2049 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4727927927927928, + "grad_norm": 24.299253463745117, + "learning_rate": 5.824155895786381e-06, + "loss": 7.052656650543213, + "loss_alignment": 0.2637939453125, + "loss_alignment_w": 0.13189697265625, + "loss_sft": 0.30893984250724316, + "loss_total_v6": 0.4407910406589508, + "step": 2050 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4730234234234234, + "grad_norm": 24.086803436279297, + "learning_rate": 5.8204097428952945e-06, + "loss": 7.358859062194824, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3250257261097431, + "loss_total_v6": 0.4599286802113056, + "step": 2051 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4732540540540541, + "grad_norm": 23.29581642150879, + "learning_rate": 5.816663116668276e-06, + "loss": 7.035894870758057, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.30632057040929794, + "loss_total_v6": 0.43974341824650764, + "step": 2052 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4734846846846847, + "grad_norm": 26.45049285888672, + "learning_rate": 5.812916019266946e-06, + "loss": 6.939155578613281, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3042416535317898, + "loss_total_v6": 0.4336972162127495, + "step": 2053 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4737153153153153, + "grad_norm": 23.575468063354492, + "learning_rate": 5.809168452853193e-06, + "loss": 7.287827491760254, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3217154033482075, + "loss_total_v6": 0.45548921450972557, + "step": 2054 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47394594594594597, + "grad_norm": 22.29548454284668, + "learning_rate": 5.8054204195891775e-06, + "loss": 7.076655387878418, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.3076169118285179, + "loss_total_v6": 0.4422909580171108, + "step": 2055 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4741765765765766, + "grad_norm": 23.733642578125, + "learning_rate": 5.801671921637329e-06, + "loss": 7.015883445739746, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.30659576691687107, + "loss_total_v6": 0.438492726534605, + "step": 2056 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4744072072072072, + "grad_norm": 23.94719123840332, + "learning_rate": 5.797922961160348e-06, + "loss": 7.784213066101074, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.35322781652212143, + "loss_total_v6": 0.48651334270834923, + "step": 2057 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47463783783783786, + "grad_norm": 27.56429672241211, + "learning_rate": 5.7941735403212e-06, + "loss": 6.567314147949219, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2792925760149956, + "loss_total_v6": 0.4104571230709553, + "step": 2058 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4748684684684685, + "grad_norm": 26.33144187927246, + "learning_rate": 5.790423661283112e-06, + "loss": 6.757554531097412, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.29566869139671326, + "loss_total_v6": 0.42234715074300766, + "step": 2059 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4750990990990991, + "grad_norm": 28.676891326904297, + "learning_rate": 5.786673326209584e-06, + "loss": 7.421187877655029, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.32922646403312683, + "loss_total_v6": 0.46382423862814903, + "step": 2060 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47532972972972976, + "grad_norm": 24.770666122436523, + "learning_rate": 5.78292253726437e-06, + "loss": 7.034258842468262, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.30998723953962326, + "loss_total_v6": 0.43964117392897606, + "step": 2061 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47556036036036037, + "grad_norm": 23.68360710144043, + "learning_rate": 5.779171296611493e-06, + "loss": 6.955306053161621, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3033437170088291, + "loss_total_v6": 0.4347066320478916, + "step": 2062 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.475790990990991, + "grad_norm": 24.141477584838867, + "learning_rate": 5.775419606415233e-06, + "loss": 7.15117073059082, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.31566155329346657, + "loss_total_v6": 0.44694817811250687, + "step": 2063 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47602162162162165, + "grad_norm": 25.57117462158203, + "learning_rate": 5.771667468840128e-06, + "loss": 7.556817054748535, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.33864932134747505, + "loss_total_v6": 0.47230106592178345, + "step": 2064 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47625225225225226, + "grad_norm": 25.15115737915039, + "learning_rate": 5.767914886050981e-06, + "loss": 7.130810260772705, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.314755255356431, + "loss_total_v6": 0.44567565247416496, + "step": 2065 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4764828828828829, + "grad_norm": 25.590394973754883, + "learning_rate": 5.7641618602128415e-06, + "loss": 6.784588813781738, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.29235345497727394, + "loss_total_v6": 0.42403680086135864, + "step": 2066 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4767135135135135, + "grad_norm": 26.32581901550293, + "learning_rate": 5.7604083934910235e-06, + "loss": 6.735317230224609, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.29078463464975357, + "loss_total_v6": 0.4209573492407799, + "step": 2067 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47694414414414416, + "grad_norm": 24.27962303161621, + "learning_rate": 5.756654488051091e-06, + "loss": 6.984072685241699, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.3045770339667797, + "loss_total_v6": 0.4365045353770256, + "step": 2068 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47717477477477477, + "grad_norm": 23.518512725830078, + "learning_rate": 5.752900146058862e-06, + "loss": 6.886679172515869, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.2973760776221752, + "loss_total_v6": 0.4304174557328224, + "step": 2069 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4774054054054054, + "grad_norm": 24.679553985595703, + "learning_rate": 5.7491453696804075e-06, + "loss": 7.083357334136963, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.3112400956451893, + "loss_total_v6": 0.442709818482399, + "step": 2070 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47763603603603605, + "grad_norm": 29.220300674438477, + "learning_rate": 5.74539016108205e-06, + "loss": 6.686758995056152, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.2848047502338886, + "loss_total_v6": 0.4179224260151386, + "step": 2071 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47786666666666666, + "grad_norm": 25.119855880737305, + "learning_rate": 5.741634522430353e-06, + "loss": 6.686368942260742, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.28630624152719975, + "loss_total_v6": 0.4178980365395546, + "step": 2072 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4780972972972973, + "grad_norm": 22.32408332824707, + "learning_rate": 5.737878455892141e-06, + "loss": 7.324420928955078, + "loss_alignment": 0.271728515625, + "loss_alignment_w": 0.1358642578125, + "loss_sft": 0.3218967877328396, + "loss_total_v6": 0.4577763006091118, + "step": 2073 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47832792792792794, + "grad_norm": 25.914403915405273, + "learning_rate": 5.734121963634477e-06, + "loss": 6.4476213455200195, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.27335292659699917, + "loss_total_v6": 0.4029763303697109, + "step": 2074 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47855855855855856, + "grad_norm": 24.822660446166992, + "learning_rate": 5.730365047824671e-06, + "loss": 6.588688850402832, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.2805369831621647, + "loss_total_v6": 0.4117930643260479, + "step": 2075 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47878918918918917, + "grad_norm": 23.749032974243164, + "learning_rate": 5.726607710630279e-06, + "loss": 7.370358467102051, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.32739240676164627, + "loss_total_v6": 0.46064741164445877, + "step": 2076 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47901981981981984, + "grad_norm": 26.42859649658203, + "learning_rate": 5.722849954219097e-06, + "loss": 7.059717655181885, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.30925907753407955, + "loss_total_v6": 0.4412323571741581, + "step": 2077 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47925045045045045, + "grad_norm": 24.556673049926758, + "learning_rate": 5.719091780759169e-06, + "loss": 6.776420593261719, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.294299628585577, + "loss_total_v6": 0.4235262982547283, + "step": 2078 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47948108108108106, + "grad_norm": 24.376256942749023, + "learning_rate": 5.715333192418772e-06, + "loss": 6.840518951416016, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.29241587966680527, + "loss_total_v6": 0.42753245308995247, + "step": 2079 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47971171171171173, + "grad_norm": 23.347307205200195, + "learning_rate": 5.711574191366427e-06, + "loss": 7.009496212005615, + "loss_alignment": 0.2667236328125, + "loss_alignment_w": 0.13336181640625, + "loss_sft": 0.30480798333883286, + "loss_total_v6": 0.43809350952506065, + "step": 2080 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.47994234234234234, + "grad_norm": 21.832075119018555, + "learning_rate": 5.707814779770892e-06, + "loss": 7.079219818115234, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.30896735563874245, + "loss_total_v6": 0.44245124980807304, + "step": 2081 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48017297297297296, + "grad_norm": 23.783750534057617, + "learning_rate": 5.704054959801162e-06, + "loss": 7.024265289306641, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.30779100209474564, + "loss_total_v6": 0.43901658803224564, + "step": 2082 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4804036036036036, + "grad_norm": 25.11775779724121, + "learning_rate": 5.700294733626468e-06, + "loss": 7.159761428833008, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.314275860786438, + "loss_total_v6": 0.4474850855767727, + "step": 2083 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48063423423423424, + "grad_norm": 24.159269332885742, + "learning_rate": 5.696534103416276e-06, + "loss": 7.353379249572754, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3248358480632305, + "loss_total_v6": 0.459586214274168, + "step": 2084 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48086486486486485, + "grad_norm": 26.003686904907227, + "learning_rate": 5.6927730713402804e-06, + "loss": 7.246707916259766, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3211138192564249, + "loss_total_v6": 0.45291922613978386, + "step": 2085 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4810954954954955, + "grad_norm": 22.343767166137695, + "learning_rate": 5.6890116395684155e-06, + "loss": 7.29726505279541, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.32451777160167694, + "loss_total_v6": 0.45607905462384224, + "step": 2086 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48132612612612613, + "grad_norm": 24.31031036376953, + "learning_rate": 5.68524981027084e-06, + "loss": 6.597237586975098, + "loss_alignment": 0.2652587890625, + "loss_alignment_w": 0.13262939453125, + "loss_sft": 0.279621671885252, + "loss_total_v6": 0.4123273678123951, + "step": 2087 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48155675675675674, + "grad_norm": 23.423978805541992, + "learning_rate": 5.681487585617942e-06, + "loss": 6.733513355255127, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.28679612278938293, + "loss_total_v6": 0.42084458842873573, + "step": 2088 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4817873873873874, + "grad_norm": 22.44976234436035, + "learning_rate": 5.677724967780341e-06, + "loss": 6.759444236755371, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.28806583397090435, + "loss_total_v6": 0.4224652610719204, + "step": 2089 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.482018018018018, + "grad_norm": 24.94840431213379, + "learning_rate": 5.673961958928882e-06, + "loss": 6.944077491760254, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.30270295962691307, + "loss_total_v6": 0.43400484696030617, + "step": 2090 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48224864864864864, + "grad_norm": 24.332073211669922, + "learning_rate": 5.670198561234635e-06, + "loss": 6.843672752380371, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.295023862272501, + "loss_total_v6": 0.4277295544743538, + "step": 2091 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4824792792792793, + "grad_norm": 28.14728355407715, + "learning_rate": 5.666434776868895e-06, + "loss": 7.138772964477539, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.31358965672552586, + "loss_total_v6": 0.4461732804775238, + "step": 2092 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4827099099099099, + "grad_norm": 25.7601375579834, + "learning_rate": 5.66267060800318e-06, + "loss": 6.70909309387207, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.2847815789282322, + "loss_total_v6": 0.4193183183670044, + "step": 2093 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48294054054054053, + "grad_norm": 25.94438934326172, + "learning_rate": 5.658906056809228e-06, + "loss": 7.409269332885742, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.3323267512023449, + "loss_total_v6": 0.4630793072283268, + "step": 2094 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4831711711711712, + "grad_norm": 26.38056755065918, + "learning_rate": 5.655141125459001e-06, + "loss": 6.875740051269531, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.2956089787185192, + "loss_total_v6": 0.4297337494790554, + "step": 2095 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4834018018018018, + "grad_norm": 24.499452590942383, + "learning_rate": 5.651375816124679e-06, + "loss": 6.948616981506348, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.30013331212103367, + "loss_total_v6": 0.4342885762453079, + "step": 2096 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4836324324324324, + "grad_norm": 22.862247467041016, + "learning_rate": 5.647610130978658e-06, + "loss": 6.6542863845825195, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.28259211406111717, + "loss_total_v6": 0.4158928915858269, + "step": 2097 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48386306306306304, + "grad_norm": 25.05283546447754, + "learning_rate": 5.643844072193554e-06, + "loss": 6.803551197052002, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.295766381546855, + "loss_total_v6": 0.4252219460904598, + "step": 2098 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4840936936936937, + "grad_norm": 24.734777450561523, + "learning_rate": 5.6400776419421955e-06, + "loss": 7.0532331466674805, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3083502762019634, + "loss_total_v6": 0.4408270791172981, + "step": 2099 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4843243243243243, + "grad_norm": 28.045589447021484, + "learning_rate": 5.6363108423976285e-06, + "loss": 6.973271369934082, + "loss_alignment": 0.2657470703125, + "loss_alignment_w": 0.13287353515625, + "loss_sft": 0.3031085133552551, + "loss_total_v6": 0.43582944199442863, + "step": 2100 + }, + { + "epoch": 0.4843243243243243, + "eval_loss": 0.5879513621330261, + "eval_loss_alignment": 0.23827902040525115, + "eval_loss_alignment_w": 0.11913951020262557, + "eval_loss_sft": 0.4688118762468502, + "eval_loss_total_v6": 0.5879513867641692, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.5964, + "eval_samples_per_second": 19.261, + "eval_steps_per_second": 2.417, + "step": 2100 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48455495495495493, + "grad_norm": 26.58577537536621, + "learning_rate": 5.6325436757331096e-06, + "loss": 6.9326581954956055, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.2975641991943121, + "loss_total_v6": 0.43329113721847534, + "step": 2101 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4847855855855856, + "grad_norm": 28.192480087280273, + "learning_rate": 5.628776144122107e-06, + "loss": 6.761745929718018, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.28866747207939625, + "loss_total_v6": 0.4226091094315052, + "step": 2102 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4850162162162162, + "grad_norm": 25.342891693115234, + "learning_rate": 5.625008249738301e-06, + "loss": 6.640827178955078, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.28690839372575283, + "loss_total_v6": 0.4150517024099827, + "step": 2103 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4852468468468468, + "grad_norm": 29.72255516052246, + "learning_rate": 5.621239994755583e-06, + "loss": 6.906744003295898, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2995151150971651, + "loss_total_v6": 0.43167148903012276, + "step": 2104 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4854774774774775, + "grad_norm": 26.134510040283203, + "learning_rate": 5.617471381348048e-06, + "loss": 7.096044540405273, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3105071820318699, + "loss_total_v6": 0.4435027725994587, + "step": 2105 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4857081081081081, + "grad_norm": 21.7513427734375, + "learning_rate": 5.61370241169e-06, + "loss": 6.615654945373535, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.28231387957930565, + "loss_total_v6": 0.41347842663526535, + "step": 2106 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4859387387387387, + "grad_norm": 22.28307342529297, + "learning_rate": 5.609933087955949e-06, + "loss": 6.792762756347656, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.2916589044034481, + "loss_total_v6": 0.42454769089818, + "step": 2107 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4861693693693694, + "grad_norm": 24.702333450317383, + "learning_rate": 5.6061634123206075e-06, + "loss": 6.727261543273926, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.28790075704455376, + "loss_total_v6": 0.42045386508107185, + "step": 2108 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4864, + "grad_norm": 23.07085418701172, + "learning_rate": 5.602393386958894e-06, + "loss": 6.859123229980469, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.300246749073267, + "loss_total_v6": 0.4286952205002308, + "step": 2109 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4866306306306306, + "grad_norm": 22.341047286987305, + "learning_rate": 5.598623014045928e-06, + "loss": 6.749984264373779, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.2903432436287403, + "loss_total_v6": 0.4218740016222, + "step": 2110 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4868612612612613, + "grad_norm": 24.69668197631836, + "learning_rate": 5.594852295757025e-06, + "loss": 6.923737525939941, + "loss_alignment": 0.2689208984375, + "loss_alignment_w": 0.13446044921875, + "loss_sft": 0.29868514090776443, + "loss_total_v6": 0.43273360282182693, + "step": 2111 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4870918918918919, + "grad_norm": 23.888978958129883, + "learning_rate": 5.5910812342677065e-06, + "loss": 7.106266021728516, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3118632026016712, + "loss_total_v6": 0.4441416412591934, + "step": 2112 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4873225225225225, + "grad_norm": 24.719308853149414, + "learning_rate": 5.587309831753688e-06, + "loss": 6.7427754402160645, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2892670836299658, + "loss_total_v6": 0.42142345011234283, + "step": 2113 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4875531531531532, + "grad_norm": 26.139436721801758, + "learning_rate": 5.583538090390882e-06, + "loss": 7.017768383026123, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3075375221669674, + "loss_total_v6": 0.438610527664423, + "step": 2114 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4877837837837838, + "grad_norm": 26.34595489501953, + "learning_rate": 5.579766012355397e-06, + "loss": 7.224305629730225, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.32053766399621964, + "loss_total_v6": 0.45151910185813904, + "step": 2115 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4880144144144144, + "grad_norm": 24.0959529876709, + "learning_rate": 5.575993599823536e-06, + "loss": 6.376646041870117, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2674826253205538, + "loss_total_v6": 0.39854035899043083, + "step": 2116 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48824504504504507, + "grad_norm": 25.089231491088867, + "learning_rate": 5.572220854971796e-06, + "loss": 6.353394031524658, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.26621247455477715, + "loss_total_v6": 0.39708712324500084, + "step": 2117 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4884756756756757, + "grad_norm": 24.637916564941406, + "learning_rate": 5.568447779976864e-06, + "loss": 7.3961639404296875, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.33108042925596237, + "loss_total_v6": 0.46226025372743607, + "step": 2118 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4887063063063063, + "grad_norm": 24.622846603393555, + "learning_rate": 5.5646743770156176e-06, + "loss": 6.94243860244751, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.30549968779087067, + "loss_total_v6": 0.43390240147709846, + "step": 2119 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48893693693693696, + "grad_norm": 30.620010375976562, + "learning_rate": 5.560900648265124e-06, + "loss": 6.660569190979004, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.28477003797888756, + "loss_total_v6": 0.41628555953502655, + "step": 2120 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4891675675675676, + "grad_norm": 26.328933715820312, + "learning_rate": 5.557126595902638e-06, + "loss": 7.2042083740234375, + "loss_alignment": 0.2696533203125, + "loss_alignment_w": 0.13482666015625, + "loss_sft": 0.31575678288936615, + "loss_total_v6": 0.45026300475001335, + "step": 2121 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4893981981981982, + "grad_norm": 27.1560001373291, + "learning_rate": 5.553352222105604e-06, + "loss": 6.6786274909973145, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.28617337718605995, + "loss_total_v6": 0.41741422191262245, + "step": 2122 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48962882882882885, + "grad_norm": 24.403953552246094, + "learning_rate": 5.549577529051647e-06, + "loss": 7.13774299621582, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3128233850002289, + "loss_total_v6": 0.4461089298129082, + "step": 2123 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.48985945945945947, + "grad_norm": 28.422225952148438, + "learning_rate": 5.54580251891858e-06, + "loss": 7.0877814292907715, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.30976186878979206, + "loss_total_v6": 0.4429863356053829, + "step": 2124 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4900900900900901, + "grad_norm": 24.10542869567871, + "learning_rate": 5.542027193884395e-06, + "loss": 6.972597122192383, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3052941281348467, + "loss_total_v6": 0.43578729778528214, + "step": 2125 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4903207207207207, + "grad_norm": 25.192424774169922, + "learning_rate": 5.538251556127272e-06, + "loss": 7.519453048706055, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3391216956079006, + "loss_total_v6": 0.4699658267199993, + "step": 2126 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49055135135135136, + "grad_norm": 25.689212799072266, + "learning_rate": 5.534475607825566e-06, + "loss": 6.739293098449707, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2886069491505623, + "loss_total_v6": 0.4212058112025261, + "step": 2127 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.490781981981982, + "grad_norm": 25.97075080871582, + "learning_rate": 5.53069935115781e-06, + "loss": 6.5703277587890625, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.2779398038983345, + "loss_total_v6": 0.4106454849243164, + "step": 2128 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4910126126126126, + "grad_norm": 25.52933692932129, + "learning_rate": 5.526922788302721e-06, + "loss": 7.610170364379883, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3387490510940552, + "loss_total_v6": 0.475635651499033, + "step": 2129 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49124324324324325, + "grad_norm": 25.06576919555664, + "learning_rate": 5.5231459214391905e-06, + "loss": 6.792794227600098, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.29558234103024006, + "loss_total_v6": 0.4245496392250061, + "step": 2130 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49147387387387387, + "grad_norm": 29.231721878051758, + "learning_rate": 5.519368752746281e-06, + "loss": 6.462194442749023, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.27156293019652367, + "loss_total_v6": 0.40388716012239456, + "step": 2131 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4917045045045045, + "grad_norm": 25.582164764404297, + "learning_rate": 5.515591284403235e-06, + "loss": 7.189082145690918, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.31568116694688797, + "loss_total_v6": 0.4493176229298115, + "step": 2132 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49193513513513515, + "grad_norm": 25.4105167388916, + "learning_rate": 5.511813518589462e-06, + "loss": 7.404152870178223, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3308168426156044, + "loss_total_v6": 0.462759580463171, + "step": 2133 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49216576576576576, + "grad_norm": 25.92514991760254, + "learning_rate": 5.508035457484549e-06, + "loss": 7.512669086456299, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.3379652835428715, + "loss_total_v6": 0.4695418179035187, + "step": 2134 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4923963963963964, + "grad_norm": 29.766796112060547, + "learning_rate": 5.504257103268253e-06, + "loss": 7.116686820983887, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3121330067515373, + "loss_total_v6": 0.444792915135622, + "step": 2135 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49262702702702704, + "grad_norm": 28.180795669555664, + "learning_rate": 5.500478458120493e-06, + "loss": 6.894009590148926, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.2994821388274431, + "loss_total_v6": 0.43087557330727577, + "step": 2136 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49285765765765766, + "grad_norm": 23.82647132873535, + "learning_rate": 5.4966995242213625e-06, + "loss": 7.1812286376953125, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.31381701678037643, + "loss_total_v6": 0.44882678985595703, + "step": 2137 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49308828828828827, + "grad_norm": 25.284799575805664, + "learning_rate": 5.492920303751122e-06, + "loss": 6.637020111083984, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.28184868209064007, + "loss_total_v6": 0.4148137830197811, + "step": 2138 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49331891891891894, + "grad_norm": 24.244121551513672, + "learning_rate": 5.489140798890193e-06, + "loss": 6.92563533782959, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.30200809240341187, + "loss_total_v6": 0.43285221233963966, + "step": 2139 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49354954954954955, + "grad_norm": 23.94703483581543, + "learning_rate": 5.485361011819163e-06, + "loss": 6.699306488037109, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.28804565221071243, + "loss_total_v6": 0.41870665550231934, + "step": 2140 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49378018018018016, + "grad_norm": 25.767047882080078, + "learning_rate": 5.481580944718785e-06, + "loss": 7.137546539306641, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3136961758136749, + "loss_total_v6": 0.44609667733311653, + "step": 2141 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49401081081081083, + "grad_norm": 26.437665939331055, + "learning_rate": 5.477800599769968e-06, + "loss": 7.3391571044921875, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3300199415534735, + "loss_total_v6": 0.4586973078548908, + "step": 2142 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49424144144144144, + "grad_norm": 26.7730655670166, + "learning_rate": 5.4740199791537875e-06, + "loss": 6.912052154541016, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.29996894858777523, + "loss_total_v6": 0.4320032633841038, + "step": 2143 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49447207207207206, + "grad_norm": 24.67458152770996, + "learning_rate": 5.470239085051473e-06, + "loss": 6.576375961303711, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.2805303577333689, + "loss_total_v6": 0.41102350875735283, + "step": 2144 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4947027027027027, + "grad_norm": 22.954551696777344, + "learning_rate": 5.466457919644415e-06, + "loss": 7.238219261169434, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.3212089166045189, + "loss_total_v6": 0.4523887038230896, + "step": 2145 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49493333333333334, + "grad_norm": 30.831825256347656, + "learning_rate": 5.462676485114157e-06, + "loss": 6.698979377746582, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.2878573704510927, + "loss_total_v6": 0.41868622973561287, + "step": 2146 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49516396396396395, + "grad_norm": 24.31522560119629, + "learning_rate": 5.458894783642402e-06, + "loss": 6.901009559631348, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.2983480505645275, + "loss_total_v6": 0.431313119828701, + "step": 2147 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4953945945945946, + "grad_norm": 26.559831619262695, + "learning_rate": 5.455112817411006e-06, + "loss": 6.783405303955078, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.292248971760273, + "loss_total_v6": 0.4239628426730633, + "step": 2148 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49562522522522523, + "grad_norm": 26.522140502929688, + "learning_rate": 5.451330588601976e-06, + "loss": 7.084873676300049, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.3075201604515314, + "loss_total_v6": 0.44280458986759186, + "step": 2149 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49585585585585584, + "grad_norm": 23.206501007080078, + "learning_rate": 5.447548099397469e-06, + "loss": 6.867471218109131, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2973810061812401, + "loss_total_v6": 0.4292169399559498, + "step": 2150 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4960864864864865, + "grad_norm": 23.106592178344727, + "learning_rate": 5.443765351979799e-06, + "loss": 6.936155319213867, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3030623197555542, + "loss_total_v6": 0.4335096962749958, + "step": 2151 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4963171171171171, + "grad_norm": 25.992782592773438, + "learning_rate": 5.439982348531423e-06, + "loss": 7.4923858642578125, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.333554282784462, + "loss_total_v6": 0.4682741351425648, + "step": 2152 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49654774774774774, + "grad_norm": 24.57164192199707, + "learning_rate": 5.436199091234945e-06, + "loss": 7.196743011474609, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3180063031613827, + "loss_total_v6": 0.4497964531183243, + "step": 2153 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4967783783783784, + "grad_norm": 27.506383895874023, + "learning_rate": 5.432415582273118e-06, + "loss": 7.046534538269043, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3100373148918152, + "loss_total_v6": 0.4404084049165249, + "step": 2154 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.497009009009009, + "grad_norm": 22.298633575439453, + "learning_rate": 5.428631823828843e-06, + "loss": 6.935231685638428, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.30137188360095024, + "loss_total_v6": 0.43345196917653084, + "step": 2155 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49723963963963963, + "grad_norm": 29.5520076751709, + "learning_rate": 5.4248478180851604e-06, + "loss": 6.927412033081055, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3025311045348644, + "loss_total_v6": 0.4329632557928562, + "step": 2156 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49747027027027024, + "grad_norm": 21.251222610473633, + "learning_rate": 5.421063567225254e-06, + "loss": 7.442031383514404, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3316277973353863, + "loss_total_v6": 0.46512696892023087, + "step": 2157 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4977009009009009, + "grad_norm": 27.63862419128418, + "learning_rate": 5.41727907343245e-06, + "loss": 7.180532932281494, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.3169168531894684, + "loss_total_v6": 0.4487833119928837, + "step": 2158 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4979315315315315, + "grad_norm": 21.187450408935547, + "learning_rate": 5.413494338890215e-06, + "loss": 6.730392932891846, + "loss_alignment": 0.2642822265625, + "loss_alignment_w": 0.13214111328125, + "loss_sft": 0.2879591342061758, + "loss_total_v6": 0.42064956575632095, + "step": 2159 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49816216216216214, + "grad_norm": 23.333969116210938, + "learning_rate": 5.4097093657821545e-06, + "loss": 7.410898685455322, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.33242858946323395, + "loss_total_v6": 0.46318116039037704, + "step": 2160 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4983927927927928, + "grad_norm": 25.545419692993164, + "learning_rate": 5.405924156292008e-06, + "loss": 7.351380348205566, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3262672871351242, + "loss_total_v6": 0.4594612680375576, + "step": 2161 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4986234234234234, + "grad_norm": 23.726030349731445, + "learning_rate": 5.40213871260366e-06, + "loss": 6.717856407165527, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2878775168210268, + "loss_total_v6": 0.41986605525016785, + "step": 2162 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.49885405405405403, + "grad_norm": 26.77463722229004, + "learning_rate": 5.39835303690112e-06, + "loss": 6.897895812988281, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.2997403349727392, + "loss_total_v6": 0.43111850321292877, + "step": 2163 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4990846846846847, + "grad_norm": 25.392520904541016, + "learning_rate": 5.394567131368539e-06, + "loss": 7.060423374176025, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3088912181556225, + "loss_total_v6": 0.4412764683365822, + "step": 2164 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4993153153153153, + "grad_norm": 24.479307174682617, + "learning_rate": 5.390780998190197e-06, + "loss": 7.197700023651123, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.3162960894405842, + "loss_total_v6": 0.4498562589287758, + "step": 2165 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4995459459459459, + "grad_norm": 20.35626792907715, + "learning_rate": 5.3869946395505045e-06, + "loss": 6.992044448852539, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.30811179243028164, + "loss_total_v6": 0.4370027743279934, + "step": 2166 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.4997765765765766, + "grad_norm": 25.088531494140625, + "learning_rate": 5.383208057634004e-06, + "loss": 7.055257797241211, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3092550113797188, + "loss_total_v6": 0.4409536346793175, + "step": 2167 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5000072072072073, + "grad_norm": 25.688804626464844, + "learning_rate": 5.379421254625367e-06, + "loss": 7.128907680511475, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3086395896971226, + "loss_total_v6": 0.44555671885609627, + "step": 2168 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5002378378378378, + "grad_norm": 26.924345016479492, + "learning_rate": 5.375634232709392e-06, + "loss": 6.8100504875183105, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.2929682545363903, + "loss_total_v6": 0.4256281591951847, + "step": 2169 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5004684684684685, + "grad_norm": 24.600690841674805, + "learning_rate": 5.371846994071005e-06, + "loss": 6.674284934997559, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.2836589030921459, + "loss_total_v6": 0.4171427898108959, + "step": 2170 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5006990990990992, + "grad_norm": 26.994783401489258, + "learning_rate": 5.368059540895252e-06, + "loss": 7.130857467651367, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.31251514330506325, + "loss_total_v6": 0.44567859917879105, + "step": 2171 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5009297297297297, + "grad_norm": 25.614580154418945, + "learning_rate": 5.364271875367311e-06, + "loss": 7.410064697265625, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.33144568651914597, + "loss_total_v6": 0.46312902867794037, + "step": 2172 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5011603603603604, + "grad_norm": 26.233938217163086, + "learning_rate": 5.360483999672475e-06, + "loss": 6.776415824890137, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.29075922444462776, + "loss_total_v6": 0.42352597042918205, + "step": 2173 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5013909909909909, + "grad_norm": 27.638626098632812, + "learning_rate": 5.356695915996162e-06, + "loss": 6.848467826843262, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.29750555381178856, + "loss_total_v6": 0.42802922427654266, + "step": 2174 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5016216216216216, + "grad_norm": 23.798484802246094, + "learning_rate": 5.3529076265239114e-06, + "loss": 7.030223846435547, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.3084685765206814, + "loss_total_v6": 0.43938900902867317, + "step": 2175 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5018522522522523, + "grad_norm": 25.75377655029297, + "learning_rate": 5.349119133441375e-06, + "loss": 7.222832202911377, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3186602760106325, + "loss_total_v6": 0.45142700523138046, + "step": 2176 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5020828828828828, + "grad_norm": 26.717336654663086, + "learning_rate": 5.345330438934331e-06, + "loss": 6.704257965087891, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.28881289437413216, + "loss_total_v6": 0.41901613771915436, + "step": 2177 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5023135135135135, + "grad_norm": 25.84540367126465, + "learning_rate": 5.341541545188666e-06, + "loss": 6.885745048522949, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.29801958054304123, + "loss_total_v6": 0.4303590655326843, + "step": 2178 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5025441441441442, + "grad_norm": 27.003673553466797, + "learning_rate": 5.337752454390387e-06, + "loss": 6.608030319213867, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.282096765935421, + "loss_total_v6": 0.4130019135773182, + "step": 2179 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5027747747747747, + "grad_norm": 23.915416717529297, + "learning_rate": 5.3339631687256085e-06, + "loss": 7.2671966552734375, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.32021236419677734, + "loss_total_v6": 0.45419978722929955, + "step": 2180 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5030054054054054, + "grad_norm": 25.609214782714844, + "learning_rate": 5.3301736903805646e-06, + "loss": 6.854475021362305, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.2967671174556017, + "loss_total_v6": 0.42840469628572464, + "step": 2181 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5032360360360361, + "grad_norm": 25.353384017944336, + "learning_rate": 5.326384021541597e-06, + "loss": 7.394762992858887, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3282768465578556, + "loss_total_v6": 0.4621727056801319, + "step": 2182 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5034666666666666, + "grad_norm": 25.122888565063477, + "learning_rate": 5.322594164395159e-06, + "loss": 6.698002815246582, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2860262766480446, + "loss_total_v6": 0.4186251610517502, + "step": 2183 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5036972972972973, + "grad_norm": 22.884658813476562, + "learning_rate": 5.318804121127808e-06, + "loss": 7.079972743988037, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3109522759914398, + "loss_total_v6": 0.4424982964992523, + "step": 2184 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.503927927927928, + "grad_norm": 26.750165939331055, + "learning_rate": 5.315013893926212e-06, + "loss": 7.000269412994385, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3023544866591692, + "loss_total_v6": 0.43751683831214905, + "step": 2185 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5041585585585585, + "grad_norm": 23.830001831054688, + "learning_rate": 5.311223484977148e-06, + "loss": 7.025556564331055, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3086193762719631, + "loss_total_v6": 0.4390972666442394, + "step": 2186 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5043891891891892, + "grad_norm": 23.07016944885254, + "learning_rate": 5.307432896467491e-06, + "loss": 7.042654991149902, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.3114885576069355, + "loss_total_v6": 0.4401659220457077, + "step": 2187 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5046198198198198, + "grad_norm": 24.873258590698242, + "learning_rate": 5.303642130584228e-06, + "loss": 6.835920333862305, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.29531752690672874, + "loss_total_v6": 0.42724500596523285, + "step": 2188 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5048504504504504, + "grad_norm": 26.358613967895508, + "learning_rate": 5.29985118951444e-06, + "loss": 6.9571533203125, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3057479988783598, + "loss_total_v6": 0.43482208997011185, + "step": 2189 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5050810810810811, + "grad_norm": 26.13385581970215, + "learning_rate": 5.296060075445315e-06, + "loss": 6.964229583740234, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.30324528738856316, + "loss_total_v6": 0.43526434153318405, + "step": 2190 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5053117117117117, + "grad_norm": 28.58495330810547, + "learning_rate": 5.292268790564138e-06, + "loss": 6.888510704040527, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.3020986709743738, + "loss_total_v6": 0.43053190410137177, + "step": 2191 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5055423423423423, + "grad_norm": 23.543230056762695, + "learning_rate": 5.288477337058293e-06, + "loss": 6.864031791687012, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.2958995755761862, + "loss_total_v6": 0.42900198325514793, + "step": 2192 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.505772972972973, + "grad_norm": 24.839643478393555, + "learning_rate": 5.284685717115262e-06, + "loss": 7.568373680114746, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.33696070313453674, + "loss_total_v6": 0.47302333638072014, + "step": 2193 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5060036036036036, + "grad_norm": 27.364110946655273, + "learning_rate": 5.2808939329226215e-06, + "loss": 7.167392253875732, + "loss_alignment": 0.2701416015625, + "loss_alignment_w": 0.13507080078125, + "loss_sft": 0.31269286200404167, + "loss_total_v6": 0.44796203076839447, + "step": 2194 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5062342342342342, + "grad_norm": 26.481569290161133, + "learning_rate": 5.2771019866680465e-06, + "loss": 7.214099884033203, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3181908130645752, + "loss_total_v6": 0.4508812576532364, + "step": 2195 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5064648648648649, + "grad_norm": 26.308107376098633, + "learning_rate": 5.273309880539301e-06, + "loss": 7.137360572814941, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.31385236978530884, + "loss_total_v6": 0.44608503952622414, + "step": 2196 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5066954954954955, + "grad_norm": 28.486753463745117, + "learning_rate": 5.2695176167242425e-06, + "loss": 7.025010585784912, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.3070593625307083, + "loss_total_v6": 0.4390631541609764, + "step": 2197 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5069261261261261, + "grad_norm": 22.72794532775879, + "learning_rate": 5.265725197410821e-06, + "loss": 6.760393142700195, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.28850664757192135, + "loss_total_v6": 0.4225245900452137, + "step": 2198 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5071567567567568, + "grad_norm": 25.07668113708496, + "learning_rate": 5.261932624787075e-06, + "loss": 6.882142543792725, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.30171595700085163, + "loss_total_v6": 0.430133905261755, + "step": 2199 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5073873873873874, + "grad_norm": 29.468212127685547, + "learning_rate": 5.258139901041131e-06, + "loss": 6.730387210845947, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.29032387025654316, + "loss_total_v6": 0.4206491969525814, + "step": 2200 + }, + { + "epoch": 0.5073873873873874, + "eval_loss": 0.5835849642753601, + "eval_loss_alignment": 0.2368097174657534, + "eval_loss_alignment_w": 0.1184048587328767, + "eval_loss_sft": 0.4651800938733092, + "eval_loss_total_v6": 0.583584953167531, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 93.4902, + "eval_samples_per_second": 18.665, + "eval_steps_per_second": 2.342, + "step": 2200 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.507618018018018, + "grad_norm": 23.11248016357422, + "learning_rate": 5.254347028361207e-06, + "loss": 6.74973201751709, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.289854496717453, + "loss_total_v6": 0.4218582697212696, + "step": 2201 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5078486486486486, + "grad_norm": 25.80480194091797, + "learning_rate": 5.250554008935596e-06, + "loss": 6.921952724456787, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.30055722780525684, + "loss_total_v6": 0.4326220527291298, + "step": 2202 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5080792792792793, + "grad_norm": 25.948694229125977, + "learning_rate": 5.246760844952692e-06, + "loss": 7.2270827293396, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.32090955786406994, + "loss_total_v6": 0.4516926556825638, + "step": 2203 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5083099099099099, + "grad_norm": 25.6557674407959, + "learning_rate": 5.242967538600957e-06, + "loss": 6.482044219970703, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2741768453270197, + "loss_total_v6": 0.40512778609991074, + "step": 2204 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5085405405405405, + "grad_norm": 22.98134422302246, + "learning_rate": 5.239174092068944e-06, + "loss": 6.801490783691406, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.29501201398670673, + "loss_total_v6": 0.4250931963324547, + "step": 2205 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5087711711711712, + "grad_norm": 24.174999237060547, + "learning_rate": 5.235380507545282e-06, + "loss": 6.707723617553711, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2894720062613487, + "loss_total_v6": 0.41923272982239723, + "step": 2206 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5090018018018018, + "grad_norm": 23.012924194335938, + "learning_rate": 5.231586787218687e-06, + "loss": 6.546586513519287, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.27990445867180824, + "loss_total_v6": 0.40916165709495544, + "step": 2207 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5092324324324324, + "grad_norm": 26.627422332763672, + "learning_rate": 5.2277929332779435e-06, + "loss": 6.987862586975098, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3035932369530201, + "loss_total_v6": 0.4367414154112339, + "step": 2208 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5094630630630631, + "grad_norm": 24.582962036132812, + "learning_rate": 5.22399894791192e-06, + "loss": 7.026784896850586, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3051256202161312, + "loss_total_v6": 0.4391740746796131, + "step": 2209 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5096936936936937, + "grad_norm": 24.898574829101562, + "learning_rate": 5.220204833309559e-06, + "loss": 6.98137092590332, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.30436241813004017, + "loss_total_v6": 0.4363356791436672, + "step": 2210 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5099243243243243, + "grad_norm": 28.75760269165039, + "learning_rate": 5.216410591659877e-06, + "loss": 6.757046699523926, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.29292090237140656, + "loss_total_v6": 0.42231544107198715, + "step": 2211 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.510154954954955, + "grad_norm": 25.583425521850586, + "learning_rate": 5.212616225151965e-06, + "loss": 6.344176769256592, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2655448578298092, + "loss_total_v6": 0.3965110406279564, + "step": 2212 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5103855855855856, + "grad_norm": 27.235958099365234, + "learning_rate": 5.208821735974984e-06, + "loss": 7.071803569793701, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3083970546722412, + "loss_total_v6": 0.4419877305626869, + "step": 2213 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5106162162162162, + "grad_norm": 24.098543167114258, + "learning_rate": 5.205027126318167e-06, + "loss": 7.2693986892700195, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.321219727396965, + "loss_total_v6": 0.4543374106287956, + "step": 2214 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5108468468468469, + "grad_norm": 27.27153205871582, + "learning_rate": 5.201232398370817e-06, + "loss": 6.8509345054626465, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.2957828678190708, + "loss_total_v6": 0.4281833842396736, + "step": 2215 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5110774774774774, + "grad_norm": 25.443130493164062, + "learning_rate": 5.1974375543223046e-06, + "loss": 6.852130889892578, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.2973377499729395, + "loss_total_v6": 0.42825816571712494, + "step": 2216 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5113081081081081, + "grad_norm": 28.66526985168457, + "learning_rate": 5.193642596362068e-06, + "loss": 7.226317405700684, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.32014460675418377, + "loss_total_v6": 0.4516448527574539, + "step": 2217 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5115387387387388, + "grad_norm": 24.495859146118164, + "learning_rate": 5.1898475266796115e-06, + "loss": 7.155909538269043, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3156983070075512, + "loss_total_v6": 0.447244331240654, + "step": 2218 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5117693693693693, + "grad_norm": 24.4652156829834, + "learning_rate": 5.186052347464498e-06, + "loss": 6.6656084060668945, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.2831624113023281, + "loss_total_v6": 0.4166005104780197, + "step": 2219 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.512, + "grad_norm": 24.14605712890625, + "learning_rate": 5.182257060906365e-06, + "loss": 6.334980487823486, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.2686322070658207, + "loss_total_v6": 0.3959362767636776, + "step": 2220 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5122306306306307, + "grad_norm": 26.58324432373047, + "learning_rate": 5.178461669194903e-06, + "loss": 7.022163391113281, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.30715607292950153, + "loss_total_v6": 0.4388852044939995, + "step": 2221 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5124612612612612, + "grad_norm": 22.439983367919922, + "learning_rate": 5.174666174519866e-06, + "loss": 6.961780548095703, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.29965901374816895, + "loss_total_v6": 0.43511129915714264, + "step": 2222 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5126918918918919, + "grad_norm": 26.56812858581543, + "learning_rate": 5.170870579071064e-06, + "loss": 7.213906288146973, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.3195672556757927, + "loss_total_v6": 0.4508691541850567, + "step": 2223 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5129225225225226, + "grad_norm": 25.582590103149414, + "learning_rate": 5.1670748850383734e-06, + "loss": 7.0482072830200195, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.30782249942421913, + "loss_total_v6": 0.44051293283700943, + "step": 2224 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5131531531531531, + "grad_norm": 26.43106460571289, + "learning_rate": 5.1632790946117195e-06, + "loss": 7.145328521728516, + "loss_alignment": 0.271484375, + "loss_alignment_w": 0.1357421875, + "loss_sft": 0.3106272164732218, + "loss_total_v6": 0.4465830437839031, + "step": 2225 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5133837837837838, + "grad_norm": 22.61912727355957, + "learning_rate": 5.159483209981085e-06, + "loss": 6.463828086853027, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2728552632033825, + "loss_total_v6": 0.4039892740547657, + "step": 2226 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5136144144144145, + "grad_norm": 24.434955596923828, + "learning_rate": 5.155687233336511e-06, + "loss": 7.4165849685668945, + "loss_alignment": 0.2679443359375, + "loss_alignment_w": 0.13397216796875, + "loss_sft": 0.3293965607881546, + "loss_total_v6": 0.4635365754365921, + "step": 2227 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.513845045045045, + "grad_norm": 25.5106258392334, + "learning_rate": 5.151891166868086e-06, + "loss": 6.998516082763672, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.30776859261095524, + "loss_total_v6": 0.4374072551727295, + "step": 2228 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5140756756756757, + "grad_norm": 21.860811233520508, + "learning_rate": 5.148095012765956e-06, + "loss": 7.074428081512451, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.31055994890630245, + "loss_total_v6": 0.4421517476439476, + "step": 2229 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5143063063063064, + "grad_norm": 23.48936653137207, + "learning_rate": 5.144298773220311e-06, + "loss": 7.00007438659668, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.3102921135723591, + "loss_total_v6": 0.4375046454370022, + "step": 2230 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5145369369369369, + "grad_norm": 24.15243911743164, + "learning_rate": 5.140502450421396e-06, + "loss": 6.781569957733154, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.291813800111413, + "loss_total_v6": 0.42384811863303185, + "step": 2231 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5147675675675676, + "grad_norm": 22.314716339111328, + "learning_rate": 5.136706046559501e-06, + "loss": 7.1134033203125, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.31572720780968666, + "loss_total_v6": 0.44458768144249916, + "step": 2232 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5149981981981981, + "grad_norm": 26.733057022094727, + "learning_rate": 5.132909563824964e-06, + "loss": 7.097886085510254, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.30903534963727, + "loss_total_v6": 0.44361788406968117, + "step": 2233 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5152288288288288, + "grad_norm": 22.914052963256836, + "learning_rate": 5.129113004408168e-06, + "loss": 7.091957092285156, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.31160973012447357, + "loss_total_v6": 0.44324731826782227, + "step": 2234 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5154594594594595, + "grad_norm": 25.00433921813965, + "learning_rate": 5.1253163704995425e-06, + "loss": 6.990032196044922, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.30606343038380146, + "loss_total_v6": 0.4368770122528076, + "step": 2235 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.51569009009009, + "grad_norm": 27.99495506286621, + "learning_rate": 5.121519664289553e-06, + "loss": 7.449092864990234, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.33072636649012566, + "loss_total_v6": 0.46556827798485756, + "step": 2236 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5159207207207207, + "grad_norm": 23.182994842529297, + "learning_rate": 5.117722887968717e-06, + "loss": 6.915266036987305, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.2992390487343073, + "loss_total_v6": 0.43220412358641624, + "step": 2237 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5161513513513514, + "grad_norm": 25.525169372558594, + "learning_rate": 5.113926043727584e-06, + "loss": 7.293938636779785, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.32237205281853676, + "loss_total_v6": 0.45587119087576866, + "step": 2238 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5163819819819819, + "grad_norm": 25.07171630859375, + "learning_rate": 5.110129133756746e-06, + "loss": 6.981966972351074, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3045522477477789, + "loss_total_v6": 0.43637293204665184, + "step": 2239 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5166126126126126, + "grad_norm": 26.350685119628906, + "learning_rate": 5.106332160246834e-06, + "loss": 7.261034965515137, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.32490842416882515, + "loss_total_v6": 0.45381467044353485, + "step": 2240 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5168432432432433, + "grad_norm": 25.115819931030273, + "learning_rate": 5.102535125388511e-06, + "loss": 7.222658157348633, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.32080090418457985, + "loss_total_v6": 0.45141611993312836, + "step": 2241 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5170738738738738, + "grad_norm": 27.36624526977539, + "learning_rate": 5.098738031372484e-06, + "loss": 7.236227035522461, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.3187345489859581, + "loss_total_v6": 0.4522642083466053, + "step": 2242 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5173045045045045, + "grad_norm": 21.937519073486328, + "learning_rate": 5.094940880389485e-06, + "loss": 6.744947910308838, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.29021158814430237, + "loss_total_v6": 0.42155923694372177, + "step": 2243 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5175351351351352, + "grad_norm": 27.122852325439453, + "learning_rate": 5.091143674630283e-06, + "loss": 7.144455909729004, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.31676775962114334, + "loss_total_v6": 0.44652850925922394, + "step": 2244 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5177657657657657, + "grad_norm": 23.623096466064453, + "learning_rate": 5.08734641628568e-06, + "loss": 7.077051162719727, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3090759925544262, + "loss_total_v6": 0.4423157200217247, + "step": 2245 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5179963963963964, + "grad_norm": 27.544719696044922, + "learning_rate": 5.083549107546505e-06, + "loss": 7.198967933654785, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31883198767900467, + "loss_total_v6": 0.4499354958534241, + "step": 2246 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.518227027027027, + "grad_norm": 28.732248306274414, + "learning_rate": 5.079751750603618e-06, + "loss": 7.3259429931640625, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3265085443854332, + "loss_total_v6": 0.4578714445233345, + "step": 2247 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5184576576576576, + "grad_norm": 25.113513946533203, + "learning_rate": 5.075954347647909e-06, + "loss": 6.543385028839111, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.2819931823760271, + "loss_total_v6": 0.40896156057715416, + "step": 2248 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5186882882882883, + "grad_norm": 23.12540054321289, + "learning_rate": 5.072156900870288e-06, + "loss": 7.03615140914917, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3067486211657524, + "loss_total_v6": 0.4397594779729843, + "step": 2249 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.518918918918919, + "grad_norm": 26.279727935791016, + "learning_rate": 5.0683594124616975e-06, + "loss": 6.663440227508545, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2837440874427557, + "loss_total_v6": 0.41646501049399376, + "step": 2250 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5191495495495495, + "grad_norm": 24.073522567749023, + "learning_rate": 5.064561884613099e-06, + "loss": 7.001480579376221, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3056192621588707, + "loss_total_v6": 0.4375925362110138, + "step": 2251 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5193801801801802, + "grad_norm": 24.018165588378906, + "learning_rate": 5.06076431951548e-06, + "loss": 7.152122497558594, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3137831948697567, + "loss_total_v6": 0.4470076635479927, + "step": 2252 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5196108108108108, + "grad_norm": 23.206953048706055, + "learning_rate": 5.056966719359847e-06, + "loss": 7.1069464683532715, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.31323324143886566, + "loss_total_v6": 0.44418417289853096, + "step": 2253 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5198414414414414, + "grad_norm": 24.419464111328125, + "learning_rate": 5.0531690863372284e-06, + "loss": 6.918259620666504, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3019896298646927, + "loss_total_v6": 0.4323912411928177, + "step": 2254 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5200720720720721, + "grad_norm": 26.102811813354492, + "learning_rate": 5.0493714226386715e-06, + "loss": 7.225611209869385, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.32170261070132256, + "loss_total_v6": 0.45160068571567535, + "step": 2255 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5203027027027027, + "grad_norm": 27.880142211914062, + "learning_rate": 5.045573730455241e-06, + "loss": 7.323159217834473, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.32360321283340454, + "loss_total_v6": 0.45769745483994484, + "step": 2256 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5205333333333333, + "grad_norm": 23.25762939453125, + "learning_rate": 5.041776011978016e-06, + "loss": 7.319257736206055, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.32375607267022133, + "loss_total_v6": 0.4574535936117172, + "step": 2257 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.520763963963964, + "grad_norm": 27.74178123474121, + "learning_rate": 5.037978269398094e-06, + "loss": 6.957073211669922, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3035762086510658, + "loss_total_v6": 0.4348170608282089, + "step": 2258 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5209945945945946, + "grad_norm": 23.72633934020996, + "learning_rate": 5.034180504906585e-06, + "loss": 6.949423789978027, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3019842505455017, + "loss_total_v6": 0.4343389831483364, + "step": 2259 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5212252252252252, + "grad_norm": 26.485010147094727, + "learning_rate": 5.030382720694612e-06, + "loss": 6.610590934753418, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.2825009264051914, + "loss_total_v6": 0.413161925971508, + "step": 2260 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5214558558558559, + "grad_norm": 25.666967391967773, + "learning_rate": 5.026584918953309e-06, + "loss": 6.610846519470215, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.2831730116158724, + "loss_total_v6": 0.4131779074668884, + "step": 2261 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5216864864864865, + "grad_norm": 24.083484649658203, + "learning_rate": 5.022787101873818e-06, + "loss": 7.400668144226074, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.33215541020035744, + "loss_total_v6": 0.46254176273941994, + "step": 2262 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5219171171171171, + "grad_norm": 25.733705520629883, + "learning_rate": 5.018989271647297e-06, + "loss": 6.833867073059082, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.297508517280221, + "loss_total_v6": 0.42711666226387024, + "step": 2263 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5221477477477477, + "grad_norm": 30.654874801635742, + "learning_rate": 5.015191430464902e-06, + "loss": 7.099961280822754, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.3091955780982971, + "loss_total_v6": 0.4437475800514221, + "step": 2264 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5223783783783784, + "grad_norm": 23.825660705566406, + "learning_rate": 5.0113935805178e-06, + "loss": 6.7607808113098145, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.29301692917943, + "loss_total_v6": 0.4225488007068634, + "step": 2265 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.522609009009009, + "grad_norm": 23.860824584960938, + "learning_rate": 5.007595723997165e-06, + "loss": 7.131518363952637, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.3134719803929329, + "loss_total_v6": 0.4457199014723301, + "step": 2266 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5228396396396396, + "grad_norm": 23.98178482055664, + "learning_rate": 5.003797863094172e-06, + "loss": 7.069332599639893, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3099973425269127, + "loss_total_v6": 0.441833283752203, + "step": 2267 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5230702702702703, + "grad_norm": 26.79676055908203, + "learning_rate": 5e-06, + "loss": 7.717331886291504, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3503294922411442, + "loss_total_v6": 0.4823332726955414, + "step": 2268 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5233009009009009, + "grad_norm": 27.212860107421875, + "learning_rate": 4.996202136905828e-06, + "loss": 6.953927993774414, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3034254349768162, + "loss_total_v6": 0.4346204996109009, + "step": 2269 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5235315315315315, + "grad_norm": 24.08469581604004, + "learning_rate": 4.9924042760028355e-06, + "loss": 7.058925628662109, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.31044558249413967, + "loss_total_v6": 0.44118287041783333, + "step": 2270 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5237621621621622, + "grad_norm": 25.710664749145508, + "learning_rate": 4.9886064194822e-06, + "loss": 7.046947956085205, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3125350810587406, + "loss_total_v6": 0.4404342509806156, + "step": 2271 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5239927927927928, + "grad_norm": 24.69013214111328, + "learning_rate": 4.984808569535101e-06, + "loss": 7.140456676483154, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.31476303189992905, + "loss_total_v6": 0.44627852737903595, + "step": 2272 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5242234234234234, + "grad_norm": 24.880460739135742, + "learning_rate": 4.981010728352705e-06, + "loss": 7.113797187805176, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.31094533763825893, + "loss_total_v6": 0.4446123279631138, + "step": 2273 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5244540540540541, + "grad_norm": 23.30506706237793, + "learning_rate": 4.977212898126183e-06, + "loss": 6.674280166625977, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.28599322587251663, + "loss_total_v6": 0.41714252531528473, + "step": 2274 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5246846846846847, + "grad_norm": 24.735424041748047, + "learning_rate": 4.973415081046692e-06, + "loss": 6.811925888061523, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.2986091375350952, + "loss_total_v6": 0.4257453717291355, + "step": 2275 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5249153153153153, + "grad_norm": 23.724895477294922, + "learning_rate": 4.969617279305388e-06, + "loss": 6.821335792541504, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.2945890985429287, + "loss_total_v6": 0.4263334907591343, + "step": 2276 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.525145945945946, + "grad_norm": 22.52695655822754, + "learning_rate": 4.965819495093417e-06, + "loss": 6.890742301940918, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.29885073006153107, + "loss_total_v6": 0.43067140504717827, + "step": 2277 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5253765765765765, + "grad_norm": 26.42122459411621, + "learning_rate": 4.9620217306019075e-06, + "loss": 6.2637410163879395, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.25868656672537327, + "loss_total_v6": 0.3914838060736656, + "step": 2278 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5256072072072072, + "grad_norm": 26.993619918823242, + "learning_rate": 4.958223988021986e-06, + "loss": 6.6059417724609375, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.28335473127663136, + "loss_total_v6": 0.4128713458776474, + "step": 2279 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5258378378378379, + "grad_norm": 26.267616271972656, + "learning_rate": 4.954426269544761e-06, + "loss": 6.985563278198242, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.30604349449276924, + "loss_total_v6": 0.43659769743680954, + "step": 2280 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5260684684684684, + "grad_norm": 27.01305389404297, + "learning_rate": 4.950628577361331e-06, + "loss": 7.295121192932129, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.32313259318470955, + "loss_total_v6": 0.45594509318470955, + "step": 2281 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5262990990990991, + "grad_norm": 24.0327091217041, + "learning_rate": 4.946830913662773e-06, + "loss": 7.227906703948975, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.31624611094594, + "loss_total_v6": 0.45174415782094, + "step": 2282 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5265297297297298, + "grad_norm": 42.58496856689453, + "learning_rate": 4.943033280640154e-06, + "loss": 6.3953471183776855, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.2695211824029684, + "loss_total_v6": 0.39970918744802475, + "step": 2283 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5267603603603603, + "grad_norm": 24.903396606445312, + "learning_rate": 4.939235680484522e-06, + "loss": 6.738583564758301, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.289783276617527, + "loss_total_v6": 0.421161450445652, + "step": 2284 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.526990990990991, + "grad_norm": 23.849138259887695, + "learning_rate": 4.935438115386902e-06, + "loss": 7.074120998382568, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3099304065108299, + "loss_total_v6": 0.4421325586736202, + "step": 2285 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5272216216216217, + "grad_norm": 24.489089965820312, + "learning_rate": 4.931640587538306e-06, + "loss": 6.42803955078125, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2719764541834593, + "loss_total_v6": 0.40175246074795723, + "step": 2286 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5274522522522522, + "grad_norm": 26.249277114868164, + "learning_rate": 4.927843099129714e-06, + "loss": 7.036360740661621, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3079213686287403, + "loss_total_v6": 0.4397725686430931, + "step": 2287 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5276828828828829, + "grad_norm": 24.738759994506836, + "learning_rate": 4.924045652352093e-06, + "loss": 6.6441755294799805, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.28633944876492023, + "loss_total_v6": 0.4152609780430794, + "step": 2288 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5279135135135136, + "grad_norm": 24.594213485717773, + "learning_rate": 4.9202482493963825e-06, + "loss": 7.554825782775879, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.34003548324108124, + "loss_total_v6": 0.47217659652233124, + "step": 2289 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5281441441441441, + "grad_norm": 27.855741500854492, + "learning_rate": 4.916450892453495e-06, + "loss": 6.586541652679443, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.281074121594429, + "loss_total_v6": 0.4116588458418846, + "step": 2290 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5283747747747748, + "grad_norm": 22.651277542114258, + "learning_rate": 4.912653583714322e-06, + "loss": 7.125626564025879, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.31133369728922844, + "loss_total_v6": 0.44535164535045624, + "step": 2291 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5286054054054055, + "grad_norm": 23.264301300048828, + "learning_rate": 4.9088563253697184e-06, + "loss": 6.806686878204346, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.2938108667731285, + "loss_total_v6": 0.4254179261624813, + "step": 2292 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.528836036036036, + "grad_norm": 26.007020950317383, + "learning_rate": 4.905059119610517e-06, + "loss": 6.756629943847656, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.29031609557569027, + "loss_total_v6": 0.4222893640398979, + "step": 2293 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5290666666666667, + "grad_norm": 24.806095123291016, + "learning_rate": 4.901261968627517e-06, + "loss": 6.567972660064697, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.2812411040067673, + "loss_total_v6": 0.4104983024299145, + "step": 2294 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5292972972972972, + "grad_norm": 26.213443756103516, + "learning_rate": 4.897464874611489e-06, + "loss": 6.828550815582275, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.29415501840412617, + "loss_total_v6": 0.4267844259738922, + "step": 2295 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5295279279279279, + "grad_norm": 27.283369064331055, + "learning_rate": 4.8936678397531675e-06, + "loss": 7.005843162536621, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3046712316572666, + "loss_total_v6": 0.4378652162849903, + "step": 2296 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5297585585585586, + "grad_norm": 24.73346519470215, + "learning_rate": 4.889870866243254e-06, + "loss": 7.038109302520752, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.30653525702655315, + "loss_total_v6": 0.4398818239569664, + "step": 2297 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5299891891891891, + "grad_norm": 24.245573043823242, + "learning_rate": 4.886073956272418e-06, + "loss": 7.134527206420898, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.31744420528411865, + "loss_total_v6": 0.44590796157717705, + "step": 2298 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5302198198198198, + "grad_norm": 23.50308609008789, + "learning_rate": 4.882277112031284e-06, + "loss": 7.49825382232666, + "loss_alignment": 0.2705078125, + "loss_alignment_w": 0.13525390625, + "loss_sft": 0.33329543098807335, + "loss_total_v6": 0.46864088252186775, + "step": 2299 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5304504504504505, + "grad_norm": 25.916873931884766, + "learning_rate": 4.878480335710448e-06, + "loss": 6.71879768371582, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.29001152515411377, + "loss_total_v6": 0.41992485895752907, + "step": 2300 + }, + { + "epoch": 0.5304504504504505, + "eval_loss": 0.5827249884605408, + "eval_loss_alignment": 0.23656669163812785, + "eval_loss_alignment_w": 0.11828334581906393, + "eval_loss_sft": 0.4644416857544826, + "eval_loss_total_v6": 0.5827250318372086, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.9292, + "eval_samples_per_second": 19.191, + "eval_steps_per_second": 2.408, + "step": 2300 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.530681081081081, + "grad_norm": 25.796979904174805, + "learning_rate": 4.87468362950046e-06, + "loss": 7.098245620727539, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.3104616403579712, + "loss_total_v6": 0.4436403512954712, + "step": 2301 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5309117117117117, + "grad_norm": 25.646778106689453, + "learning_rate": 4.870886995591832e-06, + "loss": 6.966181755065918, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3056714031845331, + "loss_total_v6": 0.43538637459278107, + "step": 2302 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5311423423423424, + "grad_norm": 21.77490234375, + "learning_rate": 4.8670904361750385e-06, + "loss": 6.790583610534668, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.29573412612080574, + "loss_total_v6": 0.42441149428486824, + "step": 2303 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5313729729729729, + "grad_norm": 25.654150009155273, + "learning_rate": 4.8632939534405e-06, + "loss": 7.116656303405762, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3125125877559185, + "loss_total_v6": 0.4447910226881504, + "step": 2304 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5316036036036036, + "grad_norm": 23.12071418762207, + "learning_rate": 4.859497549578606e-06, + "loss": 7.197808265686035, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.3189883902668953, + "loss_total_v6": 0.4498630054295063, + "step": 2305 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5318342342342343, + "grad_norm": 22.350360870361328, + "learning_rate": 4.8557012267796895e-06, + "loss": 7.057483673095703, + "loss_alignment": 0.2677001953125, + "loss_alignment_w": 0.13385009765625, + "loss_sft": 0.30753257125616074, + "loss_total_v6": 0.44109274074435234, + "step": 2306 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5320648648648648, + "grad_norm": 24.177627563476562, + "learning_rate": 4.851904987234047e-06, + "loss": 6.896204948425293, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3000465855002403, + "loss_total_v6": 0.431012786924839, + "step": 2307 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5322954954954955, + "grad_norm": 24.45041847229004, + "learning_rate": 4.848108833131915e-06, + "loss": 7.0856099128723145, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.31019069626927376, + "loss_total_v6": 0.44285061955451965, + "step": 2308 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5325261261261262, + "grad_norm": 23.098346710205078, + "learning_rate": 4.84431276666349e-06, + "loss": 7.062151908874512, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3108303137123585, + "loss_total_v6": 0.4413845017552376, + "step": 2309 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5327567567567567, + "grad_norm": 26.5402774810791, + "learning_rate": 4.840516790018916e-06, + "loss": 6.850193500518799, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2982085086405277, + "loss_total_v6": 0.4281371086835861, + "step": 2310 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5329873873873874, + "grad_norm": 26.611221313476562, + "learning_rate": 4.836720905388281e-06, + "loss": 6.813383102416992, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2970827892422676, + "loss_total_v6": 0.425836443901062, + "step": 2311 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.533218018018018, + "grad_norm": 24.097183227539062, + "learning_rate": 4.832925114961629e-06, + "loss": 6.814579963684082, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.2953723073005676, + "loss_total_v6": 0.4259112477302551, + "step": 2312 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5334486486486486, + "grad_norm": 29.22708511352539, + "learning_rate": 4.829129420928937e-06, + "loss": 6.893091201782227, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.300767557695508, + "loss_total_v6": 0.43081820756196976, + "step": 2313 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5336792792792793, + "grad_norm": 26.56529426574707, + "learning_rate": 4.825333825480136e-06, + "loss": 7.181431770324707, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3185141757130623, + "loss_total_v6": 0.4488394930958748, + "step": 2314 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5339099099099099, + "grad_norm": 22.822725296020508, + "learning_rate": 4.821538330805098e-06, + "loss": 6.732518196105957, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2896025739610195, + "loss_total_v6": 0.4207823909819126, + "step": 2315 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5341405405405405, + "grad_norm": 24.6739501953125, + "learning_rate": 4.817742939093636e-06, + "loss": 6.890065670013428, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.30060896277427673, + "loss_total_v6": 0.43062910437583923, + "step": 2316 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5343711711711712, + "grad_norm": 22.95769691467285, + "learning_rate": 4.813947652535503e-06, + "loss": 6.819155693054199, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.29683319479227066, + "loss_total_v6": 0.42619720846414566, + "step": 2317 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5346018018018018, + "grad_norm": 23.308456420898438, + "learning_rate": 4.81015247332039e-06, + "loss": 6.455322265625, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2736968845129013, + "loss_total_v6": 0.4034576267004013, + "step": 2318 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5348324324324324, + "grad_norm": 22.73480987548828, + "learning_rate": 4.806357403637934e-06, + "loss": 6.705160140991211, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.2881368361413479, + "loss_total_v6": 0.4190725050866604, + "step": 2319 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5350630630630631, + "grad_norm": 23.919645309448242, + "learning_rate": 4.802562445677696e-06, + "loss": 6.62481689453125, + "loss_alignment": 0.2637939453125, + "loss_alignment_w": 0.13189697265625, + "loss_sft": 0.28230670280754566, + "loss_total_v6": 0.4140510782599449, + "step": 2320 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5352936936936937, + "grad_norm": 24.13775062561035, + "learning_rate": 4.798767601629184e-06, + "loss": 6.976369857788086, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.30301229655742645, + "loss_total_v6": 0.43602313846349716, + "step": 2321 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5355243243243243, + "grad_norm": 25.930683135986328, + "learning_rate": 4.794972873681834e-06, + "loss": 6.83559513092041, + "loss_alignment": 0.2637939453125, + "loss_alignment_w": 0.13189697265625, + "loss_sft": 0.29554133489727974, + "loss_total_v6": 0.42722468450665474, + "step": 2322 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.535754954954955, + "grad_norm": 27.45391273498535, + "learning_rate": 4.791178264025017e-06, + "loss": 7.435245513916016, + "loss_alignment": 0.27392578125, + "loss_alignment_w": 0.136962890625, + "loss_sft": 0.3278009928762913, + "loss_total_v6": 0.4647028483450413, + "step": 2323 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5359855855855856, + "grad_norm": 24.02068328857422, + "learning_rate": 4.7873837748480374e-06, + "loss": 7.143968105316162, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3137312941253185, + "loss_total_v6": 0.44649801403284073, + "step": 2324 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5362162162162162, + "grad_norm": 23.560056686401367, + "learning_rate": 4.7835894083401235e-06, + "loss": 7.013336658477783, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.30454448238015175, + "loss_total_v6": 0.43833355233073235, + "step": 2325 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5364468468468468, + "grad_norm": 24.83784294128418, + "learning_rate": 4.779795166690443e-06, + "loss": 6.836069107055664, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.29718838818371296, + "loss_total_v6": 0.4272543117403984, + "step": 2326 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5366774774774775, + "grad_norm": 25.3743839263916, + "learning_rate": 4.776001052088081e-06, + "loss": 7.550318717956543, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.33876198902726173, + "loss_total_v6": 0.47189491987228394, + "step": 2327 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5369081081081081, + "grad_norm": 23.186269760131836, + "learning_rate": 4.772207066722057e-06, + "loss": 6.820955276489258, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2984563373029232, + "loss_total_v6": 0.4263097383081913, + "step": 2328 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5371387387387387, + "grad_norm": 24.60535430908203, + "learning_rate": 4.768413212781315e-06, + "loss": 6.911847114562988, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3044727575033903, + "loss_total_v6": 0.43199045211076736, + "step": 2329 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5373693693693694, + "grad_norm": 25.176177978515625, + "learning_rate": 4.764619492454719e-06, + "loss": 6.911217212677002, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.30046609230339527, + "loss_total_v6": 0.4319510720670223, + "step": 2330 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5376, + "grad_norm": 22.744220733642578, + "learning_rate": 4.760825907931058e-06, + "loss": 7.2655439376831055, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3209788165986538, + "loss_total_v6": 0.454096470028162, + "step": 2331 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5378306306306306, + "grad_norm": 26.247343063354492, + "learning_rate": 4.757032461399044e-06, + "loss": 7.108506202697754, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3121405001729727, + "loss_total_v6": 0.4442816264927387, + "step": 2332 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5380612612612613, + "grad_norm": 25.754167556762695, + "learning_rate": 4.753239155047311e-06, + "loss": 6.951609134674072, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3038145639002323, + "loss_total_v6": 0.4344755746424198, + "step": 2333 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5382918918918919, + "grad_norm": 25.408096313476562, + "learning_rate": 4.7494459910644044e-06, + "loss": 6.710899353027344, + "loss_alignment": 0.2637939453125, + "loss_alignment_w": 0.13189697265625, + "loss_sft": 0.28776310198009014, + "loss_total_v6": 0.4194311983883381, + "step": 2334 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5385225225225225, + "grad_norm": 24.043420791625977, + "learning_rate": 4.745652971638796e-06, + "loss": 6.9947404861450195, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3041604310274124, + "loss_total_v6": 0.4371712952852249, + "step": 2335 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5387531531531532, + "grad_norm": 31.20182228088379, + "learning_rate": 4.741860098958869e-06, + "loss": 6.707661151885986, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.2862179521471262, + "loss_total_v6": 0.41922882944345474, + "step": 2336 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5389837837837838, + "grad_norm": 24.85181999206543, + "learning_rate": 4.738067375212926e-06, + "loss": 6.573211669921875, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.2822094075381756, + "loss_total_v6": 0.4108257256448269, + "step": 2337 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5392144144144144, + "grad_norm": 36.67107391357422, + "learning_rate": 4.734274802589181e-06, + "loss": 6.881175994873047, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.29863429442048073, + "loss_total_v6": 0.43007349222898483, + "step": 2338 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5394450450450451, + "grad_norm": 25.1209774017334, + "learning_rate": 4.730482383275759e-06, + "loss": 7.03701639175415, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3067416064441204, + "loss_total_v6": 0.4398135133087635, + "step": 2339 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5396756756756756, + "grad_norm": 27.91943359375, + "learning_rate": 4.726690119460701e-06, + "loss": 6.795138359069824, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.2938215136528015, + "loss_total_v6": 0.4246961399912834, + "step": 2340 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5399063063063063, + "grad_norm": 23.69544219970703, + "learning_rate": 4.722898013331954e-06, + "loss": 6.8821001052856445, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.2964795082807541, + "loss_total_v6": 0.4301312416791916, + "step": 2341 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.540136936936937, + "grad_norm": 29.733898162841797, + "learning_rate": 4.7191060670773785e-06, + "loss": 6.543973922729492, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.27902399003505707, + "loss_total_v6": 0.40899834781885147, + "step": 2342 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5403675675675675, + "grad_norm": 23.638965606689453, + "learning_rate": 4.715314282884739e-06, + "loss": 6.641596794128418, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.28451510332524776, + "loss_total_v6": 0.4150998182594776, + "step": 2343 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5405981981981982, + "grad_norm": 24.909866333007812, + "learning_rate": 4.7115226629417075e-06, + "loss": 6.738095283508301, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.291431225836277, + "loss_total_v6": 0.421130932867527, + "step": 2344 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5408288288288289, + "grad_norm": 25.79358673095703, + "learning_rate": 4.707731209435864e-06, + "loss": 7.719830513000488, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.3486850820481777, + "loss_total_v6": 0.4824893958866596, + "step": 2345 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5410594594594594, + "grad_norm": 24.93780517578125, + "learning_rate": 4.703939924554687e-06, + "loss": 6.898202896118164, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.29892027750611305, + "loss_total_v6": 0.43113768100738525, + "step": 2346 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5412900900900901, + "grad_norm": 24.827627182006836, + "learning_rate": 4.700148810485561e-06, + "loss": 6.667941570281982, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.28489512763917446, + "loss_total_v6": 0.4167463406920433, + "step": 2347 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5415207207207208, + "grad_norm": 26.27359962463379, + "learning_rate": 4.696357869415774e-06, + "loss": 7.005246162414551, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.305000152438879, + "loss_total_v6": 0.4378278814256191, + "step": 2348 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5417513513513513, + "grad_norm": 23.144136428833008, + "learning_rate": 4.692567103532509e-06, + "loss": 6.916484832763672, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2997424341738224, + "loss_total_v6": 0.4322802796959877, + "step": 2349 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.541981981981982, + "grad_norm": 27.69058609008789, + "learning_rate": 4.688776515022855e-06, + "loss": 7.152886867523193, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.3148227632045746, + "loss_total_v6": 0.447055421769619, + "step": 2350 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5422126126126127, + "grad_norm": 25.46179962158203, + "learning_rate": 4.6849861060737895e-06, + "loss": 6.790127277374268, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.2943170201033354, + "loss_total_v6": 0.42438294366002083, + "step": 2351 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5424432432432432, + "grad_norm": 28.932735443115234, + "learning_rate": 4.681195878872195e-06, + "loss": 6.671404838562012, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.2899181507527828, + "loss_total_v6": 0.4169628247618675, + "step": 2352 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5426738738738739, + "grad_norm": 26.494338989257812, + "learning_rate": 4.677405835604842e-06, + "loss": 6.836925983428955, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.2962348647415638, + "loss_total_v6": 0.4273078702390194, + "step": 2353 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5429045045045046, + "grad_norm": 23.231250762939453, + "learning_rate": 4.673615978458402e-06, + "loss": 6.821381568908691, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2969417981803417, + "loss_total_v6": 0.4263363257050514, + "step": 2354 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5431351351351351, + "grad_norm": 23.473764419555664, + "learning_rate": 4.669826309619437e-06, + "loss": 6.773322582244873, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.29372451454401016, + "loss_total_v6": 0.42333266884088516, + "step": 2355 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5433657657657658, + "grad_norm": 27.695131301879883, + "learning_rate": 4.666036831274392e-06, + "loss": 6.724067687988281, + "loss_alignment": 0.269775390625, + "loss_alignment_w": 0.1348876953125, + "loss_sft": 0.2856564298272133, + "loss_total_v6": 0.4202542155981064, + "step": 2356 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5435963963963963, + "grad_norm": 22.104202270507812, + "learning_rate": 4.662247545609616e-06, + "loss": 6.637459754943848, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.28425654023885727, + "loss_total_v6": 0.4148412421345711, + "step": 2357 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.543827027027027, + "grad_norm": 27.910037994384766, + "learning_rate": 4.658458454811335e-06, + "loss": 7.4025983810424805, + "loss_alignment": 0.2667236328125, + "loss_alignment_w": 0.13336181640625, + "loss_sft": 0.3294074051082134, + "loss_total_v6": 0.4626624062657356, + "step": 2358 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5440576576576577, + "grad_norm": 24.677413940429688, + "learning_rate": 4.654669561065672e-06, + "loss": 6.694863319396973, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.29045349545776844, + "loss_total_v6": 0.418428972363472, + "step": 2359 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5442882882882882, + "grad_norm": 50.853946685791016, + "learning_rate": 4.650880866558626e-06, + "loss": 7.03530216217041, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3090301416814327, + "loss_total_v6": 0.4397064037621021, + "step": 2360 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5445189189189189, + "grad_norm": 25.25769805908203, + "learning_rate": 4.647092373476091e-06, + "loss": 7.262920379638672, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3224627748131752, + "loss_total_v6": 0.4539325125515461, + "step": 2361 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5447495495495496, + "grad_norm": 23.92314338684082, + "learning_rate": 4.643304084003839e-06, + "loss": 7.305537700653076, + "loss_alignment": 0.2681884765625, + "loss_alignment_w": 0.13409423828125, + "loss_sft": 0.32248660922050476, + "loss_total_v6": 0.45659609138965607, + "step": 2362 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5449801801801801, + "grad_norm": 26.932723999023438, + "learning_rate": 4.6395160003275254e-06, + "loss": 7.076979637145996, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.31201646104454994, + "loss_total_v6": 0.44231124967336655, + "step": 2363 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5452108108108108, + "grad_norm": 22.403079986572266, + "learning_rate": 4.6357281246326916e-06, + "loss": 7.185832977294922, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.31808731332421303, + "loss_total_v6": 0.4491145461797714, + "step": 2364 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5454414414414415, + "grad_norm": 27.745769500732422, + "learning_rate": 4.631940459104749e-06, + "loss": 6.3933329582214355, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2666182182729244, + "loss_total_v6": 0.3995833024382591, + "step": 2365 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.545672072072072, + "grad_norm": 25.235275268554688, + "learning_rate": 4.628153005928998e-06, + "loss": 7.111297607421875, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3117961995303631, + "loss_total_v6": 0.4444561116397381, + "step": 2366 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5459027027027027, + "grad_norm": 25.486797332763672, + "learning_rate": 4.624365767290609e-06, + "loss": 6.888245582580566, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2991218902170658, + "loss_total_v6": 0.4305153302848339, + "step": 2367 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5461333333333334, + "grad_norm": 23.178329467773438, + "learning_rate": 4.620578745374634e-06, + "loss": 6.797460556030273, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.29106750153005123, + "loss_total_v6": 0.4248412922024727, + "step": 2368 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5463639639639639, + "grad_norm": 22.573198318481445, + "learning_rate": 4.616791942365997e-06, + "loss": 7.058071136474609, + "loss_alignment": 0.27001953125, + "loss_alignment_w": 0.135009765625, + "loss_sft": 0.3064553737640381, + "loss_total_v6": 0.4411294534802437, + "step": 2369 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5465945945945946, + "grad_norm": 29.770244598388672, + "learning_rate": 4.613005360449496e-06, + "loss": 6.918875694274902, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.2995409481227398, + "loss_total_v6": 0.4324297346174717, + "step": 2370 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5468252252252253, + "grad_norm": 23.668331146240234, + "learning_rate": 4.609219001809806e-06, + "loss": 6.786441802978516, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.29375097155570984, + "loss_total_v6": 0.42415259033441544, + "step": 2371 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5470558558558558, + "grad_norm": 25.6221923828125, + "learning_rate": 4.6054328686314616e-06, + "loss": 6.978791236877441, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3053761124610901, + "loss_total_v6": 0.4361744448542595, + "step": 2372 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5472864864864865, + "grad_norm": 24.812204360961914, + "learning_rate": 4.60164696309888e-06, + "loss": 7.0558013916015625, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3103113118559122, + "loss_total_v6": 0.44098757579922676, + "step": 2373 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5475171171171171, + "grad_norm": 24.769474029541016, + "learning_rate": 4.5978612873963415e-06, + "loss": 6.787352561950684, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.2951506916433573, + "loss_total_v6": 0.4242095313966274, + "step": 2374 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5477477477477477, + "grad_norm": 24.27614974975586, + "learning_rate": 4.594075843707991e-06, + "loss": 6.970036506652832, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3053477331995964, + "loss_total_v6": 0.4356272667646408, + "step": 2375 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5479783783783784, + "grad_norm": 25.707582473754883, + "learning_rate": 4.590290634217848e-06, + "loss": 7.1317219734191895, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.31308795884251595, + "loss_total_v6": 0.44573261216282845, + "step": 2376 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.548209009009009, + "grad_norm": 24.588977813720703, + "learning_rate": 4.586505661109787e-06, + "loss": 6.676303386688232, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2854940667748451, + "loss_total_v6": 0.4172689616680145, + "step": 2377 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5484396396396396, + "grad_norm": 24.490455627441406, + "learning_rate": 4.582720926567552e-06, + "loss": 6.954046249389648, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3029598016291857, + "loss_total_v6": 0.43462788313627243, + "step": 2378 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5486702702702703, + "grad_norm": 24.439085006713867, + "learning_rate": 4.5789364327747475e-06, + "loss": 7.439548492431641, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.33457015082240105, + "loss_total_v6": 0.46497175842523575, + "step": 2379 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5489009009009009, + "grad_norm": 25.522554397583008, + "learning_rate": 4.57515218191484e-06, + "loss": 6.89884090423584, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.29793782718479633, + "loss_total_v6": 0.4311775639653206, + "step": 2380 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5491315315315315, + "grad_norm": 24.56102180480957, + "learning_rate": 4.571368176171159e-06, + "loss": 7.107476234436035, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3129764162003994, + "loss_total_v6": 0.444217249751091, + "step": 2381 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5493621621621622, + "grad_norm": 22.70916748046875, + "learning_rate": 4.567584417726883e-06, + "loss": 6.735555648803711, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.286893242970109, + "loss_total_v6": 0.42097223550081253, + "step": 2382 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5495927927927928, + "grad_norm": 25.36695098876953, + "learning_rate": 4.563800908765058e-06, + "loss": 6.740849494934082, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.29229002818465233, + "loss_total_v6": 0.42130308598279953, + "step": 2383 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5498234234234234, + "grad_norm": 26.767906188964844, + "learning_rate": 4.560017651468579e-06, + "loss": 7.038601875305176, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3090379498898983, + "loss_total_v6": 0.4399125874042511, + "step": 2384 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.550054054054054, + "grad_norm": 25.352096557617188, + "learning_rate": 4.556234648020203e-06, + "loss": 6.822966575622559, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.2925853207707405, + "loss_total_v6": 0.4264354109764099, + "step": 2385 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5502846846846847, + "grad_norm": 26.996599197387695, + "learning_rate": 4.552451900602531e-06, + "loss": 7.633289813995361, + "loss_alignment": 0.269287109375, + "loss_alignment_w": 0.1346435546875, + "loss_sft": 0.3426659405231476, + "loss_total_v6": 0.4770806171000004, + "step": 2386 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5505153153153153, + "grad_norm": 28.227649688720703, + "learning_rate": 4.548669411398025e-06, + "loss": 7.094828128814697, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3118960075080395, + "loss_total_v6": 0.4434267617762089, + "step": 2387 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.550745945945946, + "grad_norm": 25.232351303100586, + "learning_rate": 4.544887182588995e-06, + "loss": 6.539247512817383, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2790642958134413, + "loss_total_v6": 0.4087029844522476, + "step": 2388 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5509765765765766, + "grad_norm": 24.5053768157959, + "learning_rate": 4.5411052163575986e-06, + "loss": 7.271790504455566, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3232613205909729, + "loss_total_v6": 0.4544868990778923, + "step": 2389 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5512072072072072, + "grad_norm": 25.16229248046875, + "learning_rate": 4.537323514885846e-06, + "loss": 6.901182174682617, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3025396801531315, + "loss_total_v6": 0.4313238672912121, + "step": 2390 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5514378378378378, + "grad_norm": 20.052391052246094, + "learning_rate": 4.533542080355587e-06, + "loss": 7.3029584884643555, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3258806802332401, + "loss_total_v6": 0.4564348869025707, + "step": 2391 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5516684684684685, + "grad_norm": 27.099550247192383, + "learning_rate": 4.52976091494853e-06, + "loss": 7.506237983703613, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.33513718843460083, + "loss_total_v6": 0.4691398926079273, + "step": 2392 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5518990990990991, + "grad_norm": 23.34792137145996, + "learning_rate": 4.525980020846214e-06, + "loss": 6.797589302062988, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.29597358778119087, + "loss_total_v6": 0.4248493127524853, + "step": 2393 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5521297297297297, + "grad_norm": 21.345046997070312, + "learning_rate": 4.522199400230033e-06, + "loss": 6.5405144691467285, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.28019633889198303, + "loss_total_v6": 0.40878215059638023, + "step": 2394 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5523603603603604, + "grad_norm": 24.24677848815918, + "learning_rate": 4.518419055281217e-06, + "loss": 6.626779079437256, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.2829023413360119, + "loss_total_v6": 0.4141736850142479, + "step": 2395 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.552590990990991, + "grad_norm": 22.82808494567871, + "learning_rate": 4.514638988180838e-06, + "loss": 7.398822784423828, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.32961393892765045, + "loss_total_v6": 0.46242644265294075, + "step": 2396 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5528216216216216, + "grad_norm": 24.15130043029785, + "learning_rate": 4.510859201109809e-06, + "loss": 7.06944465637207, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3113928884267807, + "loss_total_v6": 0.4418402798473835, + "step": 2397 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5530522522522523, + "grad_norm": 24.185651779174805, + "learning_rate": 4.50707969624888e-06, + "loss": 7.198061466217041, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.31909578293561935, + "loss_total_v6": 0.44987885281443596, + "step": 2398 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5532828828828829, + "grad_norm": 24.514894485473633, + "learning_rate": 4.503300475778638e-06, + "loss": 7.272867679595947, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.3215433731675148, + "loss_total_v6": 0.4545542262494564, + "step": 2399 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5535135135135135, + "grad_norm": 29.044937133789062, + "learning_rate": 4.499521541879508e-06, + "loss": 7.9887776374816895, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.3641515262424946, + "loss_total_v6": 0.4992986172437668, + "step": 2400 + }, + { + "epoch": 0.5535135135135135, + "eval_loss": 0.581539511680603, + "eval_loss_alignment": 0.23632589540525115, + "eval_loss_alignment_w": 0.11816294770262557, + "eval_loss_sft": 0.4633765365826349, + "eval_loss_total_v6": 0.5815394843447971, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.7871, + "eval_samples_per_second": 19.011, + "eval_steps_per_second": 2.386, + "step": 2400 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5537441441441442, + "grad_norm": 23.890703201293945, + "learning_rate": 4.495742896731748e-06, + "loss": 6.9432759284973145, + "loss_alignment": 0.267822265625, + "loss_alignment_w": 0.1339111328125, + "loss_sft": 0.29987576231360435, + "loss_total_v6": 0.43395473808050156, + "step": 2401 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5539747747747747, + "grad_norm": 28.423500061035156, + "learning_rate": 4.4919645425154524e-06, + "loss": 7.0634002685546875, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.3088636137545109, + "loss_total_v6": 0.4414624944329262, + "step": 2402 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5542054054054054, + "grad_norm": 27.180400848388672, + "learning_rate": 4.488186481410539e-06, + "loss": 6.962559223175049, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.30137090384960175, + "loss_total_v6": 0.43515996262431145, + "step": 2403 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5544360360360361, + "grad_norm": 25.00650405883789, + "learning_rate": 4.484408715596768e-06, + "loss": 6.993022918701172, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.30707431584596634, + "loss_total_v6": 0.43706392496824265, + "step": 2404 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5546666666666666, + "grad_norm": 23.888187408447266, + "learning_rate": 4.48063124725372e-06, + "loss": 6.978054046630859, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.3078782632946968, + "loss_total_v6": 0.4361283741891384, + "step": 2405 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5548972972972973, + "grad_norm": 23.182350158691406, + "learning_rate": 4.47685407856081e-06, + "loss": 6.699715614318848, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.2865300886332989, + "loss_total_v6": 0.4187322221696377, + "step": 2406 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.555127927927928, + "grad_norm": 24.607791900634766, + "learning_rate": 4.47307721169728e-06, + "loss": 7.843229293823242, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3562296889722347, + "loss_total_v6": 0.49020184203982353, + "step": 2407 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5553585585585585, + "grad_norm": 23.082468032836914, + "learning_rate": 4.469300648842192e-06, + "loss": 6.883604049682617, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.29939636774361134, + "loss_total_v6": 0.4302252382040024, + "step": 2408 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5555891891891892, + "grad_norm": 24.340810775756836, + "learning_rate": 4.465524392174437e-06, + "loss": 6.783358573913574, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2941076084971428, + "loss_total_v6": 0.4239599071443081, + "step": 2409 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5558198198198199, + "grad_norm": 23.317941665649414, + "learning_rate": 4.461748443872729e-06, + "loss": 6.94917106628418, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.30421146750450134, + "loss_total_v6": 0.43432318046689034, + "step": 2410 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5560504504504504, + "grad_norm": 22.991863250732422, + "learning_rate": 4.457972806115607e-06, + "loss": 7.04859733581543, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.30812153220176697, + "loss_total_v6": 0.44053731858730316, + "step": 2411 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5562810810810811, + "grad_norm": 23.632768630981445, + "learning_rate": 4.454197481081423e-06, + "loss": 6.654721736907959, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.28614412248134613, + "loss_total_v6": 0.41592010483145714, + "step": 2412 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5565117117117118, + "grad_norm": 24.733606338500977, + "learning_rate": 4.450422470948354e-06, + "loss": 6.661828517913818, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.2890602331608534, + "loss_total_v6": 0.41636429354548454, + "step": 2413 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5567423423423423, + "grad_norm": 26.673662185668945, + "learning_rate": 4.446647777894397e-06, + "loss": 7.621460914611816, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.34523778036236763, + "loss_total_v6": 0.47634128481149673, + "step": 2414 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.556972972972973, + "grad_norm": 27.8245792388916, + "learning_rate": 4.442873404097362e-06, + "loss": 7.102451324462891, + "loss_alignment": 0.270263671875, + "loss_alignment_w": 0.1351318359375, + "loss_sft": 0.30877135321497917, + "loss_total_v6": 0.44390319660305977, + "step": 2415 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5572036036036037, + "grad_norm": 24.30681037902832, + "learning_rate": 4.4390993517348784e-06, + "loss": 6.798089981079102, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2950588297098875, + "loss_total_v6": 0.42488061264157295, + "step": 2416 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5574342342342342, + "grad_norm": 27.882112503051758, + "learning_rate": 4.435325622984383e-06, + "loss": 6.76116943359375, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.28969957679510117, + "loss_total_v6": 0.4225730933248997, + "step": 2417 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5576648648648649, + "grad_norm": 21.802221298217773, + "learning_rate": 4.431552220023138e-06, + "loss": 6.483359336853027, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.27537292428314686, + "loss_total_v6": 0.4052099622786045, + "step": 2418 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5578954954954954, + "grad_norm": 24.393836975097656, + "learning_rate": 4.427779145028206e-06, + "loss": 7.0598931312561035, + "loss_alignment": 0.2652587890625, + "loss_alignment_w": 0.13262939453125, + "loss_sft": 0.3087360020726919, + "loss_total_v6": 0.44124332070350647, + "step": 2419 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5581261261261261, + "grad_norm": 26.826486587524414, + "learning_rate": 4.424006400176465e-06, + "loss": 7.027684688568115, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.308523528277874, + "loss_total_v6": 0.4392302967607975, + "step": 2420 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5583567567567568, + "grad_norm": 24.625497817993164, + "learning_rate": 4.420233987644605e-06, + "loss": 7.104125499725342, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.31134791672229767, + "loss_total_v6": 0.44400782510638237, + "step": 2421 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5585873873873873, + "grad_norm": 23.04331398010254, + "learning_rate": 4.416461909609119e-06, + "loss": 7.003910064697266, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.307525884360075, + "loss_total_v6": 0.4377443864941597, + "step": 2422 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.558818018018018, + "grad_norm": 23.772661209106445, + "learning_rate": 4.412690168246315e-06, + "loss": 6.75424861907959, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.2912200763821602, + "loss_total_v6": 0.4221405163407326, + "step": 2423 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5590486486486487, + "grad_norm": 24.681495666503906, + "learning_rate": 4.408918765732295e-06, + "loss": 7.115171432495117, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.31332002580165863, + "loss_total_v6": 0.4446982145309448, + "step": 2424 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5592792792792792, + "grad_norm": 29.31487274169922, + "learning_rate": 4.405147704242976e-06, + "loss": 7.044853687286377, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.30829956382513046, + "loss_total_v6": 0.44030335545539856, + "step": 2425 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5595099099099099, + "grad_norm": 26.594541549682617, + "learning_rate": 4.401376985954073e-06, + "loss": 7.204769134521484, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.32068993151187897, + "loss_total_v6": 0.45029808208346367, + "step": 2426 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5597405405405406, + "grad_norm": 25.234100341796875, + "learning_rate": 4.397606613041106e-06, + "loss": 6.966120719909668, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.30531665682792664, + "loss_total_v6": 0.43538255989551544, + "step": 2427 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5599711711711711, + "grad_norm": 21.942781448364258, + "learning_rate": 4.393836587679394e-06, + "loss": 6.744932174682617, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.2937659062445164, + "loss_total_v6": 0.4215582571923733, + "step": 2428 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5602018018018018, + "grad_norm": 24.489112854003906, + "learning_rate": 4.3900669120440524e-06, + "loss": 7.256481647491455, + "loss_alignment": 0.2711181640625, + "loss_alignment_w": 0.13555908203125, + "loss_sft": 0.3183830250054598, + "loss_total_v6": 0.45353010669350624, + "step": 2429 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5604324324324325, + "grad_norm": 25.549386978149414, + "learning_rate": 4.386297588310002e-06, + "loss": 7.201481819152832, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3180736005306244, + "loss_total_v6": 0.4500926360487938, + "step": 2430 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.560663063063063, + "grad_norm": 25.977903366088867, + "learning_rate": 4.382528618651953e-06, + "loss": 6.998175621032715, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.30625196546316147, + "loss_total_v6": 0.43738599121570587, + "step": 2431 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5608936936936937, + "grad_norm": 27.983978271484375, + "learning_rate": 4.378760005244417e-06, + "loss": 7.65503454208374, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.34408601373434067, + "loss_total_v6": 0.47843965142965317, + "step": 2432 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5611243243243244, + "grad_norm": 25.63047981262207, + "learning_rate": 4.3749917502617e-06, + "loss": 6.6571550369262695, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.2828782219439745, + "loss_total_v6": 0.41607218608260155, + "step": 2433 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5613549549549549, + "grad_norm": 26.562652587890625, + "learning_rate": 4.371223855877895e-06, + "loss": 6.96571159362793, + "loss_alignment": 0.2642822265625, + "loss_alignment_w": 0.13214111328125, + "loss_sft": 0.3033989332616329, + "loss_total_v6": 0.4353569485247135, + "step": 2434 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5615855855855856, + "grad_norm": 28.83562660217285, + "learning_rate": 4.367456324266893e-06, + "loss": 6.974555015563965, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.3058437667787075, + "loss_total_v6": 0.4359096884727478, + "step": 2435 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5618162162162162, + "grad_norm": 33.38984298706055, + "learning_rate": 4.363689157602373e-06, + "loss": 7.6613945960998535, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.34646715968847275, + "loss_total_v6": 0.47883715108036995, + "step": 2436 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5620468468468468, + "grad_norm": 29.507722854614258, + "learning_rate": 4.359922358057807e-06, + "loss": 6.693565845489502, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2880835700780153, + "loss_total_v6": 0.4183478616178036, + "step": 2437 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5622774774774775, + "grad_norm": 23.328527450561523, + "learning_rate": 4.356155927806448e-06, + "loss": 7.143925666809082, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31502560526132584, + "loss_total_v6": 0.44649532809853554, + "step": 2438 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5625081081081081, + "grad_norm": 26.888418197631836, + "learning_rate": 4.3523898690213414e-06, + "loss": 6.85639762878418, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.2957123424857855, + "loss_total_v6": 0.42852484062314034, + "step": 2439 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5627387387387387, + "grad_norm": 27.64748764038086, + "learning_rate": 4.3486241838753225e-06, + "loss": 6.66580057144165, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.28485288470983505, + "loss_total_v6": 0.41661251336336136, + "step": 2440 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5629693693693694, + "grad_norm": 25.294189453125, + "learning_rate": 4.344858874540999e-06, + "loss": 6.6572041511535645, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.28521587885916233, + "loss_total_v6": 0.4160752557218075, + "step": 2441 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5632, + "grad_norm": 24.843780517578125, + "learning_rate": 4.3410939431907726e-06, + "loss": 7.066916465759277, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3107161298394203, + "loss_total_v6": 0.4416823051869869, + "step": 2442 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5634306306306306, + "grad_norm": 24.075292587280273, + "learning_rate": 4.337329391996821e-06, + "loss": 6.775472640991211, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.2914022319018841, + "loss_total_v6": 0.4234670475125313, + "step": 2443 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5636612612612613, + "grad_norm": 24.74888038635254, + "learning_rate": 4.333565223131107e-06, + "loss": 7.420553207397461, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3324064128100872, + "loss_total_v6": 0.4637845680117607, + "step": 2444 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5638918918918919, + "grad_norm": 23.370506286621094, + "learning_rate": 4.329801438765366e-06, + "loss": 6.973532676696777, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3066343776881695, + "loss_total_v6": 0.4358458034694195, + "step": 2445 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5641225225225225, + "grad_norm": 27.827058792114258, + "learning_rate": 4.3260380410711186e-06, + "loss": 6.694866180419922, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.2870662212371826, + "loss_total_v6": 0.4184291325509548, + "step": 2446 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5643531531531532, + "grad_norm": 27.947153091430664, + "learning_rate": 4.32227503221966e-06, + "loss": 6.799894332885742, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2930353730916977, + "loss_total_v6": 0.4249933809041977, + "step": 2447 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5645837837837838, + "grad_norm": 24.35841178894043, + "learning_rate": 4.318512414382059e-06, + "loss": 6.866962432861328, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.3010723628103733, + "loss_total_v6": 0.4291851595044136, + "step": 2448 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5648144144144144, + "grad_norm": 25.70941162109375, + "learning_rate": 4.314750189729163e-06, + "loss": 6.784838676452637, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.29267422668635845, + "loss_total_v6": 0.4240524247288704, + "step": 2449 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.565045045045045, + "grad_norm": 25.49970054626465, + "learning_rate": 4.310988360431585e-06, + "loss": 6.762458801269531, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.29127553664147854, + "loss_total_v6": 0.4226536937057972, + "step": 2450 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5652756756756757, + "grad_norm": 24.67001724243164, + "learning_rate": 4.30722692865972e-06, + "loss": 7.0985426902771, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.31366928294301033, + "loss_total_v6": 0.4436589181423187, + "step": 2451 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5655063063063063, + "grad_norm": 26.116769790649414, + "learning_rate": 4.3034658965837255e-06, + "loss": 7.020713806152344, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3059515804052353, + "loss_total_v6": 0.4387946054339409, + "step": 2452 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5657369369369369, + "grad_norm": 24.903560638427734, + "learning_rate": 4.299705266373531e-06, + "loss": 7.100954055786133, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.310691948980093, + "loss_total_v6": 0.443809624761343, + "step": 2453 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5659675675675676, + "grad_norm": 24.13538360595703, + "learning_rate": 4.29594504019884e-06, + "loss": 6.895208358764648, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3017390798777342, + "loss_total_v6": 0.43095050007104874, + "step": 2454 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5661981981981982, + "grad_norm": 25.510372161865234, + "learning_rate": 4.29218522022911e-06, + "loss": 6.858635425567627, + "loss_alignment": 0.268798828125, + "loss_alignment_w": 0.1343994140625, + "loss_sft": 0.29444840364158154, + "loss_total_v6": 0.4286647215485573, + "step": 2455 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5664288288288288, + "grad_norm": 24.09488868713379, + "learning_rate": 4.2884258086335755e-06, + "loss": 7.065948009490967, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.31077760830521584, + "loss_total_v6": 0.4416217468678951, + "step": 2456 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5666594594594595, + "grad_norm": 26.325611114501953, + "learning_rate": 4.28466680758123e-06, + "loss": 6.679950714111328, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.2898723818361759, + "loss_total_v6": 0.4174968898296356, + "step": 2457 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5668900900900901, + "grad_norm": 29.113462448120117, + "learning_rate": 4.280908219240832e-06, + "loss": 7.121448040008545, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31394121050834656, + "loss_total_v6": 0.44509050622582436, + "step": 2458 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5671207207207207, + "grad_norm": 25.837841033935547, + "learning_rate": 4.2771500457809045e-06, + "loss": 7.220874786376953, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3198196738958359, + "loss_total_v6": 0.4513046592473984, + "step": 2459 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5673513513513514, + "grad_norm": 26.558820724487305, + "learning_rate": 4.273392289369723e-06, + "loss": 7.253123760223389, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.319287046790123, + "loss_total_v6": 0.4533202387392521, + "step": 2460 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.567581981981982, + "grad_norm": 25.191551208496094, + "learning_rate": 4.26963495217533e-06, + "loss": 6.760424613952637, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.289836123585701, + "loss_total_v6": 0.422526553273201, + "step": 2461 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5678126126126126, + "grad_norm": 24.998876571655273, + "learning_rate": 4.265878036365524e-06, + "loss": 7.067364692687988, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31138496845960617, + "loss_total_v6": 0.44171028956770897, + "step": 2462 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5680432432432433, + "grad_norm": 23.06624984741211, + "learning_rate": 4.262121544107861e-06, + "loss": 6.7761664390563965, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.2942684441804886, + "loss_total_v6": 0.4235103912651539, + "step": 2463 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5682738738738738, + "grad_norm": 27.46234130859375, + "learning_rate": 4.258365477569648e-06, + "loss": 6.923491477966309, + "loss_alignment": 0.2666015625, + "loss_alignment_w": 0.13330078125, + "loss_sft": 0.2992648798972368, + "loss_total_v6": 0.4327182471752167, + "step": 2464 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5685045045045045, + "grad_norm": 23.38113784790039, + "learning_rate": 4.254609838917951e-06, + "loss": 6.860725402832031, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.2987599205225706, + "loss_total_v6": 0.42879531905055046, + "step": 2465 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5687351351351352, + "grad_norm": 25.643775939941406, + "learning_rate": 4.250854630319593e-06, + "loss": 7.182262420654297, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.31845929101109505, + "loss_total_v6": 0.44889140501618385, + "step": 2466 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5689657657657657, + "grad_norm": 22.575862884521484, + "learning_rate": 4.247099853941138e-06, + "loss": 6.80858850479126, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.29458586499094963, + "loss_total_v6": 0.42553678527474403, + "step": 2467 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5691963963963964, + "grad_norm": 21.90786361694336, + "learning_rate": 4.24334551194891e-06, + "loss": 6.924437522888184, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.30365744419395924, + "loss_total_v6": 0.4327773302793503, + "step": 2468 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5694270270270271, + "grad_norm": 22.32648468017578, + "learning_rate": 4.239591606508977e-06, + "loss": 6.744753360748291, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.2907182462513447, + "loss_total_v6": 0.4215470850467682, + "step": 2469 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5696576576576576, + "grad_norm": 24.86249542236328, + "learning_rate": 4.235838139787161e-06, + "loss": 7.416515350341797, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.3310248889029026, + "loss_total_v6": 0.463532205671072, + "step": 2470 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5698882882882883, + "grad_norm": 23.320859909057617, + "learning_rate": 4.232085113949022e-06, + "loss": 6.995655059814453, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3083679750561714, + "loss_total_v6": 0.4372284524142742, + "step": 2471 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.570118918918919, + "grad_norm": 21.569990158081055, + "learning_rate": 4.228332531159872e-06, + "loss": 6.880546569824219, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.30019713565707207, + "loss_total_v6": 0.43003417178988457, + "step": 2472 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5703495495495495, + "grad_norm": 25.039331436157227, + "learning_rate": 4.224580393584768e-06, + "loss": 6.814657688140869, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.2994512468576431, + "loss_total_v6": 0.425916101783514, + "step": 2473 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5705801801801802, + "grad_norm": 25.71427345275879, + "learning_rate": 4.220828703388507e-06, + "loss": 7.000898361206055, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.30651363357901573, + "loss_total_v6": 0.4375561289489269, + "step": 2474 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5708108108108109, + "grad_norm": 25.868938446044922, + "learning_rate": 4.217077462735632e-06, + "loss": 6.512460708618164, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.27488769218325615, + "loss_total_v6": 0.40702881291508675, + "step": 2475 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5710414414414414, + "grad_norm": 26.519214630126953, + "learning_rate": 4.213326673790418e-06, + "loss": 7.14125394821167, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31565210595726967, + "loss_total_v6": 0.44632837176322937, + "step": 2476 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5712720720720721, + "grad_norm": 25.542943954467773, + "learning_rate": 4.2095763387168895e-06, + "loss": 7.108092308044434, + "loss_alignment": 0.2657470703125, + "loss_alignment_w": 0.13287353515625, + "loss_sft": 0.31150432862341404, + "loss_total_v6": 0.4442557841539383, + "step": 2477 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5715027027027026, + "grad_norm": 25.4780330657959, + "learning_rate": 4.205826459678802e-06, + "loss": 6.906759262084961, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.29861580207943916, + "loss_total_v6": 0.43167243525385857, + "step": 2478 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5717333333333333, + "grad_norm": 22.880945205688477, + "learning_rate": 4.202077038839652e-06, + "loss": 6.887529373168945, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.29895505867898464, + "loss_total_v6": 0.4304705671966076, + "step": 2479 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.571963963963964, + "grad_norm": 25.81584358215332, + "learning_rate": 4.198328078362672e-06, + "loss": 6.684865951538086, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.2887757960706949, + "loss_total_v6": 0.41780412569642067, + "step": 2480 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5721945945945945, + "grad_norm": 24.535524368286133, + "learning_rate": 4.194579580410825e-06, + "loss": 6.456699371337891, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.27379821985960007, + "loss_total_v6": 0.40354370698332787, + "step": 2481 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5724252252252252, + "grad_norm": 20.280244827270508, + "learning_rate": 4.190831547146809e-06, + "loss": 7.3589019775390625, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.3281259536743164, + "loss_total_v6": 0.4599313735961914, + "step": 2482 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5726558558558559, + "grad_norm": 25.469310760498047, + "learning_rate": 4.1870839807330545e-06, + "loss": 7.247457504272461, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.32029092870652676, + "loss_total_v6": 0.4529660828411579, + "step": 2483 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5728864864864864, + "grad_norm": 24.063047409057617, + "learning_rate": 4.183336883331723e-06, + "loss": 7.017457485198975, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3054733909666538, + "loss_total_v6": 0.4385910779237747, + "step": 2484 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5731171171171171, + "grad_norm": 23.321231842041016, + "learning_rate": 4.179590257104708e-06, + "loss": 6.6666765213012695, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.28234415501356125, + "loss_total_v6": 0.41666726022958755, + "step": 2485 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5733477477477478, + "grad_norm": 25.955026626586914, + "learning_rate": 4.175844104213621e-06, + "loss": 6.808345317840576, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2976834550499916, + "loss_total_v6": 0.4255215860903263, + "step": 2486 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5735783783783783, + "grad_norm": 25.888545989990234, + "learning_rate": 4.172098426819812e-06, + "loss": 7.416672229766846, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.33265212550759315, + "loss_total_v6": 0.46354200690984726, + "step": 2487 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.573809009009009, + "grad_norm": 27.34287452697754, + "learning_rate": 4.1683532270843505e-06, + "loss": 6.9955291748046875, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.30536936968564987, + "loss_total_v6": 0.4372205585241318, + "step": 2488 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5740396396396397, + "grad_norm": 27.347742080688477, + "learning_rate": 4.164608507168034e-06, + "loss": 6.67839241027832, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.28582301549613476, + "loss_total_v6": 0.4173995442688465, + "step": 2489 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5742702702702702, + "grad_norm": 24.596776962280273, + "learning_rate": 4.1608642692313765e-06, + "loss": 6.551122188568115, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.27832636423408985, + "loss_total_v6": 0.4094451405107975, + "step": 2490 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5745009009009009, + "grad_norm": 26.00006866455078, + "learning_rate": 4.1571205154346215e-06, + "loss": 7.271535396575928, + "loss_alignment": 0.271240234375, + "loss_alignment_w": 0.1356201171875, + "loss_sft": 0.3186829797923565, + "loss_total_v6": 0.4544709511101246, + "step": 2491 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5747315315315316, + "grad_norm": 27.910491943359375, + "learning_rate": 4.1533772479377325e-06, + "loss": 6.537539482116699, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.2769281379878521, + "loss_total_v6": 0.4085962064564228, + "step": 2492 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5749621621621621, + "grad_norm": 25.30719757080078, + "learning_rate": 4.149634468900386e-06, + "loss": 7.0332746505737305, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3070418071001768, + "loss_total_v6": 0.43957965075969696, + "step": 2493 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5751927927927928, + "grad_norm": 25.840322494506836, + "learning_rate": 4.145892180481983e-06, + "loss": 6.850373268127441, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2981586828827858, + "loss_total_v6": 0.4281483143568039, + "step": 2494 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5754234234234235, + "grad_norm": 26.966398239135742, + "learning_rate": 4.1421503848416385e-06, + "loss": 6.829459190368652, + "loss_alignment": 0.253662109375, + "loss_alignment_w": 0.1268310546875, + "loss_sft": 0.30020851269364357, + "loss_total_v6": 0.42684121057391167, + "step": 2495 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.575654054054054, + "grad_norm": 27.550914764404297, + "learning_rate": 4.138409084138185e-06, + "loss": 6.903441429138184, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.3014754932373762, + "loss_total_v6": 0.43146511539816856, + "step": 2496 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5758846846846847, + "grad_norm": 24.822267532348633, + "learning_rate": 4.134668280530166e-06, + "loss": 6.764767646789551, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.2895277254283428, + "loss_total_v6": 0.4227979704737663, + "step": 2497 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5761153153153153, + "grad_norm": 24.396299362182617, + "learning_rate": 4.130927976175842e-06, + "loss": 7.075875759124756, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.30991804599761963, + "loss_total_v6": 0.44224225357174873, + "step": 2498 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5763459459459459, + "grad_norm": 24.381778717041016, + "learning_rate": 4.127188173233185e-06, + "loss": 6.924419403076172, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.30196258798241615, + "loss_total_v6": 0.43277620151638985, + "step": 2499 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5765765765765766, + "grad_norm": 22.65204620361328, + "learning_rate": 4.123448873859875e-06, + "loss": 7.171811103820801, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.31631068512797356, + "loss_total_v6": 0.44823817908763885, + "step": 2500 + }, + { + "epoch": 0.5765765765765766, + "eval_loss": 0.5793817043304443, + "eval_loss_alignment": 0.23473619434931506, + "eval_loss_alignment_w": 0.11736809717465753, + "eval_loss_sft": 0.46201360258847884, + "eval_loss_total_v6": 0.5793817001628822, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 92.2111, + "eval_samples_per_second": 18.924, + "eval_steps_per_second": 2.375, + "step": 2500 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5768072072072072, + "grad_norm": 24.933141708374023, + "learning_rate": 4.119710080213306e-06, + "loss": 6.755455017089844, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.29303504526615143, + "loss_total_v6": 0.42221594229340553, + "step": 2501 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5770378378378378, + "grad_norm": 24.118099212646484, + "learning_rate": 4.1159717944505724e-06, + "loss": 7.116981506347656, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3139977306127548, + "loss_total_v6": 0.4448113329708576, + "step": 2502 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5772684684684685, + "grad_norm": 28.05778694152832, + "learning_rate": 4.1122340187284845e-06, + "loss": 7.279869079589844, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.32472755014896393, + "loss_total_v6": 0.45499182865023613, + "step": 2503 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5774990990990991, + "grad_norm": 25.049713134765625, + "learning_rate": 4.108496755203553e-06, + "loss": 6.395083427429199, + "loss_alignment": 0.2642822265625, + "loss_alignment_w": 0.13214111328125, + "loss_sft": 0.26753636077046394, + "loss_total_v6": 0.39969272911548615, + "step": 2504 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5777297297297297, + "grad_norm": 23.581756591796875, + "learning_rate": 4.104760006031995e-06, + "loss": 6.641650676727295, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.2882110960781574, + "loss_total_v6": 0.41510315984487534, + "step": 2505 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5779603603603604, + "grad_norm": 26.234018325805664, + "learning_rate": 4.101023773369732e-06, + "loss": 6.431942462921143, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.2747686207294464, + "loss_total_v6": 0.4019963964819908, + "step": 2506 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.578190990990991, + "grad_norm": 24.15885353088379, + "learning_rate": 4.097288059372383e-06, + "loss": 6.901241779327393, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.30121590569615364, + "loss_total_v6": 0.43132760003209114, + "step": 2507 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5784216216216216, + "grad_norm": 25.955408096313477, + "learning_rate": 4.0935528661952716e-06, + "loss": 7.08454704284668, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.3144272714853287, + "loss_total_v6": 0.4427842013537884, + "step": 2508 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5786522522522523, + "grad_norm": 25.057933807373047, + "learning_rate": 4.08981819599342e-06, + "loss": 6.733030319213867, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2889784500002861, + "loss_total_v6": 0.4208143949508667, + "step": 2509 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5788828828828829, + "grad_norm": 27.441097259521484, + "learning_rate": 4.08608405092155e-06, + "loss": 7.217021942138672, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.3223559930920601, + "loss_total_v6": 0.451063871383667, + "step": 2510 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5791135135135135, + "grad_norm": 25.309158325195312, + "learning_rate": 4.082350433134076e-06, + "loss": 6.434200286865234, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.271323936060071, + "loss_total_v6": 0.40213751792907715, + "step": 2511 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5793441441441441, + "grad_norm": 25.766738891601562, + "learning_rate": 4.078617344785113e-06, + "loss": 6.556785583496094, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.27866506204009056, + "loss_total_v6": 0.40979909896850586, + "step": 2512 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5795747747747748, + "grad_norm": 21.231904983520508, + "learning_rate": 4.074884788028469e-06, + "loss": 7.285765647888184, + "loss_alignment": 0.273681640625, + "loss_alignment_w": 0.1368408203125, + "loss_sft": 0.31851955130696297, + "loss_total_v6": 0.45536036789417267, + "step": 2513 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5798054054054054, + "grad_norm": 22.79694366455078, + "learning_rate": 4.0711527650176434e-06, + "loss": 6.891252517700195, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.2991419956088066, + "loss_total_v6": 0.4307032711803913, + "step": 2514 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.580036036036036, + "grad_norm": 25.629240036010742, + "learning_rate": 4.067421277905833e-06, + "loss": 6.876645088195801, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.29880886897444725, + "loss_total_v6": 0.42979032173752785, + "step": 2515 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5802666666666667, + "grad_norm": 25.36403465270996, + "learning_rate": 4.063690328845917e-06, + "loss": 6.721185207366943, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.2906337436288595, + "loss_total_v6": 0.42007406428456306, + "step": 2516 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5804972972972973, + "grad_norm": 26.05082130432129, + "learning_rate": 4.059959919990471e-06, + "loss": 6.712538242340088, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.2906884402036667, + "loss_total_v6": 0.4195336550474167, + "step": 2517 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5807279279279279, + "grad_norm": 26.44779396057129, + "learning_rate": 4.056230053491759e-06, + "loss": 7.236556529998779, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.3208608105778694, + "loss_total_v6": 0.4522847719490528, + "step": 2518 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5809585585585586, + "grad_norm": 23.096651077270508, + "learning_rate": 4.052500731501726e-06, + "loss": 7.284550666809082, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.32473018020391464, + "loss_total_v6": 0.45528439059853554, + "step": 2519 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5811891891891892, + "grad_norm": 25.33966827392578, + "learning_rate": 4.04877195617201e-06, + "loss": 6.789990425109863, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.29531557485461235, + "loss_total_v6": 0.42437441274523735, + "step": 2520 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5814198198198198, + "grad_norm": 26.64027214050293, + "learning_rate": 4.045043729653927e-06, + "loss": 6.6716156005859375, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.28594872914254665, + "loss_total_v6": 0.4169759601354599, + "step": 2521 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5816504504504505, + "grad_norm": 24.24052619934082, + "learning_rate": 4.041316054098485e-06, + "loss": 6.787069320678711, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2943547908216715, + "loss_total_v6": 0.42419181764125824, + "step": 2522 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.581881081081081, + "grad_norm": 23.420495986938477, + "learning_rate": 4.037588931656362e-06, + "loss": 7.095454692840576, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3159329853951931, + "loss_total_v6": 0.4434659257531166, + "step": 2523 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5821117117117117, + "grad_norm": 24.993240356445312, + "learning_rate": 4.033862364477927e-06, + "loss": 7.334836959838867, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.3242262788116932, + "loss_total_v6": 0.4584273062646389, + "step": 2524 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5823423423423424, + "grad_norm": 26.604373931884766, + "learning_rate": 4.030136354713227e-06, + "loss": 6.936994552612305, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.30499159544706345, + "loss_total_v6": 0.43356217071413994, + "step": 2525 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.582572972972973, + "grad_norm": 24.554563522338867, + "learning_rate": 4.026410904511984e-06, + "loss": 6.751431465148926, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.29104405641555786, + "loss_total_v6": 0.42196448147296906, + "step": 2526 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5828036036036036, + "grad_norm": 23.806015014648438, + "learning_rate": 4.0226860160236e-06, + "loss": 7.020913124084473, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.30851224437355995, + "loss_total_v6": 0.43880704045295715, + "step": 2527 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5830342342342343, + "grad_norm": 24.08253288269043, + "learning_rate": 4.018961691397148e-06, + "loss": 6.71716833114624, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.2896655425429344, + "loss_total_v6": 0.41982302069664, + "step": 2528 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5832648648648648, + "grad_norm": 28.35443878173828, + "learning_rate": 4.015237932781385e-06, + "loss": 6.551481246948242, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.27949321269989014, + "loss_total_v6": 0.40946755930781364, + "step": 2529 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5834954954954955, + "grad_norm": 25.9288330078125, + "learning_rate": 4.011514742324732e-06, + "loss": 6.855151176452637, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3004867359995842, + "loss_total_v6": 0.4284469448029995, + "step": 2530 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5837261261261262, + "grad_norm": 25.99205780029297, + "learning_rate": 4.007792122175287e-06, + "loss": 6.528388023376465, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2785229030996561, + "loss_total_v6": 0.40802424773573875, + "step": 2531 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5839567567567567, + "grad_norm": 25.396400451660156, + "learning_rate": 4.004070074480821e-06, + "loss": 6.822688102722168, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.2975880317389965, + "loss_total_v6": 0.4264179915189743, + "step": 2532 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5841873873873874, + "grad_norm": 23.669601440429688, + "learning_rate": 4.000348601388767e-06, + "loss": 6.880311965942383, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3014947324991226, + "loss_total_v6": 0.4300195015966892, + "step": 2533 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5844180180180181, + "grad_norm": 24.802400588989258, + "learning_rate": 3.9966277050462345e-06, + "loss": 6.538105010986328, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.28129695914685726, + "loss_total_v6": 0.4086315557360649, + "step": 2534 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5846486486486486, + "grad_norm": 26.14944076538086, + "learning_rate": 3.992907387599996e-06, + "loss": 6.526741027832031, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.27755019068717957, + "loss_total_v6": 0.40792129561305046, + "step": 2535 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5848792792792793, + "grad_norm": 22.44434928894043, + "learning_rate": 3.989187651196493e-06, + "loss": 6.558257102966309, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.2824191488325596, + "loss_total_v6": 0.4098910763859749, + "step": 2536 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.58510990990991, + "grad_norm": 26.293794631958008, + "learning_rate": 3.985468497981826e-06, + "loss": 6.789632797241211, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.29675803519785404, + "loss_total_v6": 0.4243520312011242, + "step": 2537 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5853405405405405, + "grad_norm": 26.161136627197266, + "learning_rate": 3.981749930101764e-06, + "loss": 6.8791022300720215, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.29896245151758194, + "loss_total_v6": 0.42994390055537224, + "step": 2538 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5855711711711712, + "grad_norm": 27.699548721313477, + "learning_rate": 3.978031949701742e-06, + "loss": 7.069883346557617, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31125245802104473, + "loss_total_v6": 0.4418677017092705, + "step": 2539 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5858018018018017, + "grad_norm": 24.4456729888916, + "learning_rate": 3.974314558926844e-06, + "loss": 6.603283882141113, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2832649387419224, + "loss_total_v6": 0.4127052500844002, + "step": 2540 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5860324324324324, + "grad_norm": 26.273523330688477, + "learning_rate": 3.970597759921826e-06, + "loss": 7.029336929321289, + "loss_alignment": 0.270751953125, + "loss_alignment_w": 0.1353759765625, + "loss_sft": 0.3043238129466772, + "loss_total_v6": 0.43933356180787086, + "step": 2541 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5862630630630631, + "grad_norm": 25.62277603149414, + "learning_rate": 3.966881554831094e-06, + "loss": 6.597269058227539, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.2850557491183281, + "loss_total_v6": 0.4123293049633503, + "step": 2542 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5864936936936936, + "grad_norm": 25.201717376708984, + "learning_rate": 3.963165945798718e-06, + "loss": 6.973967552185059, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.30847733467817307, + "loss_total_v6": 0.43587296828627586, + "step": 2543 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5867243243243243, + "grad_norm": 24.340221405029297, + "learning_rate": 3.959450934968422e-06, + "loss": 6.705530643463135, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.2883430980145931, + "loss_total_v6": 0.41909565031528473, + "step": 2544 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.586954954954955, + "grad_norm": 26.739959716796875, + "learning_rate": 3.955736524483579e-06, + "loss": 7.024506568908691, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.3089657500386238, + "loss_total_v6": 0.4390316605567932, + "step": 2545 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5871855855855855, + "grad_norm": 25.64743995666504, + "learning_rate": 3.952022716487225e-06, + "loss": 6.973285675048828, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.30515407770872116, + "loss_total_v6": 0.43583035096526146, + "step": 2546 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5874162162162162, + "grad_norm": 24.52659797668457, + "learning_rate": 3.9483095131220406e-06, + "loss": 6.868075370788574, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.30195063538849354, + "loss_total_v6": 0.4292547255754471, + "step": 2547 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5876468468468469, + "grad_norm": 23.80904197692871, + "learning_rate": 3.944596916530365e-06, + "loss": 7.21992301940918, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.32000434398651123, + "loss_total_v6": 0.45124517008662224, + "step": 2548 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5878774774774774, + "grad_norm": 24.7640380859375, + "learning_rate": 3.940884928854179e-06, + "loss": 7.0823073387146, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.31375325471162796, + "loss_total_v6": 0.44264423102140427, + "step": 2549 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5881081081081081, + "grad_norm": 24.204227447509766, + "learning_rate": 3.937173552235117e-06, + "loss": 7.2015275955200195, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.32015163637697697, + "loss_total_v6": 0.4500954784452915, + "step": 2550 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5883387387387388, + "grad_norm": 25.06484603881836, + "learning_rate": 3.933462788814463e-06, + "loss": 7.26796293258667, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.32552454620599747, + "loss_total_v6": 0.45424768701195717, + "step": 2551 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5885693693693693, + "grad_norm": 23.050559997558594, + "learning_rate": 3.929752640733141e-06, + "loss": 6.717263221740723, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.29196030274033546, + "loss_total_v6": 0.41982895508408546, + "step": 2552 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5888, + "grad_norm": 22.578031539916992, + "learning_rate": 3.926043110131727e-06, + "loss": 7.252984046936035, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3231845572590828, + "loss_total_v6": 0.4533115215599537, + "step": 2553 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5890306306306307, + "grad_norm": 27.060209274291992, + "learning_rate": 3.922334199150433e-06, + "loss": 7.147688865661621, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3155202381312847, + "loss_total_v6": 0.4467305690050125, + "step": 2554 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5892612612612612, + "grad_norm": 24.12108612060547, + "learning_rate": 3.9186259099291205e-06, + "loss": 7.564822196960449, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.34096546471118927, + "loss_total_v6": 0.47280141338706017, + "step": 2555 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5894918918918919, + "grad_norm": 28.858631134033203, + "learning_rate": 3.914918244607287e-06, + "loss": 6.971351623535156, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3038125354796648, + "loss_total_v6": 0.43570949882268906, + "step": 2556 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5897225225225226, + "grad_norm": 24.410991668701172, + "learning_rate": 3.911211205324074e-06, + "loss": 7.366904258728027, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.3296636901795864, + "loss_total_v6": 0.4604315087199211, + "step": 2557 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5899531531531531, + "grad_norm": 27.89724349975586, + "learning_rate": 3.907504794218262e-06, + "loss": 6.892665863037109, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.2991387788206339, + "loss_total_v6": 0.43079160526394844, + "step": 2558 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5901837837837838, + "grad_norm": 21.93444061279297, + "learning_rate": 3.903799013428264e-06, + "loss": 6.851343154907227, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.29574737697839737, + "loss_total_v6": 0.42820892855525017, + "step": 2559 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5904144144144144, + "grad_norm": 26.232528686523438, + "learning_rate": 3.900093865092134e-06, + "loss": 7.148978233337402, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.31528033688664436, + "loss_total_v6": 0.44681110978126526, + "step": 2560 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.590645045045045, + "grad_norm": 27.202560424804688, + "learning_rate": 3.896389351347559e-06, + "loss": 6.992485046386719, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3067660313099623, + "loss_total_v6": 0.4370303191244602, + "step": 2561 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5908756756756757, + "grad_norm": 25.33956527709961, + "learning_rate": 3.892685474331863e-06, + "loss": 7.124557971954346, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.3131590262055397, + "loss_total_v6": 0.4452848695218563, + "step": 2562 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5911063063063063, + "grad_norm": 27.85540771484375, + "learning_rate": 3.888982236181997e-06, + "loss": 7.136578559875488, + "loss_alignment": 0.2696533203125, + "loss_alignment_w": 0.13482666015625, + "loss_sft": 0.3109501376748085, + "loss_total_v6": 0.4460361823439598, + "step": 2563 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5913369369369369, + "grad_norm": 25.29630470275879, + "learning_rate": 3.8852796390345465e-06, + "loss": 6.741320610046387, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.289313493296504, + "loss_total_v6": 0.421332523226738, + "step": 2564 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5915675675675676, + "grad_norm": 23.069795608520508, + "learning_rate": 3.8815776850257325e-06, + "loss": 7.102181911468506, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.31232511810958385, + "loss_total_v6": 0.4438863880932331, + "step": 2565 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5917981981981982, + "grad_norm": 25.435163497924805, + "learning_rate": 3.877876376291393e-06, + "loss": 7.145201683044434, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3186606839299202, + "loss_total_v6": 0.4465751051902771, + "step": 2566 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5920288288288288, + "grad_norm": 26.181964874267578, + "learning_rate": 3.874175714967007e-06, + "loss": 6.815096855163574, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2961370199918747, + "loss_total_v6": 0.4259435310959816, + "step": 2567 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5922594594594595, + "grad_norm": 23.477083206176758, + "learning_rate": 3.870475703187667e-06, + "loss": 7.022554397583008, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.309270977973938, + "loss_total_v6": 0.438909649848938, + "step": 2568 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5924900900900901, + "grad_norm": 24.14260482788086, + "learning_rate": 3.866776343088103e-06, + "loss": 7.128724098205566, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3156777136027813, + "loss_total_v6": 0.4455452598631382, + "step": 2569 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5927207207207207, + "grad_norm": 23.43929672241211, + "learning_rate": 3.863077636802658e-06, + "loss": 6.900890350341797, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3032844327390194, + "loss_total_v6": 0.431305643171072, + "step": 2570 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5929513513513514, + "grad_norm": 25.02945899963379, + "learning_rate": 3.859379586465306e-06, + "loss": 6.877718448638916, + "loss_alignment": 0.2672119140625, + "loss_alignment_w": 0.13360595703125, + "loss_sft": 0.2964498233050108, + "loss_total_v6": 0.42985741421580315, + "step": 2571 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.593181981981982, + "grad_norm": 24.620885848999023, + "learning_rate": 3.85568219420964e-06, + "loss": 7.309996604919434, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3250083290040493, + "loss_total_v6": 0.456874780356884, + "step": 2572 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5934126126126126, + "grad_norm": 23.38779640197754, + "learning_rate": 3.85198546216887e-06, + "loss": 6.939840316772461, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.3037198707461357, + "loss_total_v6": 0.4337400197982788, + "step": 2573 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5936432432432432, + "grad_norm": 27.409931182861328, + "learning_rate": 3.848289392475832e-06, + "loss": 6.860170841217041, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.29492586478590965, + "loss_total_v6": 0.42876069247722626, + "step": 2574 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5938738738738739, + "grad_norm": 26.88656234741211, + "learning_rate": 3.844593987262973e-06, + "loss": 7.06854248046875, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3088188208639622, + "loss_total_v6": 0.44178391247987747, + "step": 2575 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5941045045045045, + "grad_norm": 24.86758804321289, + "learning_rate": 3.840899248662358e-06, + "loss": 7.127292633056641, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.31389451399445534, + "loss_total_v6": 0.44545579701662064, + "step": 2576 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5943351351351351, + "grad_norm": 24.137067794799805, + "learning_rate": 3.837205178805674e-06, + "loss": 7.541621208190918, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.3408429026603699, + "loss_total_v6": 0.47135133668780327, + "step": 2577 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5945657657657658, + "grad_norm": 23.727420806884766, + "learning_rate": 3.833511779824212e-06, + "loss": 7.096368789672852, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3125568628311157, + "loss_total_v6": 0.44352302700281143, + "step": 2578 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5947963963963964, + "grad_norm": 24.002748489379883, + "learning_rate": 3.829819053848884e-06, + "loss": 6.677351951599121, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.28957269340753555, + "loss_total_v6": 0.41733453050255775, + "step": 2579 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.595027027027027, + "grad_norm": 24.15886688232422, + "learning_rate": 3.826127003010208e-06, + "loss": 7.301805019378662, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3260069973766804, + "loss_total_v6": 0.4563628248870373, + "step": 2580 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5952576576576577, + "grad_norm": 23.98444938659668, + "learning_rate": 3.822435629438317e-06, + "loss": 6.956418991088867, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.30124654807150364, + "loss_total_v6": 0.4347761981189251, + "step": 2581 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5954882882882883, + "grad_norm": 24.76620101928711, + "learning_rate": 3.818744935262951e-06, + "loss": 7.50185489654541, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.3377014063298702, + "loss_total_v6": 0.46886593848466873, + "step": 2582 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5957189189189189, + "grad_norm": 25.69097137451172, + "learning_rate": 3.8150549226134575e-06, + "loss": 6.732691764831543, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2909867335110903, + "loss_total_v6": 0.42079323902726173, + "step": 2583 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5959495495495496, + "grad_norm": 25.49319076538086, + "learning_rate": 3.811365593618795e-06, + "loss": 6.7689666748046875, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.29157544299960136, + "loss_total_v6": 0.42306042462587357, + "step": 2584 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5961801801801802, + "grad_norm": 21.65517807006836, + "learning_rate": 3.807676950407519e-06, + "loss": 7.050539016723633, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.3111115675419569, + "loss_total_v6": 0.44065869599580765, + "step": 2585 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5964108108108108, + "grad_norm": 23.467647552490234, + "learning_rate": 3.8039889951077984e-06, + "loss": 6.831912040710449, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.297859376296401, + "loss_total_v6": 0.4269945099949837, + "step": 2586 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5966414414414415, + "grad_norm": 22.743268966674805, + "learning_rate": 3.8003017298474e-06, + "loss": 7.006215572357178, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.3095467872917652, + "loss_total_v6": 0.4378884732723236, + "step": 2587 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.596872072072072, + "grad_norm": 27.601253509521484, + "learning_rate": 3.7966151567536956e-06, + "loss": 7.195247650146484, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.3187367878854275, + "loss_total_v6": 0.4497029706835747, + "step": 2588 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5971027027027027, + "grad_norm": 25.624271392822266, + "learning_rate": 3.7929292779536515e-06, + "loss": 7.218871116638184, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3236770071089268, + "loss_total_v6": 0.4511794447898865, + "step": 2589 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5973333333333334, + "grad_norm": 24.02661895751953, + "learning_rate": 3.789244095573842e-06, + "loss": 6.915070056915283, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.302232775837183, + "loss_total_v6": 0.4321918897330761, + "step": 2590 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5975639639639639, + "grad_norm": 26.744001388549805, + "learning_rate": 3.785559611740436e-06, + "loss": 7.169554710388184, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.3156813997775316, + "loss_total_v6": 0.4480971731245518, + "step": 2591 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5977945945945946, + "grad_norm": 25.80350112915039, + "learning_rate": 3.7818758285791958e-06, + "loss": 6.913821220397949, + "loss_alignment": 0.27099609375, + "loss_alignment_w": 0.135498046875, + "loss_sft": 0.2964479327201843, + "loss_total_v6": 0.43211381882429123, + "step": 2592 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5980252252252253, + "grad_norm": 30.6024227142334, + "learning_rate": 3.778192748215484e-06, + "loss": 6.780852794647217, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.29222677648067474, + "loss_total_v6": 0.42380330711603165, + "step": 2593 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5982558558558558, + "grad_norm": 23.717985153198242, + "learning_rate": 3.7745103727742554e-06, + "loss": 6.606919288635254, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.28439241275191307, + "loss_total_v6": 0.41293245181441307, + "step": 2594 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5984864864864865, + "grad_norm": 25.668649673461914, + "learning_rate": 3.770828704380062e-06, + "loss": 7.189828872680664, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.3189474157989025, + "loss_total_v6": 0.44936428591609, + "step": 2595 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5987171171171172, + "grad_norm": 23.85968017578125, + "learning_rate": 3.7671477451570393e-06, + "loss": 6.953528881072998, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3028969541192055, + "loss_total_v6": 0.434595562517643, + "step": 2596 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5989477477477477, + "grad_norm": 22.682918548583984, + "learning_rate": 3.7634674972289227e-06, + "loss": 7.4141364097595215, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.329533401876688, + "loss_total_v6": 0.4633835144340992, + "step": 2597 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5991783783783784, + "grad_norm": 28.49032211303711, + "learning_rate": 3.7597879627190337e-06, + "loss": 6.734280586242676, + "loss_alignment": 0.2685546875, + "loss_alignment_w": 0.13427734375, + "loss_sft": 0.28685931116342545, + "loss_total_v6": 0.42089251056313515, + "step": 2598 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5994090090090091, + "grad_norm": 27.46858024597168, + "learning_rate": 3.7561091437502808e-06, + "loss": 6.81239128112793, + "loss_alignment": 0.252197265625, + "loss_alignment_w": 0.1260986328125, + "loss_sft": 0.2996910735964775, + "loss_total_v6": 0.4257744438946247, + "step": 2599 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5996396396396396, + "grad_norm": 24.24721908569336, + "learning_rate": 3.7524310424451637e-06, + "loss": 7.0350847244262695, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.30927592143416405, + "loss_total_v6": 0.43969278782606125, + "step": 2600 + }, + { + "epoch": 0.5996396396396396, + "eval_loss": 0.5781945586204529, + "eval_loss_alignment": 0.23413643336187215, + "eval_loss_alignment_w": 0.11706821668093607, + "eval_loss_sft": 0.46112631860116843, + "eval_loss_total_v6": 0.5781945352821045, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 92.1562, + "eval_samples_per_second": 18.935, + "eval_steps_per_second": 2.376, + "step": 2600 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.5998702702702703, + "grad_norm": 29.397541046142578, + "learning_rate": 3.748753660925761e-06, + "loss": 6.695189476013184, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.2889632359147072, + "loss_total_v6": 0.4184493198990822, + "step": 2601 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6001009009009008, + "grad_norm": 28.914390563964844, + "learning_rate": 3.7450770013137417e-06, + "loss": 6.596539497375488, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2827671319246292, + "loss_total_v6": 0.4122837148606777, + "step": 2602 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6003315315315315, + "grad_norm": 30.340356826782227, + "learning_rate": 3.74140106573036e-06, + "loss": 6.65403938293457, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.28637612238526344, + "loss_total_v6": 0.41587746515870094, + "step": 2603 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6005621621621622, + "grad_norm": 26.023056030273438, + "learning_rate": 3.737725856296446e-06, + "loss": 6.9387311935424805, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.30171268060803413, + "loss_total_v6": 0.43367068469524384, + "step": 2604 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6007927927927927, + "grad_norm": 26.279211044311523, + "learning_rate": 3.734051375132417e-06, + "loss": 7.37476921081543, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.32997215166687965, + "loss_total_v6": 0.46092309057712555, + "step": 2605 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6010234234234234, + "grad_norm": 24.835634231567383, + "learning_rate": 3.7303776243582636e-06, + "loss": 6.5049262046813965, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.2749813701957464, + "loss_total_v6": 0.4065578915178776, + "step": 2606 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6012540540540541, + "grad_norm": 23.921485900878906, + "learning_rate": 3.726704606093561e-06, + "loss": 7.15805721282959, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.3175262585282326, + "loss_total_v6": 0.44737856090068817, + "step": 2607 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6014846846846846, + "grad_norm": 24.542438507080078, + "learning_rate": 3.7230323224574583e-06, + "loss": 6.6123046875, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.28509524650871754, + "loss_total_v6": 0.4132690653204918, + "step": 2608 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6017153153153153, + "grad_norm": 21.950716018676758, + "learning_rate": 3.7193607755686836e-06, + "loss": 7.09348726272583, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3129108101129532, + "loss_total_v6": 0.4433429501950741, + "step": 2609 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.601945945945946, + "grad_norm": 26.706586837768555, + "learning_rate": 3.7156899675455383e-06, + "loss": 7.029363632202148, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.3096660487353802, + "loss_total_v6": 0.4393352307379246, + "step": 2610 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6021765765765765, + "grad_norm": 25.84774398803711, + "learning_rate": 3.7120199005058955e-06, + "loss": 7.032537937164307, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.31158868223428726, + "loss_total_v6": 0.43953362479805946, + "step": 2611 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6024072072072072, + "grad_norm": 25.789419174194336, + "learning_rate": 3.7083505765672044e-06, + "loss": 6.419793128967285, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2724376246333122, + "loss_total_v6": 0.40123704448342323, + "step": 2612 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6026378378378379, + "grad_norm": 25.303464889526367, + "learning_rate": 3.704681997846482e-06, + "loss": 6.81716775894165, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.29626647755503654, + "loss_total_v6": 0.42607298865914345, + "step": 2613 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6028684684684684, + "grad_norm": 27.626955032348633, + "learning_rate": 3.7010141664603206e-06, + "loss": 7.042318344116211, + "loss_alignment": 0.2674560546875, + "loss_alignment_w": 0.13372802734375, + "loss_sft": 0.3064016103744507, + "loss_total_v6": 0.4401448965072632, + "step": 2614 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6030990990990991, + "grad_norm": 27.022563934326172, + "learning_rate": 3.6973470845248727e-06, + "loss": 7.108127593994141, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3145582489669323, + "loss_total_v6": 0.4442579671740532, + "step": 2615 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6033297297297298, + "grad_norm": 27.01211929321289, + "learning_rate": 3.6936807541558674e-06, + "loss": 7.035548210144043, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.3120209760963917, + "loss_total_v6": 0.4397217743098736, + "step": 2616 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6035603603603603, + "grad_norm": 23.500600814819336, + "learning_rate": 3.690015177468597e-06, + "loss": 7.0806097984313965, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.3135402947664261, + "loss_total_v6": 0.442538108676672, + "step": 2617 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.603790990990991, + "grad_norm": 26.635995864868164, + "learning_rate": 3.6863503565779156e-06, + "loss": 7.061102867126465, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.30934568122029305, + "loss_total_v6": 0.44131895154714584, + "step": 2618 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6040216216216217, + "grad_norm": 20.995237350463867, + "learning_rate": 3.682686293598246e-06, + "loss": 6.926119804382324, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.30209940299391747, + "loss_total_v6": 0.43288248777389526, + "step": 2619 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6042522522522522, + "grad_norm": 27.32746124267578, + "learning_rate": 3.6790229906435706e-06, + "loss": 6.688530921936035, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.28860814683139324, + "loss_total_v6": 0.4180331975221634, + "step": 2620 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6044828828828829, + "grad_norm": 22.6960391998291, + "learning_rate": 3.675360449827435e-06, + "loss": 6.781686782836914, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.2978483568876982, + "loss_total_v6": 0.4238554276525974, + "step": 2621 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6047135135135135, + "grad_norm": 25.28272247314453, + "learning_rate": 3.6716986732629445e-06, + "loss": 6.660974502563477, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2852989211678505, + "loss_total_v6": 0.4163108989596367, + "step": 2622 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6049441441441441, + "grad_norm": 25.720848083496094, + "learning_rate": 3.668037663062762e-06, + "loss": 7.102004051208496, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.311428964138031, + "loss_total_v6": 0.4438752718269825, + "step": 2623 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6051747747747748, + "grad_norm": 25.524147033691406, + "learning_rate": 3.6643774213391115e-06, + "loss": 7.03705358505249, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.30746110156178474, + "loss_total_v6": 0.43981584906578064, + "step": 2624 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6054054054054054, + "grad_norm": 26.908655166625977, + "learning_rate": 3.6607179502037708e-06, + "loss": 7.547576904296875, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3400707058608532, + "loss_total_v6": 0.4717235416173935, + "step": 2625 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.605636036036036, + "grad_norm": 28.55823516845703, + "learning_rate": 3.657059251768077e-06, + "loss": 6.557774543762207, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2783759292215109, + "loss_total_v6": 0.40986092388629913, + "step": 2626 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6058666666666667, + "grad_norm": 24.023029327392578, + "learning_rate": 3.6534013281429137e-06, + "loss": 6.631692409515381, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2843385636806488, + "loss_total_v6": 0.4144807793200016, + "step": 2627 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6060972972972973, + "grad_norm": 26.473264694213867, + "learning_rate": 3.6497441814387246e-06, + "loss": 6.613204002380371, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.2832898572087288, + "loss_total_v6": 0.4133252538740635, + "step": 2628 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6063279279279279, + "grad_norm": 23.919002532958984, + "learning_rate": 3.646087813765504e-06, + "loss": 6.662437438964844, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.28362036496400833, + "loss_total_v6": 0.41640233993530273, + "step": 2629 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6065585585585586, + "grad_norm": 23.381467819213867, + "learning_rate": 3.642432227232793e-06, + "loss": 7.452388763427734, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.3321988359093666, + "loss_total_v6": 0.4657742790877819, + "step": 2630 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6067891891891892, + "grad_norm": 27.122684478759766, + "learning_rate": 3.6387774239496893e-06, + "loss": 6.941104888916016, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.303386939689517, + "loss_total_v6": 0.43381906673312187, + "step": 2631 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6070198198198198, + "grad_norm": 27.798568725585938, + "learning_rate": 3.6351234060248288e-06, + "loss": 7.079723358154297, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3101279754191637, + "loss_total_v6": 0.44248269870877266, + "step": 2632 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6072504504504505, + "grad_norm": 23.30961799621582, + "learning_rate": 3.631470175566402e-06, + "loss": 6.675086975097656, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.288897018879652, + "loss_total_v6": 0.417192917317152, + "step": 2633 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6074810810810811, + "grad_norm": 25.639249801635742, + "learning_rate": 3.627817734682141e-06, + "loss": 6.866770267486572, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.30095355212688446, + "loss_total_v6": 0.42917314171791077, + "step": 2634 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6077117117117117, + "grad_norm": 24.91571044921875, + "learning_rate": 3.624166085479325e-06, + "loss": 7.300708770751953, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.32557223178446293, + "loss_total_v6": 0.4562942795455456, + "step": 2635 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6079423423423423, + "grad_norm": 26.938438415527344, + "learning_rate": 3.6205152300647784e-06, + "loss": 7.043242454528809, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30876342952251434, + "loss_total_v6": 0.44020263850688934, + "step": 2636 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.608172972972973, + "grad_norm": 25.62368392944336, + "learning_rate": 3.6168651705448575e-06, + "loss": 7.404203414916992, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3308657221496105, + "loss_total_v6": 0.4627627022564411, + "step": 2637 + }, + { + "action_cond_aligned": 0.96875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6084036036036036, + "grad_norm": 23.97366714477539, + "learning_rate": 3.613215909025475e-06, + "loss": 6.903541564941406, + "loss_alignment": 0.253662109375, + "loss_alignment_w": 0.1268310546875, + "loss_sft": 0.3046402856707573, + "loss_total_v6": 0.4314713552594185, + "step": 2638 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6086342342342342, + "grad_norm": 23.557844161987305, + "learning_rate": 3.6095674476120667e-06, + "loss": 7.381253242492676, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.32840902358293533, + "loss_total_v6": 0.46132833138108253, + "step": 2639 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6088648648648649, + "grad_norm": 23.601858139038086, + "learning_rate": 3.6059197884096207e-06, + "loss": 7.382547855377197, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.32905450463294983, + "loss_total_v6": 0.4614092484116554, + "step": 2640 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6090954954954955, + "grad_norm": 26.412179946899414, + "learning_rate": 3.6022729335226525e-06, + "loss": 6.567419052124023, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.27792586758732796, + "loss_total_v6": 0.41046370193362236, + "step": 2641 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6093261261261261, + "grad_norm": 24.41840362548828, + "learning_rate": 3.598626885055219e-06, + "loss": 7.097872734069824, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.3142225109040737, + "loss_total_v6": 0.4436170347034931, + "step": 2642 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6095567567567568, + "grad_norm": 24.5627498626709, + "learning_rate": 3.594981645110912e-06, + "loss": 7.052816390991211, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3076681159436703, + "loss_total_v6": 0.4408010430634022, + "step": 2643 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6097873873873874, + "grad_norm": 24.290721893310547, + "learning_rate": 3.5913372157928515e-06, + "loss": 7.3672285079956055, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.32904307916760445, + "loss_total_v6": 0.46045179292559624, + "step": 2644 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.610018018018018, + "grad_norm": 27.85612678527832, + "learning_rate": 3.5876935992036963e-06, + "loss": 6.915818214416504, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.3033018596470356, + "loss_total_v6": 0.4322386346757412, + "step": 2645 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6102486486486487, + "grad_norm": 24.94359588623047, + "learning_rate": 3.5840507974456306e-06, + "loss": 7.250086784362793, + "loss_alignment": 0.2677001953125, + "loss_alignment_w": 0.13385009765625, + "loss_sft": 0.31976860016584396, + "loss_total_v6": 0.45313040912151337, + "step": 2646 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6104792792792793, + "grad_norm": 25.662519454956055, + "learning_rate": 3.5804088126203752e-06, + "loss": 6.61851167678833, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.2873752433806658, + "loss_total_v6": 0.41365698352456093, + "step": 2647 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6107099099099099, + "grad_norm": 27.89476203918457, + "learning_rate": 3.576767646829171e-06, + "loss": 6.439874649047852, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.27210579439997673, + "loss_total_v6": 0.4024921618402004, + "step": 2648 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6109405405405406, + "grad_norm": 23.785634994506836, + "learning_rate": 3.573127302172793e-06, + "loss": 6.817686080932617, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.2942999377846718, + "loss_total_v6": 0.4261053651571274, + "step": 2649 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6111711711711711, + "grad_norm": 25.913127899169922, + "learning_rate": 3.5694877807515414e-06, + "loss": 7.272269248962402, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.32477132603526115, + "loss_total_v6": 0.45451681315898895, + "step": 2650 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6114018018018018, + "grad_norm": 27.533611297607422, + "learning_rate": 3.565849084665238e-06, + "loss": 6.60354471206665, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.2826403733342886, + "loss_total_v6": 0.41272154822945595, + "step": 2651 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6116324324324325, + "grad_norm": 24.17313575744629, + "learning_rate": 3.562211216013235e-06, + "loss": 6.759076118469238, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.2931545786559582, + "loss_total_v6": 0.4224422797560692, + "step": 2652 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.611863063063063, + "grad_norm": 25.2744197845459, + "learning_rate": 3.5585741768943982e-06, + "loss": 6.522708892822266, + "loss_alignment": 0.253662109375, + "loss_alignment_w": 0.1268310546875, + "loss_sft": 0.2809603288769722, + "loss_total_v6": 0.407669298350811, + "step": 2653 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6120936936936937, + "grad_norm": 23.6917724609375, + "learning_rate": 3.5549379694071235e-06, + "loss": 7.022754669189453, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3098022975027561, + "loss_total_v6": 0.4389221854507923, + "step": 2654 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6123243243243244, + "grad_norm": 24.320798873901367, + "learning_rate": 3.55130259564932e-06, + "loss": 6.61339807510376, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.28107419796288013, + "loss_total_v6": 0.4133373759686947, + "step": 2655 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6125549549549549, + "grad_norm": 28.98994255065918, + "learning_rate": 3.5476680577184207e-06, + "loss": 7.276267051696777, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.324578694999218, + "loss_total_v6": 0.4547666907310486, + "step": 2656 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6127855855855856, + "grad_norm": 24.1190128326416, + "learning_rate": 3.5440343577113767e-06, + "loss": 7.045636177062988, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3124988153576851, + "loss_total_v6": 0.4403522312641144, + "step": 2657 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6130162162162163, + "grad_norm": 26.638124465942383, + "learning_rate": 3.5404014977246492e-06, + "loss": 7.258739471435547, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.3251006416976452, + "loss_total_v6": 0.4536712057888508, + "step": 2658 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6132468468468468, + "grad_norm": 23.01515769958496, + "learning_rate": 3.536769479854222e-06, + "loss": 7.059300422668457, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3097060061991215, + "loss_total_v6": 0.44120626151561737, + "step": 2659 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6134774774774775, + "grad_norm": 26.033964157104492, + "learning_rate": 3.5331383061955883e-06, + "loss": 7.300865173339844, + "loss_alignment": 0.26806640625, + "loss_alignment_w": 0.134033203125, + "loss_sft": 0.32233189418911934, + "loss_total_v6": 0.45630406215786934, + "step": 2660 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6137081081081082, + "grad_norm": 25.44736099243164, + "learning_rate": 3.529507978843757e-06, + "loss": 6.774211406707764, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.2923457380384207, + "loss_total_v6": 0.42338821664452553, + "step": 2661 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6139387387387387, + "grad_norm": 25.774181365966797, + "learning_rate": 3.5258784998932505e-06, + "loss": 6.817873954772949, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.29841630905866623, + "loss_total_v6": 0.42611709982156754, + "step": 2662 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6141693693693694, + "grad_norm": 22.552616119384766, + "learning_rate": 3.522249871438093e-06, + "loss": 6.949546813964844, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.3039145227521658, + "loss_total_v6": 0.43434665352106094, + "step": 2663 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6144, + "grad_norm": 22.865427017211914, + "learning_rate": 3.518622095571831e-06, + "loss": 6.928436756134033, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.30348018929362297, + "loss_total_v6": 0.4330272972583771, + "step": 2664 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6146306306306306, + "grad_norm": 24.149892807006836, + "learning_rate": 3.5149951743875064e-06, + "loss": 6.498944282531738, + "loss_alignment": 0.252197265625, + "loss_alignment_w": 0.1260986328125, + "loss_sft": 0.28028375282883644, + "loss_total_v6": 0.40618401020765305, + "step": 2665 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6148612612612613, + "grad_norm": 22.898881912231445, + "learning_rate": 3.511369109977677e-06, + "loss": 6.913002967834473, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.30379730090498924, + "loss_total_v6": 0.43206268176436424, + "step": 2666 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6150918918918918, + "grad_norm": 24.733171463012695, + "learning_rate": 3.507743904434401e-06, + "loss": 7.3245038986206055, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.3257166538387537, + "loss_total_v6": 0.45778147131204605, + "step": 2667 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6153225225225225, + "grad_norm": 23.22998046875, + "learning_rate": 3.504119559849245e-06, + "loss": 6.753376483917236, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.29374435171484947, + "loss_total_v6": 0.42208602651953697, + "step": 2668 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6155531531531532, + "grad_norm": 24.777393341064453, + "learning_rate": 3.500496078313277e-06, + "loss": 6.501548767089844, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.2789817098528147, + "loss_total_v6": 0.40634681284427643, + "step": 2669 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6157837837837837, + "grad_norm": 23.88528060913086, + "learning_rate": 3.496873461917064e-06, + "loss": 6.659308433532715, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.28467599488794804, + "loss_total_v6": 0.4162067584693432, + "step": 2670 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6160144144144144, + "grad_norm": 24.76435661315918, + "learning_rate": 3.4932517127506802e-06, + "loss": 6.8148603439331055, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2949931062757969, + "loss_total_v6": 0.4259287603199482, + "step": 2671 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6162450450450451, + "grad_norm": 28.5228328704834, + "learning_rate": 3.4896308329036944e-06, + "loss": 6.855415344238281, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.29705478623509407, + "loss_total_v6": 0.42846347764134407, + "step": 2672 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6164756756756756, + "grad_norm": 24.5661678314209, + "learning_rate": 3.486010824465177e-06, + "loss": 7.10023307800293, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.3113945536315441, + "loss_total_v6": 0.4437645524740219, + "step": 2673 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6167063063063063, + "grad_norm": 22.571624755859375, + "learning_rate": 3.482391689523692e-06, + "loss": 6.993438720703125, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3060017004609108, + "loss_total_v6": 0.4370899498462677, + "step": 2674 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.616936936936937, + "grad_norm": 28.529848098754883, + "learning_rate": 3.478773430167302e-06, + "loss": 6.971080303192139, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.30594702810049057, + "loss_total_v6": 0.43569251149892807, + "step": 2675 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6171675675675675, + "grad_norm": 25.999141693115234, + "learning_rate": 3.475156048483567e-06, + "loss": 6.885982513427734, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2982938326895237, + "loss_total_v6": 0.4303739108145237, + "step": 2676 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6173981981981982, + "grad_norm": 27.51850700378418, + "learning_rate": 3.4715395465595337e-06, + "loss": 7.042250156402588, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.30909814313054085, + "loss_total_v6": 0.44014063104987144, + "step": 2677 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6176288288288289, + "grad_norm": 23.173404693603516, + "learning_rate": 3.46792392648175e-06, + "loss": 6.973527908325195, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.3068172000348568, + "loss_total_v6": 0.4358455166220665, + "step": 2678 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6178594594594594, + "grad_norm": 24.18910789489746, + "learning_rate": 3.4643091903362447e-06, + "loss": 6.900008201599121, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2997807767242193, + "loss_total_v6": 0.43125050514936447, + "step": 2679 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6180900900900901, + "grad_norm": 26.22867774963379, + "learning_rate": 3.4606953402085465e-06, + "loss": 6.647910118103027, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.2893347255885601, + "loss_total_v6": 0.4154943637549877, + "step": 2680 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6183207207207208, + "grad_norm": 23.363927841186523, + "learning_rate": 3.457082378183665e-06, + "loss": 7.0067243576049805, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.306694682687521, + "loss_total_v6": 0.437920268625021, + "step": 2681 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6185513513513513, + "grad_norm": 25.440710067749023, + "learning_rate": 3.4534703063461033e-06, + "loss": 7.234961032867432, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3211120739579201, + "loss_total_v6": 0.4521850720047951, + "step": 2682 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.618781981981982, + "grad_norm": 21.590778350830078, + "learning_rate": 3.449859126779851e-06, + "loss": 7.759370803833008, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3522549644112587, + "loss_total_v6": 0.4849606566131115, + "step": 2683 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6190126126126126, + "grad_norm": 29.361839294433594, + "learning_rate": 3.446248841568375e-06, + "loss": 6.7293548583984375, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.29166318103671074, + "loss_total_v6": 0.42058468982577324, + "step": 2684 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6192432432432432, + "grad_norm": 26.16796875, + "learning_rate": 3.442639452794635e-06, + "loss": 6.500438690185547, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.27630302496254444, + "loss_total_v6": 0.4062773995101452, + "step": 2685 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6194738738738739, + "grad_norm": 25.8586368560791, + "learning_rate": 3.439030962541069e-06, + "loss": 7.08535099029541, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.31255489960312843, + "loss_total_v6": 0.44283443689346313, + "step": 2686 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6197045045045045, + "grad_norm": 26.016525268554688, + "learning_rate": 3.435423372889598e-06, + "loss": 6.852015495300293, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.29726951755583286, + "loss_total_v6": 0.4282509684562683, + "step": 2687 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6199351351351351, + "grad_norm": 30.02544593811035, + "learning_rate": 3.431816685921625e-06, + "loss": 6.930058002471924, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3023913037031889, + "loss_total_v6": 0.43312861770391464, + "step": 2688 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6201657657657658, + "grad_norm": 25.33783531188965, + "learning_rate": 3.4282109037180276e-06, + "loss": 6.978063106536865, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3051017187535763, + "loss_total_v6": 0.4361289478838444, + "step": 2689 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6203963963963964, + "grad_norm": 26.938379287719727, + "learning_rate": 3.4246060283591674e-06, + "loss": 7.134767532348633, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.31710823625326157, + "loss_total_v6": 0.44592295587062836, + "step": 2690 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.620627027027027, + "grad_norm": 26.060571670532227, + "learning_rate": 3.4210020619248762e-06, + "loss": 6.332061290740967, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2660693619400263, + "loss_total_v6": 0.39575381577014923, + "step": 2691 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6208576576576577, + "grad_norm": 24.470579147338867, + "learning_rate": 3.417399006494466e-06, + "loss": 6.75101375579834, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.2891563680022955, + "loss_total_v6": 0.42193834856152534, + "step": 2692 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6210882882882883, + "grad_norm": 23.921010971069336, + "learning_rate": 3.413796864146722e-06, + "loss": 6.728936672210693, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.29255253821611404, + "loss_total_v6": 0.42055852711200714, + "step": 2693 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6213189189189189, + "grad_norm": 24.899721145629883, + "learning_rate": 3.4101956369599033e-06, + "loss": 6.867003440856934, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.29845043644309044, + "loss_total_v6": 0.42918773740530014, + "step": 2694 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6215495495495496, + "grad_norm": 25.040786743164062, + "learning_rate": 3.406595327011741e-06, + "loss": 7.198624134063721, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.31803232058882713, + "loss_total_v6": 0.44991401582956314, + "step": 2695 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6217801801801802, + "grad_norm": 27.731327056884766, + "learning_rate": 3.402995936379433e-06, + "loss": 7.140005111694336, + "loss_alignment": 0.26708984375, + "loss_alignment_w": 0.133544921875, + "loss_sft": 0.3133004680275917, + "loss_total_v6": 0.4462502971291542, + "step": 2696 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6220108108108108, + "grad_norm": 24.57418441772461, + "learning_rate": 3.3993974671396523e-06, + "loss": 6.847855567932129, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.2954226080328226, + "loss_total_v6": 0.42799097672104836, + "step": 2697 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6222414414414414, + "grad_norm": 23.71955680847168, + "learning_rate": 3.395799921368538e-06, + "loss": 7.284064292907715, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.32416577637195587, + "loss_total_v6": 0.4552540257573128, + "step": 2698 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6224720720720721, + "grad_norm": 26.694217681884766, + "learning_rate": 3.3922033011416967e-06, + "loss": 6.990574836730957, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.30724173970520496, + "loss_total_v6": 0.4369109384715557, + "step": 2699 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6227027027027027, + "grad_norm": 25.083019256591797, + "learning_rate": 3.388607608534199e-06, + "loss": 6.903256893157959, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.30218110233545303, + "loss_total_v6": 0.43145355582237244, + "step": 2700 + }, + { + "epoch": 0.6227027027027027, + "eval_loss": 0.5753805041313171, + "eval_loss_alignment": 0.2343058825627854, + "eval_loss_alignment_w": 0.1171529412813927, + "eval_loss_sft": 0.45822757942661574, + "eval_loss_total_v6": 0.5753805206484719, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.8307, + "eval_samples_per_second": 19.002, + "eval_steps_per_second": 2.385, + "step": 2700 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6229333333333333, + "grad_norm": 27.433677673339844, + "learning_rate": 3.3850128456205835e-06, + "loss": 6.943798065185547, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.30404351465404034, + "loss_total_v6": 0.4339873604476452, + "step": 2701 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.623163963963964, + "grad_norm": 24.77096939086914, + "learning_rate": 3.3814190144748514e-06, + "loss": 7.0660600662231445, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.3093960490077734, + "loss_total_v6": 0.44162872806191444, + "step": 2702 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6233945945945946, + "grad_norm": 25.168445587158203, + "learning_rate": 3.3778261171704652e-06, + "loss": 7.148140907287598, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.31632667407393456, + "loss_total_v6": 0.44675879552960396, + "step": 2703 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6236252252252252, + "grad_norm": 24.517894744873047, + "learning_rate": 3.3742341557803504e-06, + "loss": 6.754698753356934, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.2937201391905546, + "loss_total_v6": 0.42216864973306656, + "step": 2704 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6238558558558559, + "grad_norm": 23.964479446411133, + "learning_rate": 3.370643132376889e-06, + "loss": 6.616097927093506, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.28250941447913647, + "loss_total_v6": 0.4135061278939247, + "step": 2705 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6240864864864865, + "grad_norm": 23.537824630737305, + "learning_rate": 3.3670530490319265e-06, + "loss": 6.997928619384766, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.30675528198480606, + "loss_total_v6": 0.43737052381038666, + "step": 2706 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6243171171171171, + "grad_norm": 22.976757049560547, + "learning_rate": 3.3634639078167626e-06, + "loss": 7.0132060050964355, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3080152999609709, + "loss_total_v6": 0.4383253753185272, + "step": 2707 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6245477477477478, + "grad_norm": 26.093521118164062, + "learning_rate": 3.3598757108021546e-06, + "loss": 6.973981857299805, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3040989637374878, + "loss_total_v6": 0.435873843729496, + "step": 2708 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6247783783783784, + "grad_norm": 24.68427276611328, + "learning_rate": 3.356288460058318e-06, + "loss": 7.051067352294922, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3112818952649832, + "loss_total_v6": 0.4406916908919811, + "step": 2709 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.625009009009009, + "grad_norm": 24.737041473388672, + "learning_rate": 3.3527021576549146e-06, + "loss": 6.953578948974609, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3053567111492157, + "loss_total_v6": 0.4345986545085907, + "step": 2710 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6252396396396397, + "grad_norm": 23.75922966003418, + "learning_rate": 3.3491168056610674e-06, + "loss": 6.589072227478027, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.28204100392758846, + "loss_total_v6": 0.4118170030415058, + "step": 2711 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6254702702702702, + "grad_norm": 25.31610870361328, + "learning_rate": 3.3455324061453454e-06, + "loss": 7.222161293029785, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3206325173377991, + "loss_total_v6": 0.4513850659132004, + "step": 2712 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6257009009009009, + "grad_norm": 21.657482147216797, + "learning_rate": 3.3419489611757727e-06, + "loss": 6.908400058746338, + "loss_alignment": 0.25048828125, + "loss_alignment_w": 0.125244140625, + "loss_sft": 0.3064393252134323, + "loss_total_v6": 0.431775014847517, + "step": 2713 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6259315315315316, + "grad_norm": 24.604473114013672, + "learning_rate": 3.3383664728198174e-06, + "loss": 6.97233772277832, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3070785030722618, + "loss_total_v6": 0.4357711263000965, + "step": 2714 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6261621621621621, + "grad_norm": 25.040023803710938, + "learning_rate": 3.3347849431443996e-06, + "loss": 7.022152900695801, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.31205350160598755, + "loss_total_v6": 0.43888456001877785, + "step": 2715 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6263927927927928, + "grad_norm": 33.1684684753418, + "learning_rate": 3.331204374215888e-06, + "loss": 7.03641414642334, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.31122058629989624, + "loss_total_v6": 0.43977587670087814, + "step": 2716 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6266234234234235, + "grad_norm": 26.547332763671875, + "learning_rate": 3.3276247681000894e-06, + "loss": 6.754264831542969, + "loss_alignment": 0.2520751953125, + "loss_alignment_w": 0.12603759765625, + "loss_sft": 0.29598185420036316, + "loss_total_v6": 0.42214153707027435, + "step": 2717 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.626854054054054, + "grad_norm": 23.280908584594727, + "learning_rate": 3.3240461268622627e-06, + "loss": 6.445042610168457, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.27450399100780487, + "loss_total_v6": 0.40281515568494797, + "step": 2718 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6270846846846847, + "grad_norm": 23.015670776367188, + "learning_rate": 3.320468452567106e-06, + "loss": 6.919675827026367, + "loss_alignment": 0.268310546875, + "loss_alignment_w": 0.1341552734375, + "loss_sft": 0.2983397636562586, + "loss_total_v6": 0.43247976154088974, + "step": 2719 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6273153153153154, + "grad_norm": 26.174297332763672, + "learning_rate": 3.316891747278761e-06, + "loss": 7.416377067565918, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.33396119251847267, + "loss_total_v6": 0.46352357044816017, + "step": 2720 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6275459459459459, + "grad_norm": 26.389734268188477, + "learning_rate": 3.3133160130608132e-06, + "loss": 7.555912971496582, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.3418734520673752, + "loss_total_v6": 0.4722445569932461, + "step": 2721 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6277765765765766, + "grad_norm": 23.52341079711914, + "learning_rate": 3.3097412519762797e-06, + "loss": 6.826684951782227, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.2983719129115343, + "loss_total_v6": 0.42666781321167946, + "step": 2722 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6280072072072072, + "grad_norm": 21.95928192138672, + "learning_rate": 3.306167466087625e-06, + "loss": 6.683071136474609, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2876565642654896, + "loss_total_v6": 0.4176919385790825, + "step": 2723 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6282378378378378, + "grad_norm": 24.782377243041992, + "learning_rate": 3.3025946574567442e-06, + "loss": 6.46652889251709, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.27766269259154797, + "loss_total_v6": 0.4041580408811569, + "step": 2724 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6284684684684685, + "grad_norm": 24.192951202392578, + "learning_rate": 3.2990228281449755e-06, + "loss": 6.8336076736450195, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2970650587230921, + "loss_total_v6": 0.42710046097636223, + "step": 2725 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.628699099099099, + "grad_norm": 24.716066360473633, + "learning_rate": 3.2954519802130833e-06, + "loss": 6.700754642486572, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.291447289288044, + "loss_total_v6": 0.41879716143012047, + "step": 2726 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6289297297297297, + "grad_norm": 26.43661117553711, + "learning_rate": 3.291882115721274e-06, + "loss": 7.003229141235352, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.307071341201663, + "loss_total_v6": 0.43770183622837067, + "step": 2727 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6291603603603604, + "grad_norm": 25.637773513793945, + "learning_rate": 3.2883132367291824e-06, + "loss": 7.146884441375732, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.31296752765774727, + "loss_total_v6": 0.44668029248714447, + "step": 2728 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6293909909909909, + "grad_norm": 23.110836029052734, + "learning_rate": 3.2847453452958755e-06, + "loss": 7.029993534088135, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3077980652451515, + "loss_total_v6": 0.4393745884299278, + "step": 2729 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6296216216216216, + "grad_norm": 23.791399002075195, + "learning_rate": 3.281178443479852e-06, + "loss": 6.8784332275390625, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.3024301342666149, + "loss_total_v6": 0.4299020543694496, + "step": 2730 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6298522522522523, + "grad_norm": 23.46596908569336, + "learning_rate": 3.2776125333390345e-06, + "loss": 7.130213737487793, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.31549612432718277, + "loss_total_v6": 0.44563834741711617, + "step": 2731 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6300828828828828, + "grad_norm": 25.109210968017578, + "learning_rate": 3.274047616930781e-06, + "loss": 6.478158950805664, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.27454433776438236, + "loss_total_v6": 0.4048849120736122, + "step": 2732 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6303135135135135, + "grad_norm": 24.70287322998047, + "learning_rate": 3.270483696311869e-06, + "loss": 7.286547660827637, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.323527492582798, + "loss_total_v6": 0.4554092139005661, + "step": 2733 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6305441441441442, + "grad_norm": 22.921552658081055, + "learning_rate": 3.2669207735385074e-06, + "loss": 7.049108505249023, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.31001508608460426, + "loss_total_v6": 0.44056927412748337, + "step": 2734 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6307747747747747, + "grad_norm": 25.720579147338867, + "learning_rate": 3.2633588506663273e-06, + "loss": 7.2375054359436035, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3207828290760517, + "loss_total_v6": 0.4523441009223461, + "step": 2735 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6310054054054054, + "grad_norm": 22.9486026763916, + "learning_rate": 3.2597979297503777e-06, + "loss": 7.090240955352783, + "loss_alignment": 0.25048828125, + "loss_alignment_w": 0.125244140625, + "loss_sft": 0.31768227741122246, + "loss_total_v6": 0.44314004108309746, + "step": 2736 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6312360360360361, + "grad_norm": 24.01329231262207, + "learning_rate": 3.2562380128451416e-06, + "loss": 7.043781280517578, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30908703804016113, + "loss_total_v6": 0.44023633375763893, + "step": 2737 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6314666666666666, + "grad_norm": 26.88271713256836, + "learning_rate": 3.252679102004509e-06, + "loss": 6.705783843994141, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.2880232371389866, + "loss_total_v6": 0.4191114939749241, + "step": 2738 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6316972972972973, + "grad_norm": 24.585735321044922, + "learning_rate": 3.249121199281799e-06, + "loss": 7.121574878692627, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.3178859055042267, + "loss_total_v6": 0.4450984187424183, + "step": 2739 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.631927927927928, + "grad_norm": 25.18929100036621, + "learning_rate": 3.2455643067297437e-06, + "loss": 6.924091339111328, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3036205880343914, + "loss_total_v6": 0.4327557012438774, + "step": 2740 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6321585585585585, + "grad_norm": 24.971912384033203, + "learning_rate": 3.2420084264004966e-06, + "loss": 6.804828643798828, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.29325221478939056, + "loss_total_v6": 0.42530176416039467, + "step": 2741 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6323891891891892, + "grad_norm": 26.025157928466797, + "learning_rate": 3.2384535603456255e-06, + "loss": 7.097114562988281, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.31196262687444687, + "loss_total_v6": 0.44356967881321907, + "step": 2742 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6326198198198199, + "grad_norm": 23.89362907409668, + "learning_rate": 3.2348997106161107e-06, + "loss": 6.8100714683532715, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.29559406265616417, + "loss_total_v6": 0.42562947049736977, + "step": 2743 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6328504504504504, + "grad_norm": 25.168542861938477, + "learning_rate": 3.231346879262349e-06, + "loss": 6.9234395027160645, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.3044343199580908, + "loss_total_v6": 0.43271496891975403, + "step": 2744 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6330810810810811, + "grad_norm": 25.00030517578125, + "learning_rate": 3.2277950683341487e-06, + "loss": 6.719841480255127, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.28910019248723984, + "loss_total_v6": 0.41999008879065514, + "step": 2745 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6333117117117117, + "grad_norm": 23.02676010131836, + "learning_rate": 3.224244279880728e-06, + "loss": 6.795130252838135, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.2960945777595043, + "loss_total_v6": 0.4246956445276737, + "step": 2746 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6335423423423423, + "grad_norm": 25.287513732910156, + "learning_rate": 3.2206945159507213e-06, + "loss": 6.912494659423828, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3025906328111887, + "loss_total_v6": 0.43203094974160194, + "step": 2747 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.633772972972973, + "grad_norm": 25.294164657592773, + "learning_rate": 3.217145778592162e-06, + "loss": 7.056425094604492, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.31286800652742386, + "loss_total_v6": 0.44102656468749046, + "step": 2748 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6340036036036036, + "grad_norm": 28.51250648498535, + "learning_rate": 3.213598069852499e-06, + "loss": 6.957402229309082, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3028796147555113, + "loss_total_v6": 0.43483762443065643, + "step": 2749 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6342342342342342, + "grad_norm": 25.34953498840332, + "learning_rate": 3.2100513917785837e-06, + "loss": 6.8301167488098145, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.29596190340816975, + "loss_total_v6": 0.4268823005259037, + "step": 2750 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6344648648648649, + "grad_norm": 22.601085662841797, + "learning_rate": 3.2065057464166767e-06, + "loss": 6.92225456237793, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3030022159218788, + "loss_total_v6": 0.4326408952474594, + "step": 2751 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6346954954954955, + "grad_norm": 24.477184295654297, + "learning_rate": 3.202961135812437e-06, + "loss": 7.208859443664551, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.3193128816783428, + "loss_total_v6": 0.4505537115037441, + "step": 2752 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6349261261261261, + "grad_norm": 25.707746505737305, + "learning_rate": 3.199417562010931e-06, + "loss": 6.906403541564941, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.30227097123861313, + "loss_total_v6": 0.43165023624897003, + "step": 2753 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6351567567567568, + "grad_norm": 27.742168426513672, + "learning_rate": 3.195875027056629e-06, + "loss": 6.870913028717041, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3018991071730852, + "loss_total_v6": 0.42943206802010536, + "step": 2754 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6353873873873874, + "grad_norm": 23.307479858398438, + "learning_rate": 3.1923335329933947e-06, + "loss": 7.222240447998047, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.3202865347266197, + "loss_total_v6": 0.4513900354504585, + "step": 2755 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.635618018018018, + "grad_norm": 22.31134033203125, + "learning_rate": 3.1887930818644997e-06, + "loss": 6.891315937042236, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3022435028105974, + "loss_total_v6": 0.43070724979043007, + "step": 2756 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6358486486486487, + "grad_norm": 25.49868392944336, + "learning_rate": 3.185253675712605e-06, + "loss": 7.031349182128906, + "loss_alignment": 0.267333984375, + "loss_alignment_w": 0.1336669921875, + "loss_sft": 0.305655000731349, + "loss_total_v6": 0.43945932388305664, + "step": 2757 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6360792792792793, + "grad_norm": 21.760374069213867, + "learning_rate": 3.181715316579776e-06, + "loss": 7.127300262451172, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.31696198508143425, + "loss_total_v6": 0.44545625150203705, + "step": 2758 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6363099099099099, + "grad_norm": 25.777297973632812, + "learning_rate": 3.1781780065074697e-06, + "loss": 7.139031410217285, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.313544824719429, + "loss_total_v6": 0.4461894743144512, + "step": 2759 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6365405405405405, + "grad_norm": 27.1320743560791, + "learning_rate": 3.1746417475365403e-06, + "loss": 7.040655136108398, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.3103717565536499, + "loss_total_v6": 0.4400409385561943, + "step": 2760 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6367711711711712, + "grad_norm": 25.264387130737305, + "learning_rate": 3.171106541707236e-06, + "loss": 7.129443645477295, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3161346670240164, + "loss_total_v6": 0.44559023529291153, + "step": 2761 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6370018018018018, + "grad_norm": 25.772384643554688, + "learning_rate": 3.167572391059194e-06, + "loss": 7.002206802368164, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.30806031078100204, + "loss_total_v6": 0.43763793259859085, + "step": 2762 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6372324324324324, + "grad_norm": 25.414451599121094, + "learning_rate": 3.1640392976314472e-06, + "loss": 7.150035858154297, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.31357648968696594, + "loss_total_v6": 0.44687726348638535, + "step": 2763 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6374630630630631, + "grad_norm": 24.731428146362305, + "learning_rate": 3.160507263462413e-06, + "loss": 6.838118553161621, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.2966146022081375, + "loss_total_v6": 0.4273824021220207, + "step": 2764 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6376936936936937, + "grad_norm": 26.411346435546875, + "learning_rate": 3.1569762905899027e-06, + "loss": 7.190406322479248, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3175644353032112, + "loss_total_v6": 0.4494003802537918, + "step": 2765 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6379243243243243, + "grad_norm": 27.37496566772461, + "learning_rate": 3.153446381051113e-06, + "loss": 6.90219259262085, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.30104646645486355, + "loss_total_v6": 0.4313870370388031, + "step": 2766 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.638154954954955, + "grad_norm": 25.026287078857422, + "learning_rate": 3.1499175368826267e-06, + "loss": 7.109896659851074, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.3138448968529701, + "loss_total_v6": 0.44436856731772423, + "step": 2767 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6383855855855856, + "grad_norm": 22.546634674072266, + "learning_rate": 3.1463897601204165e-06, + "loss": 7.059988021850586, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3118089288473129, + "loss_total_v6": 0.44124923273921013, + "step": 2768 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6386162162162162, + "grad_norm": 26.305482864379883, + "learning_rate": 3.1428630527998293e-06, + "loss": 7.3287506103515625, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.32732487469911575, + "loss_total_v6": 0.45804693177342415, + "step": 2769 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6388468468468469, + "grad_norm": 23.558687210083008, + "learning_rate": 3.1393374169556064e-06, + "loss": 6.657111167907715, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.28798719495534897, + "loss_total_v6": 0.41606946289539337, + "step": 2770 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6390774774774775, + "grad_norm": 21.594728469848633, + "learning_rate": 3.1358128546218623e-06, + "loss": 6.526180267333984, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.27731676772236824, + "loss_total_v6": 0.40788624435663223, + "step": 2771 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6393081081081081, + "grad_norm": 23.181079864501953, + "learning_rate": 3.132289367832097e-06, + "loss": 6.809686660766602, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.295768428593874, + "loss_total_v6": 0.4256054423749447, + "step": 2772 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6395387387387388, + "grad_norm": 26.520689010620117, + "learning_rate": 3.1287669586191906e-06, + "loss": 7.17772102355957, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31719886884093285, + "loss_total_v6": 0.44860755652189255, + "step": 2773 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6397693693693693, + "grad_norm": 25.14118194580078, + "learning_rate": 3.1252456290153952e-06, + "loss": 7.179208755493164, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.32099974527955055, + "loss_total_v6": 0.44870054349303246, + "step": 2774 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.64, + "grad_norm": 21.16287612915039, + "learning_rate": 3.121725381052347e-06, + "loss": 6.920247554779053, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.30533347837626934, + "loss_total_v6": 0.4325154721736908, + "step": 2775 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6402306306306307, + "grad_norm": 28.579795837402344, + "learning_rate": 3.1182062167610533e-06, + "loss": 7.022994041442871, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3058805000036955, + "loss_total_v6": 0.43893714994192123, + "step": 2776 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6404612612612612, + "grad_norm": 24.301315307617188, + "learning_rate": 3.114688138171902e-06, + "loss": 6.946265697479248, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3058457188308239, + "loss_total_v6": 0.434141606092453, + "step": 2777 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6406918918918919, + "grad_norm": 26.052635192871094, + "learning_rate": 3.1111711473146456e-06, + "loss": 6.778485298156738, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.2934978473931551, + "loss_total_v6": 0.42365531995892525, + "step": 2778 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6409225225225226, + "grad_norm": 23.017311096191406, + "learning_rate": 3.107655246218417e-06, + "loss": 7.280086994171143, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3272436335682869, + "loss_total_v6": 0.4550054520368576, + "step": 2779 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6411531531531531, + "grad_norm": 25.367462158203125, + "learning_rate": 3.104140436911719e-06, + "loss": 6.673398494720459, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.28793701715767384, + "loss_total_v6": 0.4170874170958996, + "step": 2780 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6413837837837838, + "grad_norm": 25.518871307373047, + "learning_rate": 3.10062672142242e-06, + "loss": 6.859073638916016, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.2988398112356663, + "loss_total_v6": 0.42869212105870247, + "step": 2781 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6416144144144145, + "grad_norm": 25.50597381591797, + "learning_rate": 3.097114101777764e-06, + "loss": 6.775156021118164, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.29559389874339104, + "loss_total_v6": 0.42344727367162704, + "step": 2782 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.641845045045045, + "grad_norm": 23.37404441833496, + "learning_rate": 3.093602580004356e-06, + "loss": 6.62727689743042, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.28233836963772774, + "loss_total_v6": 0.41420481726527214, + "step": 2783 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6420756756756757, + "grad_norm": 26.7510929107666, + "learning_rate": 3.090092158128172e-06, + "loss": 6.691160202026367, + "loss_alignment": 0.2515869140625, + "loss_alignment_w": 0.12579345703125, + "loss_sft": 0.29222096130251884, + "loss_total_v6": 0.41819752380251884, + "step": 2784 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6423063063063063, + "grad_norm": 21.46351432800293, + "learning_rate": 3.0865828381745515e-06, + "loss": 7.01124906539917, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3062908500432968, + "loss_total_v6": 0.4382030703127384, + "step": 2785 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6425369369369369, + "grad_norm": 21.68536376953125, + "learning_rate": 3.0830746221681994e-06, + "loss": 6.5588483810424805, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.27945015765726566, + "loss_total_v6": 0.4099280424416065, + "step": 2786 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6427675675675676, + "grad_norm": 25.019472122192383, + "learning_rate": 3.0795675121331845e-06, + "loss": 7.087907314300537, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.3131876662373543, + "loss_total_v6": 0.442994199693203, + "step": 2787 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6429981981981981, + "grad_norm": 24.025497436523438, + "learning_rate": 3.0760615100929348e-06, + "loss": 7.062003135681152, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3099970147013664, + "loss_total_v6": 0.4413751810789108, + "step": 2788 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6432288288288288, + "grad_norm": 25.915281295776367, + "learning_rate": 3.0725566180702415e-06, + "loss": 6.8482208251953125, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.2966356389224529, + "loss_total_v6": 0.42801379784941673, + "step": 2789 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6434594594594595, + "grad_norm": 23.3400821685791, + "learning_rate": 3.069052838087252e-06, + "loss": 6.669724464416504, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.2900267094373703, + "loss_total_v6": 0.4168577529489994, + "step": 2790 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.64369009009009, + "grad_norm": 27.67003059387207, + "learning_rate": 3.065550172165477e-06, + "loss": 6.745833396911621, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.2918996326625347, + "loss_total_v6": 0.4216145910322666, + "step": 2791 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6439207207207207, + "grad_norm": 27.589984893798828, + "learning_rate": 3.0620486223257795e-06, + "loss": 6.7128448486328125, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.2881441116333008, + "loss_total_v6": 0.4195527955889702, + "step": 2792 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6441513513513514, + "grad_norm": 25.64695930480957, + "learning_rate": 3.058548190588383e-06, + "loss": 7.18947172164917, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3181469067931175, + "loss_total_v6": 0.4493419751524925, + "step": 2793 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6443819819819819, + "grad_norm": 25.351055145263672, + "learning_rate": 3.055048878972864e-06, + "loss": 6.8903069496154785, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.303370613604784, + "loss_total_v6": 0.4306441731750965, + "step": 2794 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6446126126126126, + "grad_norm": 26.247371673583984, + "learning_rate": 3.0515506894981505e-06, + "loss": 6.778958320617676, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.2937410846352577, + "loss_total_v6": 0.4236849173903465, + "step": 2795 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6448432432432433, + "grad_norm": 26.16205596923828, + "learning_rate": 3.0480536241825263e-06, + "loss": 6.798883438110352, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2959323935210705, + "loss_total_v6": 0.4249302074313164, + "step": 2796 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6450738738738738, + "grad_norm": 25.417091369628906, + "learning_rate": 3.0445576850436253e-06, + "loss": 6.829679012298584, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.2940729930996895, + "loss_total_v6": 0.4268549531698227, + "step": 2797 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6453045045045045, + "grad_norm": 28.120267868041992, + "learning_rate": 3.0410628740984347e-06, + "loss": 6.786350250244141, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2942640706896782, + "loss_total_v6": 0.4241468831896782, + "step": 2798 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6455351351351352, + "grad_norm": 23.859479904174805, + "learning_rate": 3.037569193363282e-06, + "loss": 7.329946517944336, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3266519419848919, + "loss_total_v6": 0.4581216648221016, + "step": 2799 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6457657657657657, + "grad_norm": 26.82548713684082, + "learning_rate": 3.0340766448538535e-06, + "loss": 7.43757438659668, + "loss_alignment": 0.26904296875, + "loss_alignment_w": 0.134521484375, + "loss_sft": 0.3305557742714882, + "loss_total_v6": 0.4648483879864216, + "step": 2800 + }, + { + "epoch": 0.6457657657657657, + "eval_loss": 0.5748540163040161, + "eval_loss_alignment": 0.2341431221461187, + "eval_loss_alignment_w": 0.11707156107305935, + "eval_loss_sft": 0.4577825135598172, + "eval_loss_total_v6": 0.5748540745818451, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.5864, + "eval_samples_per_second": 19.053, + "eval_steps_per_second": 2.391, + "step": 2800 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6459963963963964, + "grad_norm": 25.56105613708496, + "learning_rate": 3.030585230585176e-06, + "loss": 6.82680606842041, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2943053599447012, + "loss_total_v6": 0.42667535692453384, + "step": 2801 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6462270270270271, + "grad_norm": 27.650278091430664, + "learning_rate": 3.0270949525716216e-06, + "loss": 6.636543273925781, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.28796815872192383, + "loss_total_v6": 0.4147839583456516, + "step": 2802 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6464576576576576, + "grad_norm": 25.815725326538086, + "learning_rate": 3.0236058128269117e-06, + "loss": 6.956839561462402, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.30542321875691414, + "loss_total_v6": 0.43480248376727104, + "step": 2803 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6466882882882883, + "grad_norm": 25.742950439453125, + "learning_rate": 3.020117813364104e-06, + "loss": 6.994595527648926, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.30732520669698715, + "loss_total_v6": 0.43716222420334816, + "step": 2804 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.646918918918919, + "grad_norm": 25.471118927001953, + "learning_rate": 3.016630956195604e-06, + "loss": 7.176718711853027, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3198217749595642, + "loss_total_v6": 0.4485449120402336, + "step": 2805 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6471495495495495, + "grad_norm": 27.178730010986328, + "learning_rate": 3.0131452433331574e-06, + "loss": 6.943717002868652, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.30551856756210327, + "loss_total_v6": 0.4339823089540005, + "step": 2806 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6473801801801802, + "grad_norm": 22.663002014160156, + "learning_rate": 3.009660676787846e-06, + "loss": 7.0786662101745605, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.31210658326745033, + "loss_total_v6": 0.44241663068532944, + "step": 2807 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6476108108108108, + "grad_norm": 25.590137481689453, + "learning_rate": 3.0061772585700955e-06, + "loss": 7.055273056030273, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3121093586087227, + "loss_total_v6": 0.4409545622766018, + "step": 2808 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6478414414414414, + "grad_norm": 26.730934143066406, + "learning_rate": 3.002694990689663e-06, + "loss": 7.12095832824707, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31468877755105495, + "loss_total_v6": 0.445059884339571, + "step": 2809 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6480720720720721, + "grad_norm": 24.931856155395508, + "learning_rate": 2.999213875155648e-06, + "loss": 6.9755024909973145, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.30436186119914055, + "loss_total_v6": 0.43596889823675156, + "step": 2810 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6483027027027027, + "grad_norm": 26.283977508544922, + "learning_rate": 2.99573391397648e-06, + "loss": 7.518999099731445, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.33820831775665283, + "loss_total_v6": 0.46993744373321533, + "step": 2811 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6485333333333333, + "grad_norm": 26.700197219848633, + "learning_rate": 2.992255109159926e-06, + "loss": 6.869965553283691, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2966213822364807, + "loss_total_v6": 0.429372850805521, + "step": 2812 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.648763963963964, + "grad_norm": 23.20920753479004, + "learning_rate": 2.988777462713086e-06, + "loss": 7.112489700317383, + "loss_alignment": 0.2637939453125, + "loss_alignment_w": 0.13189697265625, + "loss_sft": 0.31258783861994743, + "loss_total_v6": 0.44453058391809464, + "step": 2813 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6489945945945946, + "grad_norm": 27.116783142089844, + "learning_rate": 2.985300976642388e-06, + "loss": 6.715512752532959, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.28956208005547523, + "loss_total_v6": 0.41971954703330994, + "step": 2814 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6492252252252252, + "grad_norm": 26.31810188293457, + "learning_rate": 2.981825652953596e-06, + "loss": 6.993851184844971, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.3088045474141836, + "loss_total_v6": 0.43711570277810097, + "step": 2815 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6494558558558559, + "grad_norm": 24.456512451171875, + "learning_rate": 2.9783514936517964e-06, + "loss": 6.742432594299316, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.2889252249151468, + "loss_total_v6": 0.4214020296931267, + "step": 2816 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6496864864864865, + "grad_norm": 25.004920959472656, + "learning_rate": 2.9748785007414105e-06, + "loss": 6.609189033508301, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.28384762071073055, + "loss_total_v6": 0.4130743145942688, + "step": 2817 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6499171171171171, + "grad_norm": 26.875646591186523, + "learning_rate": 2.9714066762261825e-06, + "loss": 6.782750606536865, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.2937186788767576, + "loss_total_v6": 0.4239219203591347, + "step": 2818 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6501477477477478, + "grad_norm": 24.177658081054688, + "learning_rate": 2.9679360221091834e-06, + "loss": 7.126542091369629, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.3138934001326561, + "loss_total_v6": 0.4454089067876339, + "step": 2819 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6503783783783784, + "grad_norm": 28.59421730041504, + "learning_rate": 2.964466540392812e-06, + "loss": 6.970284461975098, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.3028760813176632, + "loss_total_v6": 0.4356427974998951, + "step": 2820 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.650609009009009, + "grad_norm": 24.75724220275879, + "learning_rate": 2.9609982330787846e-06, + "loss": 6.941006660461426, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3054712079465389, + "loss_total_v6": 0.4338128939270973, + "step": 2821 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6508396396396396, + "grad_norm": 27.64405059814453, + "learning_rate": 2.9575311021681454e-06, + "loss": 7.289743423461914, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.32409345731139183, + "loss_total_v6": 0.45560894906520844, + "step": 2822 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6510702702702703, + "grad_norm": 26.830575942993164, + "learning_rate": 2.954065149661256e-06, + "loss": 6.744859218597412, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2894278559833765, + "loss_total_v6": 0.42155370488762856, + "step": 2823 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6513009009009009, + "grad_norm": 24.81128692626953, + "learning_rate": 2.950600377557804e-06, + "loss": 6.77811861038208, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.2917506732046604, + "loss_total_v6": 0.4236323982477188, + "step": 2824 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6515315315315315, + "grad_norm": 25.441247940063477, + "learning_rate": 2.9471367878567864e-06, + "loss": 7.075879096984863, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.31292422488331795, + "loss_total_v6": 0.44224245846271515, + "step": 2825 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6517621621621622, + "grad_norm": 26.391748428344727, + "learning_rate": 2.943674382556526e-06, + "loss": 7.232186317443848, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.3207860551774502, + "loss_total_v6": 0.4520116411149502, + "step": 2826 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6519927927927928, + "grad_norm": 25.20650863647461, + "learning_rate": 2.9402131636546614e-06, + "loss": 6.195096015930176, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.25848560594022274, + "loss_total_v6": 0.3871934898197651, + "step": 2827 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6522234234234234, + "grad_norm": 23.867578506469727, + "learning_rate": 2.9367531331481436e-06, + "loss": 6.994041442871094, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.30619192495942116, + "loss_total_v6": 0.43712758272886276, + "step": 2828 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6524540540540541, + "grad_norm": 25.199613571166992, + "learning_rate": 2.9332942930332404e-06, + "loss": 7.063787460327148, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3088420704007149, + "loss_total_v6": 0.4414867237210274, + "step": 2829 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6526846846846847, + "grad_norm": 24.05300521850586, + "learning_rate": 2.929836645305529e-06, + "loss": 6.821029186248779, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.29357810132205486, + "loss_total_v6": 0.4263143166899681, + "step": 2830 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6529153153153153, + "grad_norm": 25.153779983520508, + "learning_rate": 2.9263801919599038e-06, + "loss": 6.8911285400390625, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.298874881118536, + "loss_total_v6": 0.4306955561041832, + "step": 2831 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.653145945945946, + "grad_norm": 26.278730392456055, + "learning_rate": 2.9229249349905686e-06, + "loss": 6.502908706665039, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.27841053158044815, + "loss_total_v6": 0.40643177554011345, + "step": 2832 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6533765765765766, + "grad_norm": 25.451038360595703, + "learning_rate": 2.919470876391034e-06, + "loss": 6.606986999511719, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.28497645258903503, + "loss_total_v6": 0.41293666884303093, + "step": 2833 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6536072072072072, + "grad_norm": 22.560251235961914, + "learning_rate": 2.916018018154123e-06, + "loss": 6.8108015060424805, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.29696720093488693, + "loss_total_v6": 0.42567509040236473, + "step": 2834 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6538378378378379, + "grad_norm": 23.239269256591797, + "learning_rate": 2.9125663622719624e-06, + "loss": 6.798772811889648, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.29423175379633904, + "loss_total_v6": 0.424923300743103, + "step": 2835 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6540684684684684, + "grad_norm": 27.42072105407715, + "learning_rate": 2.9091159107359913e-06, + "loss": 7.005547046661377, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.30884888768196106, + "loss_total_v6": 0.43784669786691666, + "step": 2836 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6542990990990991, + "grad_norm": 22.570913314819336, + "learning_rate": 2.905666665536945e-06, + "loss": 6.818379878997803, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.29344305396080017, + "loss_total_v6": 0.42614874616265297, + "step": 2837 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6545297297297298, + "grad_norm": 23.694517135620117, + "learning_rate": 2.90221862866487e-06, + "loss": 7.014381408691406, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3078751564025879, + "loss_total_v6": 0.4383988417685032, + "step": 2838 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6547603603603603, + "grad_norm": 25.11659049987793, + "learning_rate": 2.898771802109115e-06, + "loss": 6.958351135253906, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.30271004140377045, + "loss_total_v6": 0.43489693105220795, + "step": 2839 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.654990990990991, + "grad_norm": 26.358388900756836, + "learning_rate": 2.8953261878583263e-06, + "loss": 6.776167869567871, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.29127782955765724, + "loss_total_v6": 0.42351050302386284, + "step": 2840 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6552216216216217, + "grad_norm": 27.091259002685547, + "learning_rate": 2.8918817879004535e-06, + "loss": 6.64798641204834, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2851585615426302, + "loss_total_v6": 0.41549915075302124, + "step": 2841 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6554522522522522, + "grad_norm": 24.408403396606445, + "learning_rate": 2.888438604222746e-06, + "loss": 6.892646789550781, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.30197574198246, + "loss_total_v6": 0.43079042434692383, + "step": 2842 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6556828828828829, + "grad_norm": 28.063922882080078, + "learning_rate": 2.884996638811752e-06, + "loss": 6.76186466217041, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.29485467448830605, + "loss_total_v6": 0.42261652275919914, + "step": 2843 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6559135135135136, + "grad_norm": 25.51102638244629, + "learning_rate": 2.881555893653314e-06, + "loss": 6.772088050842285, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.2942272089421749, + "loss_total_v6": 0.423255518078804, + "step": 2844 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6561441441441441, + "grad_norm": 23.947118759155273, + "learning_rate": 2.878116370732573e-06, + "loss": 6.7040300369262695, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.28925638273358345, + "loss_total_v6": 0.41900185495615005, + "step": 2845 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6563747747747748, + "grad_norm": 25.50556755065918, + "learning_rate": 2.8746780720339646e-06, + "loss": 6.737109661102295, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.29007265716791153, + "loss_total_v6": 0.42106935754418373, + "step": 2846 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6566054054054054, + "grad_norm": 26.788183212280273, + "learning_rate": 2.871240999541215e-06, + "loss": 6.899221420288086, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.3024781830608845, + "loss_total_v6": 0.4312013313174248, + "step": 2847 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.656836036036036, + "grad_norm": 27.166608810424805, + "learning_rate": 2.867805155237349e-06, + "loss": 6.997331619262695, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.3074198942631483, + "loss_total_v6": 0.43733322620391846, + "step": 2848 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6570666666666667, + "grad_norm": 27.84768295288086, + "learning_rate": 2.8643705411046753e-06, + "loss": 7.103123188018799, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.31204822286963463, + "loss_total_v6": 0.4439451992511749, + "step": 2849 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6572972972972972, + "grad_norm": 26.734886169433594, + "learning_rate": 2.8609371591247982e-06, + "loss": 6.936123371124268, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30188537761569023, + "loss_total_v6": 0.43350769951939583, + "step": 2850 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6575279279279279, + "grad_norm": 26.595745086669922, + "learning_rate": 2.85750501127861e-06, + "loss": 6.630908966064453, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.2873718962073326, + "loss_total_v6": 0.4144318290054798, + "step": 2851 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6577585585585586, + "grad_norm": 25.58757209777832, + "learning_rate": 2.854074099546291e-06, + "loss": 6.909102439880371, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.30066959373652935, + "loss_total_v6": 0.4318188801407814, + "step": 2852 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6579891891891891, + "grad_norm": 27.03459930419922, + "learning_rate": 2.8506444259073095e-06, + "loss": 6.756077766418457, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.29105981811881065, + "loss_total_v6": 0.42225487530231476, + "step": 2853 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6582198198198198, + "grad_norm": 25.404460906982422, + "learning_rate": 2.847215992340414e-06, + "loss": 6.533850193023682, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.27768938802182674, + "loss_total_v6": 0.408365648239851, + "step": 2854 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6584504504504505, + "grad_norm": 27.411991119384766, + "learning_rate": 2.8437888008236457e-06, + "loss": 7.016129493713379, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.3065653573721647, + "loss_total_v6": 0.4385081082582474, + "step": 2855 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.658681081081081, + "grad_norm": 25.710834503173828, + "learning_rate": 2.8403628533343207e-06, + "loss": 6.731625556945801, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.28884488344192505, + "loss_total_v6": 0.42072660103440285, + "step": 2856 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6589117117117117, + "grad_norm": 24.492347717285156, + "learning_rate": 2.8369381518490437e-06, + "loss": 7.517444133758545, + "loss_alignment": 0.2642822265625, + "loss_alignment_w": 0.13214111328125, + "loss_sft": 0.33757708221673965, + "loss_total_v6": 0.46984026208519936, + "step": 2857 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6591423423423424, + "grad_norm": 23.192089080810547, + "learning_rate": 2.8335146983437e-06, + "loss": 7.029665946960449, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.30626697465777397, + "loss_total_v6": 0.43935413286089897, + "step": 2858 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6593729729729729, + "grad_norm": 22.47234344482422, + "learning_rate": 2.8300924947934483e-06, + "loss": 7.218686103820801, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.3218190912157297, + "loss_total_v6": 0.45116785541176796, + "step": 2859 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6596036036036036, + "grad_norm": 28.72235107421875, + "learning_rate": 2.826671543172738e-06, + "loss": 7.03769588470459, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.30725711584091187, + "loss_total_v6": 0.43985600024461746, + "step": 2860 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6598342342342343, + "grad_norm": 22.881650924682617, + "learning_rate": 2.8232518454552843e-06, + "loss": 7.326164722442627, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.3303981423377991, + "loss_total_v6": 0.4578853063285351, + "step": 2861 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6600648648648648, + "grad_norm": 25.857629776000977, + "learning_rate": 2.8198334036140873e-06, + "loss": 6.708871841430664, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2897268831729889, + "loss_total_v6": 0.4193045124411583, + "step": 2862 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6602954954954955, + "grad_norm": 25.193857192993164, + "learning_rate": 2.8164162196214163e-06, + "loss": 7.200750350952148, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.3212932087481022, + "loss_total_v6": 0.4500468783080578, + "step": 2863 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6605261261261262, + "grad_norm": 29.046037673950195, + "learning_rate": 2.8130002954488185e-06, + "loss": 6.5506205558776855, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.28185029327869415, + "loss_total_v6": 0.40941378101706505, + "step": 2864 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6607567567567567, + "grad_norm": 22.711957931518555, + "learning_rate": 2.8095856330671164e-06, + "loss": 7.351747989654541, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3298761323094368, + "loss_total_v6": 0.45948426425457, + "step": 2865 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6609873873873874, + "grad_norm": 24.640499114990234, + "learning_rate": 2.806172234446397e-06, + "loss": 6.976243019104004, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.30671221390366554, + "loss_total_v6": 0.43601519614458084, + "step": 2866 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.661218018018018, + "grad_norm": 28.616361618041992, + "learning_rate": 2.8027601015560255e-06, + "loss": 7.109832286834717, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3153667077422142, + "loss_total_v6": 0.4443645179271698, + "step": 2867 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6614486486486486, + "grad_norm": 23.273250579833984, + "learning_rate": 2.799349236364634e-06, + "loss": 7.09385871887207, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.31325447373092175, + "loss_total_v6": 0.443366177380085, + "step": 2868 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6616792792792793, + "grad_norm": 24.804662704467773, + "learning_rate": 2.795939640840125e-06, + "loss": 6.719020366668701, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.2899796664714813, + "loss_total_v6": 0.4199387691915035, + "step": 2869 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6619099099099099, + "grad_norm": 27.885147094726562, + "learning_rate": 2.7925313169496637e-06, + "loss": 7.084144592285156, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.31157922744750977, + "loss_total_v6": 0.44275902211666107, + "step": 2870 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6621405405405405, + "grad_norm": 25.029855728149414, + "learning_rate": 2.7891242666596856e-06, + "loss": 6.714783668518066, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.2886620406061411, + "loss_total_v6": 0.41967400535941124, + "step": 2871 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6623711711711712, + "grad_norm": 27.32436180114746, + "learning_rate": 2.7857184919358936e-06, + "loss": 6.652324199676514, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.28939695842564106, + "loss_total_v6": 0.4157702513039112, + "step": 2872 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6626018018018018, + "grad_norm": 25.77863121032715, + "learning_rate": 2.782313994743247e-06, + "loss": 6.722386837005615, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.28806910291314125, + "loss_total_v6": 0.42014917358756065, + "step": 2873 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6628324324324324, + "grad_norm": 21.029075622558594, + "learning_rate": 2.7789107770459767e-06, + "loss": 7.020400047302246, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.3076714910566807, + "loss_total_v6": 0.4387750066816807, + "step": 2874 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6630630630630631, + "grad_norm": 24.38153839111328, + "learning_rate": 2.7755088408075682e-06, + "loss": 7.1349992752075195, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.31372005119919777, + "loss_total_v6": 0.44593746215105057, + "step": 2875 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6632936936936937, + "grad_norm": 24.287919998168945, + "learning_rate": 2.7721081879907717e-06, + "loss": 6.862993240356445, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3008548226207495, + "loss_total_v6": 0.4289370961487293, + "step": 2876 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6635243243243243, + "grad_norm": 26.778085708618164, + "learning_rate": 2.7687088205575964e-06, + "loss": 6.147108554840088, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2526788078248501, + "loss_total_v6": 0.3841942958533764, + "step": 2877 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.663754954954955, + "grad_norm": 24.272830963134766, + "learning_rate": 2.7653107404693096e-06, + "loss": 6.871560573577881, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.29943713918328285, + "loss_total_v6": 0.42947254329919815, + "step": 2878 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6639855855855856, + "grad_norm": 25.277565002441406, + "learning_rate": 2.761913949686438e-06, + "loss": 7.054225921630859, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3083512857556343, + "loss_total_v6": 0.4408891312777996, + "step": 2879 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6642162162162162, + "grad_norm": 30.381811141967773, + "learning_rate": 2.758518450168758e-06, + "loss": 6.924821853637695, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.30004991963505745, + "loss_total_v6": 0.43280138820409775, + "step": 2880 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6644468468468469, + "grad_norm": 27.334041595458984, + "learning_rate": 2.755124243875309e-06, + "loss": 6.516301155090332, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2759974170476198, + "loss_total_v6": 0.40726879239082336, + "step": 2881 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6646774774774775, + "grad_norm": 25.433561325073242, + "learning_rate": 2.751731332764377e-06, + "loss": 7.318723678588867, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.3275221399962902, + "loss_total_v6": 0.457420215010643, + "step": 2882 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6649081081081081, + "grad_norm": 26.402244567871094, + "learning_rate": 2.7483397187935084e-06, + "loss": 6.948089122772217, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.3044337872415781, + "loss_total_v6": 0.43425557017326355, + "step": 2883 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6651387387387387, + "grad_norm": 28.971895217895508, + "learning_rate": 2.74494940391949e-06, + "loss": 7.130480766296387, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.31606216356158257, + "loss_total_v6": 0.44565505161881447, + "step": 2884 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6653693693693694, + "grad_norm": 32.35963821411133, + "learning_rate": 2.741560390098374e-06, + "loss": 6.807837009429932, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.2930893041193485, + "loss_total_v6": 0.42548980936408043, + "step": 2885 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6656, + "grad_norm": 26.88909912109375, + "learning_rate": 2.7381726792854522e-06, + "loss": 6.6624436378479, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.2858332861214876, + "loss_total_v6": 0.41640273854136467, + "step": 2886 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6658306306306306, + "grad_norm": 23.99942970275879, + "learning_rate": 2.7347862734352637e-06, + "loss": 6.594614028930664, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.2840658463537693, + "loss_total_v6": 0.4121633730828762, + "step": 2887 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6660612612612613, + "grad_norm": 24.7177791595459, + "learning_rate": 2.731401174501601e-06, + "loss": 6.696165084838867, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2887037806212902, + "loss_total_v6": 0.4185102917253971, + "step": 2888 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6662918918918919, + "grad_norm": 25.129802703857422, + "learning_rate": 2.7280173844374954e-06, + "loss": 7.0216851234436035, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3080569691956043, + "loss_total_v6": 0.4388553090393543, + "step": 2889 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6665225225225225, + "grad_norm": 24.197528839111328, + "learning_rate": 2.724634905195229e-06, + "loss": 6.5981035232543945, + "loss_alignment": 0.251708984375, + "loss_alignment_w": 0.1258544921875, + "loss_sft": 0.28616076707839966, + "loss_total_v6": 0.41238146647810936, + "step": 2890 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6667531531531532, + "grad_norm": 25.562374114990234, + "learning_rate": 2.721253738726326e-06, + "loss": 6.561611175537109, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2795770224183798, + "loss_total_v6": 0.41010070592164993, + "step": 2891 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6669837837837838, + "grad_norm": 24.45745086669922, + "learning_rate": 2.717873886981551e-06, + "loss": 6.875370025634766, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.30057550594210625, + "loss_total_v6": 0.42971064895391464, + "step": 2892 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6672144144144144, + "grad_norm": 24.46587562561035, + "learning_rate": 2.7144953519109113e-06, + "loss": 6.6245880126953125, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.28633594140410423, + "loss_total_v6": 0.41403675079345703, + "step": 2893 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6674450450450451, + "grad_norm": 24.63313865661621, + "learning_rate": 2.711118135463656e-06, + "loss": 6.384911060333252, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.2727904785424471, + "loss_total_v6": 0.39905695989727974, + "step": 2894 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6676756756756757, + "grad_norm": 26.948570251464844, + "learning_rate": 2.7077422395882745e-06, + "loss": 6.830363750457764, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.29931899532675743, + "loss_total_v6": 0.42689773440361023, + "step": 2895 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6679063063063063, + "grad_norm": 22.693056106567383, + "learning_rate": 2.7043676662324882e-06, + "loss": 7.147738456726074, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.31773586198687553, + "loss_total_v6": 0.44673366472125053, + "step": 2896 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.668136936936937, + "grad_norm": 24.37588882446289, + "learning_rate": 2.7009944173432606e-06, + "loss": 7.046377182006836, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3089746329933405, + "loss_total_v6": 0.44039858877658844, + "step": 2897 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6683675675675675, + "grad_norm": 24.990137100219727, + "learning_rate": 2.6976224948667924e-06, + "loss": 6.901678085327148, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3003276288509369, + "loss_total_v6": 0.4313548617064953, + "step": 2898 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6685981981981982, + "grad_norm": 27.045995712280273, + "learning_rate": 2.694251900748512e-06, + "loss": 7.088740348815918, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3117749337106943, + "loss_total_v6": 0.44304628297686577, + "step": 2899 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6688288288288289, + "grad_norm": 24.04540252685547, + "learning_rate": 2.69088263693309e-06, + "loss": 6.775186061859131, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.29371889866888523, + "loss_total_v6": 0.4234491176903248, + "step": 2900 + }, + { + "epoch": 0.6688288288288289, + "eval_loss": 0.5734794735908508, + "eval_loss_alignment": 0.2340427903824201, + "eval_loss_alignment_w": 0.11702139519121005, + "eval_loss_sft": 0.45645810007946946, + "eval_loss_total_v6": 0.5734794950240278, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 92.4309, + "eval_samples_per_second": 18.879, + "eval_steps_per_second": 2.369, + "step": 2900 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6690594594594594, + "grad_norm": 26.061052322387695, + "learning_rate": 2.687514705364422e-06, + "loss": 6.716819763183594, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2900862544775009, + "loss_total_v6": 0.4198012351989746, + "step": 2901 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6692900900900901, + "grad_norm": 25.85629653930664, + "learning_rate": 2.6841481079856393e-06, + "loss": 6.9720659255981445, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.30593233555555344, + "loss_total_v6": 0.43575410172343254, + "step": 2902 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6695207207207208, + "grad_norm": 23.974319458007812, + "learning_rate": 2.680782846739103e-06, + "loss": 6.290264129638672, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.26051211543381214, + "loss_total_v6": 0.393141508102417, + "step": 2903 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6697513513513513, + "grad_norm": 23.66832160949707, + "learning_rate": 2.6774189235664026e-06, + "loss": 6.945804119110107, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.3049776069819927, + "loss_total_v6": 0.4341127499938011, + "step": 2904 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.669981981981982, + "grad_norm": 24.012279510498047, + "learning_rate": 2.6740563404083588e-06, + "loss": 6.516178131103516, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2785532455891371, + "loss_total_v6": 0.40726112946867943, + "step": 2905 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6702126126126126, + "grad_norm": 25.784860610961914, + "learning_rate": 2.6706950992050097e-06, + "loss": 7.03605842590332, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.31165611743927, + "loss_total_v6": 0.4397536665201187, + "step": 2906 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6704432432432432, + "grad_norm": 26.928956985473633, + "learning_rate": 2.667335201895631e-06, + "loss": 7.354509353637695, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.32817185297608376, + "loss_total_v6": 0.45965683832764626, + "step": 2907 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6706738738738739, + "grad_norm": 24.00600242614746, + "learning_rate": 2.6639766504187147e-06, + "loss": 6.795172691345215, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.29411356523633003, + "loss_total_v6": 0.42469828203320503, + "step": 2908 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6709045045045045, + "grad_norm": 24.210979461669922, + "learning_rate": 2.6606194467119816e-06, + "loss": 7.117671012878418, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3157345838844776, + "loss_total_v6": 0.4448544420301914, + "step": 2909 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6711351351351351, + "grad_norm": 25.050159454345703, + "learning_rate": 2.6572635927123674e-06, + "loss": 6.728351593017578, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2916004993021488, + "loss_total_v6": 0.4205219969153404, + "step": 2910 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6713657657657658, + "grad_norm": 26.545021057128906, + "learning_rate": 2.6539090903560396e-06, + "loss": 6.769113540649414, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.29570447094738483, + "loss_total_v6": 0.4230695813894272, + "step": 2911 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6715963963963963, + "grad_norm": 26.22976303100586, + "learning_rate": 2.650555941578381e-06, + "loss": 6.682163715362549, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.28613498620688915, + "loss_total_v6": 0.4176352359354496, + "step": 2912 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.671827027027027, + "grad_norm": 23.55146598815918, + "learning_rate": 2.647204148313989e-06, + "loss": 6.78275728225708, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2942378856241703, + "loss_total_v6": 0.4239223375916481, + "step": 2913 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6720576576576577, + "grad_norm": 25.345338821411133, + "learning_rate": 2.6438537124966866e-06, + "loss": 6.909562587738037, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.30400952510535717, + "loss_total_v6": 0.4318476617336273, + "step": 2914 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6722882882882882, + "grad_norm": 25.623838424682617, + "learning_rate": 2.640504636059506e-06, + "loss": 7.027802467346191, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3059215862303972, + "loss_total_v6": 0.43923764303326607, + "step": 2915 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6725189189189189, + "grad_norm": 25.723831176757812, + "learning_rate": 2.6371569209347016e-06, + "loss": 7.418569087982178, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.33460172824561596, + "loss_total_v6": 0.4636605754494667, + "step": 2916 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6727495495495496, + "grad_norm": 23.34383773803711, + "learning_rate": 2.6338105690537402e-06, + "loss": 7.022249221801758, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.31115927547216415, + "loss_total_v6": 0.43889058381319046, + "step": 2917 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6729801801801801, + "grad_norm": 28.378694534301758, + "learning_rate": 2.630465582347299e-06, + "loss": 6.702165126800537, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.287507152184844, + "loss_total_v6": 0.41888532042503357, + "step": 2918 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6732108108108108, + "grad_norm": 26.031322479248047, + "learning_rate": 2.6271219627452726e-06, + "loss": 5.993798732757568, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.24375304020941257, + "loss_total_v6": 0.3746124133467674, + "step": 2919 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6734414414414415, + "grad_norm": 23.27215003967285, + "learning_rate": 2.6237797121767634e-06, + "loss": 7.161279678344727, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3199249133467674, + "loss_total_v6": 0.4475799649953842, + "step": 2920 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.673672072072072, + "grad_norm": 22.095142364501953, + "learning_rate": 2.6204388325700877e-06, + "loss": 6.9572343826293945, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.3039677646011114, + "loss_total_v6": 0.43482714518904686, + "step": 2921 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6739027027027027, + "grad_norm": 23.255502700805664, + "learning_rate": 2.6170993258527634e-06, + "loss": 6.6335673332214355, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.282731506973505, + "loss_total_v6": 0.4145979695022106, + "step": 2922 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6741333333333334, + "grad_norm": 24.308311462402344, + "learning_rate": 2.613761193951524e-06, + "loss": 6.86292028427124, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.3001635707914829, + "loss_total_v6": 0.4289325065910816, + "step": 2923 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6743639639639639, + "grad_norm": 26.273998260498047, + "learning_rate": 2.6104244387923082e-06, + "loss": 6.809408664703369, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.29840604588389397, + "loss_total_v6": 0.42558804526925087, + "step": 2924 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6745945945945946, + "grad_norm": 24.730093002319336, + "learning_rate": 2.6070890623002557e-06, + "loss": 6.753332614898682, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.2930397242307663, + "loss_total_v6": 0.4220832958817482, + "step": 2925 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6748252252252253, + "grad_norm": 25.618959426879883, + "learning_rate": 2.603755066399718e-06, + "loss": 6.8777594566345215, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.3034103624522686, + "loss_total_v6": 0.4298599623143673, + "step": 2926 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6750558558558558, + "grad_norm": 26.584003448486328, + "learning_rate": 2.600422453014241e-06, + "loss": 7.427940368652344, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.334882240742445, + "loss_total_v6": 0.4642462506890297, + "step": 2927 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6752864864864865, + "grad_norm": 23.55197525024414, + "learning_rate": 2.5970912240665815e-06, + "loss": 6.810601711273193, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.2970920465886593, + "loss_total_v6": 0.4256626032292843, + "step": 2928 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6755171171171171, + "grad_norm": 26.324039459228516, + "learning_rate": 2.5937613814786944e-06, + "loss": 6.237514019012451, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.26081632822752, + "loss_total_v6": 0.3898446336388588, + "step": 2929 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6757477477477477, + "grad_norm": 23.88416862487793, + "learning_rate": 2.5904329271717333e-06, + "loss": 7.583212852478027, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3444799818098545, + "loss_total_v6": 0.4739508032798767, + "step": 2930 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6759783783783784, + "grad_norm": 25.405025482177734, + "learning_rate": 2.587105863066054e-06, + "loss": 6.589839458465576, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.28349278680980206, + "loss_total_v6": 0.4118649661540985, + "step": 2931 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.676209009009009, + "grad_norm": 23.297216415405273, + "learning_rate": 2.5837801910812054e-06, + "loss": 7.145540237426758, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.31579791754484177, + "loss_total_v6": 0.44659626856446266, + "step": 2932 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6764396396396396, + "grad_norm": 25.751375198364258, + "learning_rate": 2.58045591313594e-06, + "loss": 7.227694511413574, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.3192845806479454, + "loss_total_v6": 0.4517308808863163, + "step": 2933 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6766702702702703, + "grad_norm": 27.33783531188965, + "learning_rate": 2.577133031148198e-06, + "loss": 7.2471394538879395, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.32141545601189137, + "loss_total_v6": 0.4529462158679962, + "step": 2934 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6769009009009009, + "grad_norm": 24.890701293945312, + "learning_rate": 2.57381154703512e-06, + "loss": 6.621481895446777, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.2873625047504902, + "loss_total_v6": 0.4138426035642624, + "step": 2935 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6771315315315315, + "grad_norm": 26.56466293334961, + "learning_rate": 2.5704914627130374e-06, + "loss": 6.989654064178467, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.3067569360136986, + "loss_total_v6": 0.4368533827364445, + "step": 2936 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6773621621621622, + "grad_norm": 33.589210510253906, + "learning_rate": 2.567172780097477e-06, + "loss": 6.8073930740356445, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.29376348853111267, + "loss_total_v6": 0.4254620857536793, + "step": 2937 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6775927927927928, + "grad_norm": 26.842166900634766, + "learning_rate": 2.563855501103156e-06, + "loss": 6.935689926147461, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.3054135926067829, + "loss_total_v6": 0.4334806203842163, + "step": 2938 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6778234234234234, + "grad_norm": 26.263185501098633, + "learning_rate": 2.5605396276439764e-06, + "loss": 6.915298938751221, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.30348303727805614, + "loss_total_v6": 0.4322061873972416, + "step": 2939 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6780540540540541, + "grad_norm": 24.15688133239746, + "learning_rate": 2.5572251616330372e-06, + "loss": 6.719276428222656, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.2893242947757244, + "loss_total_v6": 0.4199547953903675, + "step": 2940 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6782846846846847, + "grad_norm": 24.7607479095459, + "learning_rate": 2.553912104982619e-06, + "loss": 6.749846935272217, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.2918452899903059, + "loss_total_v6": 0.42186543345451355, + "step": 2941 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6785153153153153, + "grad_norm": 24.12826156616211, + "learning_rate": 2.550600459604195e-06, + "loss": 6.781276702880859, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.2922990210354328, + "loss_total_v6": 0.423829797655344, + "step": 2942 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.678745945945946, + "grad_norm": 24.73476791381836, + "learning_rate": 2.5472902274084172e-06, + "loss": 7.044389724731445, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.31040678918361664, + "loss_total_v6": 0.44027434661984444, + "step": 2943 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6789765765765766, + "grad_norm": 26.087438583374023, + "learning_rate": 2.5439814103051286e-06, + "loss": 6.1641035079956055, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.25708267092704773, + "loss_total_v6": 0.38525649532675743, + "step": 2944 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6792072072072072, + "grad_norm": 23.99854850769043, + "learning_rate": 2.540674010203353e-06, + "loss": 6.753686428070068, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.2918563783168793, + "loss_total_v6": 0.42210540920495987, + "step": 2945 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6794378378378378, + "grad_norm": 24.031848907470703, + "learning_rate": 2.5373680290112973e-06, + "loss": 6.519442558288574, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.27927607111632824, + "loss_total_v6": 0.4074651561677456, + "step": 2946 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6796684684684685, + "grad_norm": 27.33869171142578, + "learning_rate": 2.5340634686363513e-06, + "loss": 6.8650054931640625, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.29736422561109066, + "loss_total_v6": 0.4290628246963024, + "step": 2947 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6798990990990991, + "grad_norm": 27.416763305664062, + "learning_rate": 2.5307603309850793e-06, + "loss": 7.161395072937012, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.3171703312546015, + "loss_total_v6": 0.44758719205856323, + "step": 2948 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6801297297297297, + "grad_norm": 29.429548263549805, + "learning_rate": 2.527458617963231e-06, + "loss": 6.294859886169434, + "loss_alignment": 0.24609375, + "loss_alignment_w": 0.123046875, + "loss_sft": 0.27044286020100117, + "loss_total_v6": 0.3934287317097187, + "step": 2949 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6803603603603604, + "grad_norm": 26.453798294067383, + "learning_rate": 2.5241583314757327e-06, + "loss": 7.236208915710449, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3230516277253628, + "loss_total_v6": 0.45226307958364487, + "step": 2950 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.680590990990991, + "grad_norm": 26.194316864013672, + "learning_rate": 2.5208594734266845e-06, + "loss": 7.221663475036621, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3194570019841194, + "loss_total_v6": 0.4513539709150791, + "step": 2951 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6808216216216216, + "grad_norm": 27.64337158203125, + "learning_rate": 2.5175620457193674e-06, + "loss": 6.766855239868164, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.2926946934312582, + "loss_total_v6": 0.42292844876646996, + "step": 2952 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6810522522522523, + "grad_norm": 26.294044494628906, + "learning_rate": 2.514266050256231e-06, + "loss": 6.466203689575195, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2742244079709053, + "loss_total_v6": 0.4041377417743206, + "step": 2953 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6812828828828829, + "grad_norm": 25.997066497802734, + "learning_rate": 2.5109714889389048e-06, + "loss": 6.704220771789551, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2890851963311434, + "loss_total_v6": 0.41901378706097603, + "step": 2954 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6815135135135135, + "grad_norm": 24.54662322998047, + "learning_rate": 2.5076783636681863e-06, + "loss": 7.057555675506592, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.3115806430578232, + "loss_total_v6": 0.4410972446203232, + "step": 2955 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6817441441441442, + "grad_norm": 27.470212936401367, + "learning_rate": 2.504386676344047e-06, + "loss": 6.586113929748535, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.282725865021348, + "loss_total_v6": 0.41163211688399315, + "step": 2956 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6819747747747748, + "grad_norm": 29.253610610961914, + "learning_rate": 2.5010964288656302e-06, + "loss": 6.476247787475586, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.275752417743206, + "loss_total_v6": 0.4047654904425144, + "step": 2957 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6822054054054054, + "grad_norm": 26.25323486328125, + "learning_rate": 2.497807623131243e-06, + "loss": 6.74760627746582, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.29402459785342216, + "loss_total_v6": 0.42172540351748466, + "step": 2958 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6824360360360361, + "grad_norm": 23.401649475097656, + "learning_rate": 2.4945202610383673e-06, + "loss": 7.52449893951416, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3397116959095001, + "loss_total_v6": 0.4702811576426029, + "step": 2959 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6826666666666666, + "grad_norm": 27.881603240966797, + "learning_rate": 2.4912343444836444e-06, + "loss": 6.770493507385254, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2937613148242235, + "loss_total_v6": 0.4231558330357075, + "step": 2960 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6828972972972973, + "grad_norm": 28.58349609375, + "learning_rate": 2.4879498753628885e-06, + "loss": 6.766191482543945, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.29420959018170834, + "loss_total_v6": 0.4228869490325451, + "step": 2961 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.683127927927928, + "grad_norm": 26.7301082611084, + "learning_rate": 2.484666855571076e-06, + "loss": 6.7195820808410645, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.2942262031137943, + "loss_total_v6": 0.4199738837778568, + "step": 2962 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6833585585585585, + "grad_norm": 25.4509334564209, + "learning_rate": 2.4813852870023473e-06, + "loss": 6.826753616333008, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2967282496392727, + "loss_total_v6": 0.426672101020813, + "step": 2963 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6835891891891892, + "grad_norm": 25.637296676635742, + "learning_rate": 2.4781051715500076e-06, + "loss": 7.157795429229736, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3143971599638462, + "loss_total_v6": 0.4473622255027294, + "step": 2964 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6838198198198199, + "grad_norm": 25.285045623779297, + "learning_rate": 2.474826511106517e-06, + "loss": 6.6538920402526855, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.28645844385027885, + "loss_total_v6": 0.41586825251579285, + "step": 2965 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6840504504504504, + "grad_norm": 25.259904861450195, + "learning_rate": 2.4715493075635062e-06, + "loss": 6.988094329833984, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.30395863950252533, + "loss_total_v6": 0.4367558993399143, + "step": 2966 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6842810810810811, + "grad_norm": 26.22556495666504, + "learning_rate": 2.468273562811755e-06, + "loss": 6.788181304931641, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.2958281058818102, + "loss_total_v6": 0.42426132038235664, + "step": 2967 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6845117117117117, + "grad_norm": 23.77396583557129, + "learning_rate": 2.4649992787412104e-06, + "loss": 6.771395683288574, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.29343623109161854, + "loss_total_v6": 0.4232122302055359, + "step": 2968 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6847423423423423, + "grad_norm": 24.93951988220215, + "learning_rate": 2.4617264572409684e-06, + "loss": 7.064362525939941, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3080540560185909, + "loss_total_v6": 0.44152267277240753, + "step": 2969 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.684972972972973, + "grad_norm": 25.463024139404297, + "learning_rate": 2.458455100199288e-06, + "loss": 7.130613327026367, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.31622302904725075, + "loss_total_v6": 0.44566333666443825, + "step": 2970 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6852036036036036, + "grad_norm": 23.385452270507812, + "learning_rate": 2.4551852095035812e-06, + "loss": 7.02726936340332, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.3094283305108547, + "loss_total_v6": 0.4392043463885784, + "step": 2971 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6854342342342342, + "grad_norm": 23.20110511779785, + "learning_rate": 2.4519167870404126e-06, + "loss": 7.19688081741333, + "loss_alignment": 0.249267578125, + "loss_alignment_w": 0.1246337890625, + "loss_sft": 0.32535434886813164, + "loss_total_v6": 0.44980503618717194, + "step": 2972 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6856648648648649, + "grad_norm": 24.837060928344727, + "learning_rate": 2.448649834695503e-06, + "loss": 6.922957420349121, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.30451101809740067, + "loss_total_v6": 0.43268483504652977, + "step": 2973 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6858954954954954, + "grad_norm": 23.016475677490234, + "learning_rate": 2.4453843543537185e-06, + "loss": 6.945523262023926, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3062418270856142, + "loss_total_v6": 0.43409522622823715, + "step": 2974 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6861261261261261, + "grad_norm": 24.01790428161621, + "learning_rate": 2.442120347899083e-06, + "loss": 7.253729820251465, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.32538259960711, + "loss_total_v6": 0.45335808768868446, + "step": 2975 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6863567567567568, + "grad_norm": 23.676706314086914, + "learning_rate": 2.4388578172147676e-06, + "loss": 6.751718044281006, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.2951818127185106, + "loss_total_v6": 0.4219823628664017, + "step": 2976 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6865873873873873, + "grad_norm": 22.799480438232422, + "learning_rate": 2.4355967641830893e-06, + "loss": 7.260673522949219, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.3239550441503525, + "loss_total_v6": 0.4537920765578747, + "step": 2977 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.686818018018018, + "grad_norm": 24.587629318237305, + "learning_rate": 2.432337190685517e-06, + "loss": 7.635886192321777, + "loss_alignment": 0.266357421875, + "loss_alignment_w": 0.1331787109375, + "loss_sft": 0.3437742367386818, + "loss_total_v6": 0.4772428907454014, + "step": 2978 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6870486486486487, + "grad_norm": 26.786537170410156, + "learning_rate": 2.429079098602661e-06, + "loss": 6.906580448150635, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.3043419420719147, + "loss_total_v6": 0.4316612742841244, + "step": 2979 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6872792792792792, + "grad_norm": 27.215106964111328, + "learning_rate": 2.4258224898142806e-06, + "loss": 6.965368270874023, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3057578504085541, + "loss_total_v6": 0.435335498303175, + "step": 2980 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6875099099099099, + "grad_norm": 24.67622184753418, + "learning_rate": 2.4225673661992792e-06, + "loss": 6.96675443649292, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.306683748960495, + "loss_total_v6": 0.4354221485555172, + "step": 2981 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6877405405405406, + "grad_norm": 27.375457763671875, + "learning_rate": 2.419313729635703e-06, + "loss": 6.869070053100586, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.2992509640753269, + "loss_total_v6": 0.4293168634176254, + "step": 2982 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6879711711711711, + "grad_norm": 25.85500144958496, + "learning_rate": 2.41606158200074e-06, + "loss": 6.689743995666504, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.288561899214983, + "loss_total_v6": 0.4181090146303177, + "step": 2983 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6882018018018018, + "grad_norm": 27.674728393554688, + "learning_rate": 2.4128109251707154e-06, + "loss": 6.998058319091797, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.3057715743780136, + "loss_total_v6": 0.4373786225914955, + "step": 2984 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6884324324324325, + "grad_norm": 26.236207962036133, + "learning_rate": 2.4095617610211033e-06, + "loss": 6.7188191413879395, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.287983451038599, + "loss_total_v6": 0.4199262037873268, + "step": 2985 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.688663063063063, + "grad_norm": 24.233701705932617, + "learning_rate": 2.406314091426505e-06, + "loss": 6.794600009918213, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.29474917240440845, + "loss_total_v6": 0.424662496894598, + "step": 2986 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6888936936936937, + "grad_norm": 24.306692123413086, + "learning_rate": 2.4030679182606687e-06, + "loss": 7.131869316101074, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.31788842752575874, + "loss_total_v6": 0.44574181735515594, + "step": 2987 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6891243243243244, + "grad_norm": 26.227563858032227, + "learning_rate": 2.3998232433964762e-06, + "loss": 7.1462202072143555, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3178088106215, + "loss_total_v6": 0.4466387517750263, + "step": 2988 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6893549549549549, + "grad_norm": 26.974117279052734, + "learning_rate": 2.396580068705944e-06, + "loss": 6.9537553787231445, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.30725985020399094, + "loss_total_v6": 0.43460970744490623, + "step": 2989 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6895855855855856, + "grad_norm": 22.228744506835938, + "learning_rate": 2.393338396060226e-06, + "loss": 7.1240997314453125, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.31136035546660423, + "loss_total_v6": 0.44525622576475143, + "step": 2990 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6898162162162162, + "grad_norm": 25.43008804321289, + "learning_rate": 2.390098227329605e-06, + "loss": 7.186068058013916, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.32019246742129326, + "loss_total_v6": 0.44912924244999886, + "step": 2991 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6900468468468468, + "grad_norm": 27.129737854003906, + "learning_rate": 2.3868595643834995e-06, + "loss": 7.027813911437988, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.31124766916036606, + "loss_total_v6": 0.43923838809132576, + "step": 2992 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6902774774774775, + "grad_norm": 26.329240798950195, + "learning_rate": 2.3836224090904566e-06, + "loss": 7.23500919342041, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.32433467358350754, + "loss_total_v6": 0.45218808203935623, + "step": 2993 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6905081081081081, + "grad_norm": 23.182748794555664, + "learning_rate": 2.3803867633181575e-06, + "loss": 6.787217140197754, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2940283231437206, + "loss_total_v6": 0.42420104891061783, + "step": 2994 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6907387387387387, + "grad_norm": 24.360801696777344, + "learning_rate": 2.377152628933408e-06, + "loss": 7.060722351074219, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.3095354996621609, + "loss_total_v6": 0.44129515811800957, + "step": 2995 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6909693693693694, + "grad_norm": 24.419273376464844, + "learning_rate": 2.373920007802144e-06, + "loss": 6.650455474853516, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.28509924933314323, + "loss_total_v6": 0.41565345227718353, + "step": 2996 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6912, + "grad_norm": 27.160118103027344, + "learning_rate": 2.3706889017894295e-06, + "loss": 7.3477277755737305, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.32945696264505386, + "loss_total_v6": 0.45923296734690666, + "step": 2997 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6914306306306306, + "grad_norm": 24.46828842163086, + "learning_rate": 2.367459312759453e-06, + "loss": 7.283612251281738, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.32238274812698364, + "loss_total_v6": 0.45522577315568924, + "step": 2998 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6916612612612613, + "grad_norm": 23.6061954498291, + "learning_rate": 2.3642312425755303e-06, + "loss": 6.72006893157959, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.29186101630330086, + "loss_total_v6": 0.42000431567430496, + "step": 2999 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6918918918918919, + "grad_norm": 24.14432144165039, + "learning_rate": 2.361004693100094e-06, + "loss": 6.825839996337891, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.29757142812013626, + "loss_total_v6": 0.42661499604582787, + "step": 3000 + }, + { + "epoch": 0.6918918918918919, + "eval_loss": 0.5713410973548889, + "eval_loss_alignment": 0.23260247217465754, + "eval_loss_alignment_w": 0.11630123608732877, + "eval_loss_sft": 0.4550398687181407, + "eval_loss_total_v6": 0.5713411045503126, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 92.2209, + "eval_samples_per_second": 18.922, + "eval_steps_per_second": 2.375, + "step": 3000 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6921225225225225, + "grad_norm": 24.001022338867188, + "learning_rate": 2.357779666194706e-06, + "loss": 7.276130676269531, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3266606256365776, + "loss_total_v6": 0.4547581747174263, + "step": 3001 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6923531531531532, + "grad_norm": 27.37128448486328, + "learning_rate": 2.3545561637200498e-06, + "loss": 6.6010308265686035, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.28411594219505787, + "loss_total_v6": 0.4125644341111183, + "step": 3002 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6925837837837838, + "grad_norm": 25.469388961791992, + "learning_rate": 2.351334187535923e-06, + "loss": 6.522342681884766, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.27980827540159225, + "loss_total_v6": 0.40764642134308815, + "step": 3003 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6928144144144144, + "grad_norm": 24.522043228149414, + "learning_rate": 2.3481137395012516e-06, + "loss": 7.50197172164917, + "loss_alignment": 0.2662353515625, + "loss_alignment_w": 0.13311767578125, + "loss_sft": 0.3358165882527828, + "loss_total_v6": 0.4688732400536537, + "step": 3004 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.693045045045045, + "grad_norm": 24.296232223510742, + "learning_rate": 2.3448948214740703e-06, + "loss": 7.174524784088135, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.3173805810511112, + "loss_total_v6": 0.4484077990055084, + "step": 3005 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6932756756756757, + "grad_norm": 25.376150131225586, + "learning_rate": 2.3416774353115387e-06, + "loss": 6.806146144866943, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2932582702487707, + "loss_total_v6": 0.42538412660360336, + "step": 3006 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6935063063063063, + "grad_norm": 24.938074111938477, + "learning_rate": 2.3384615828699286e-06, + "loss": 7.0874528884887695, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3144562803208828, + "loss_total_v6": 0.4429658092558384, + "step": 3007 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6937369369369369, + "grad_norm": 24.405006408691406, + "learning_rate": 2.3352472660046293e-06, + "loss": 6.777947425842285, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.2919993922114372, + "loss_total_v6": 0.4236217103898525, + "step": 3008 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6939675675675676, + "grad_norm": 25.046504974365234, + "learning_rate": 2.332034486570145e-06, + "loss": 6.924306869506836, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3041833806782961, + "loss_total_v6": 0.43276919052004814, + "step": 3009 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6941981981981982, + "grad_norm": 26.60991096496582, + "learning_rate": 2.328823246420086e-06, + "loss": 6.943626403808594, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.3052229918539524, + "loss_total_v6": 0.4339766539633274, + "step": 3010 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6944288288288288, + "grad_norm": 25.050207138061523, + "learning_rate": 2.325613547407184e-06, + "loss": 6.870588302612305, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.2998188864439726, + "loss_total_v6": 0.42941178381443024, + "step": 3011 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6946594594594595, + "grad_norm": 25.460657119750977, + "learning_rate": 2.322405391383273e-06, + "loss": 6.75020694732666, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.2921118959784508, + "loss_total_v6": 0.42188791185617447, + "step": 3012 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6948900900900901, + "grad_norm": 26.105382919311523, + "learning_rate": 2.3191987801993026e-06, + "loss": 6.496145248413086, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.2800477724522352, + "loss_total_v6": 0.40600908175110817, + "step": 3013 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6951207207207207, + "grad_norm": 25.387174606323242, + "learning_rate": 2.315993715705329e-06, + "loss": 6.541530609130859, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.28053453378379345, + "loss_total_v6": 0.4088456928730011, + "step": 3014 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6953513513513514, + "grad_norm": 22.682043075561523, + "learning_rate": 2.3127901997505163e-06, + "loss": 6.666051864624023, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.2879813723266125, + "loss_total_v6": 0.41662822663784027, + "step": 3015 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.695581981981982, + "grad_norm": 22.754924774169922, + "learning_rate": 2.309588234183137e-06, + "loss": 7.392765045166016, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.3317682780325413, + "loss_total_v6": 0.46204782649874687, + "step": 3016 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6958126126126126, + "grad_norm": 25.484956741333008, + "learning_rate": 2.3063878208505637e-06, + "loss": 6.631492614746094, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.28609609976410866, + "loss_total_v6": 0.41446828097105026, + "step": 3017 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6960432432432433, + "grad_norm": 26.33336067199707, + "learning_rate": 2.3031889615992804e-06, + "loss": 7.445487976074219, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.33321715146303177, + "loss_total_v6": 0.46534300222992897, + "step": 3018 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6962738738738738, + "grad_norm": 25.104585647583008, + "learning_rate": 2.2999916582748673e-06, + "loss": 6.812532424926758, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.29387104138731956, + "loss_total_v6": 0.42578326538205147, + "step": 3019 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6965045045045045, + "grad_norm": 27.895509719848633, + "learning_rate": 2.296795912722014e-06, + "loss": 7.217392921447754, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.32266905158758163, + "loss_total_v6": 0.4510870426893234, + "step": 3020 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6967351351351352, + "grad_norm": 24.15335464477539, + "learning_rate": 2.293601726784505e-06, + "loss": 6.694464683532715, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.28778880275785923, + "loss_total_v6": 0.4184040389955044, + "step": 3021 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6969657657657657, + "grad_norm": 24.720529556274414, + "learning_rate": 2.2904091023052295e-06, + "loss": 6.540475368499756, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.28081949800252914, + "loss_total_v6": 0.40877971053123474, + "step": 3022 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6971963963963964, + "grad_norm": 22.476154327392578, + "learning_rate": 2.2872180411261736e-06, + "loss": 6.681628227233887, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.2911521941423416, + "loss_total_v6": 0.4176017791032791, + "step": 3023 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6974270270270271, + "grad_norm": 24.396533966064453, + "learning_rate": 2.2840285450884232e-06, + "loss": 7.0850443840026855, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.31473300606012344, + "loss_total_v6": 0.44281526282429695, + "step": 3024 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6976576576576576, + "grad_norm": 24.114694595336914, + "learning_rate": 2.280840616032161e-06, + "loss": 6.636229515075684, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.28614798560738564, + "loss_total_v6": 0.4147643409669399, + "step": 3025 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6978882882882883, + "grad_norm": 25.952594757080078, + "learning_rate": 2.277654255796662e-06, + "loss": 6.9632487297058105, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.30544231832027435, + "loss_total_v6": 0.43520306795835495, + "step": 3026 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.698118918918919, + "grad_norm": 25.14967918395996, + "learning_rate": 2.2744694662203022e-06, + "loss": 6.899781227111816, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.30208592489361763, + "loss_total_v6": 0.43123631179332733, + "step": 3027 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6983495495495495, + "grad_norm": 25.37269401550293, + "learning_rate": 2.2712862491405437e-06, + "loss": 6.58449125289917, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.2815410904586315, + "loss_total_v6": 0.411530714482069, + "step": 3028 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6985801801801802, + "grad_norm": 25.408586502075195, + "learning_rate": 2.2681046063939477e-06, + "loss": 7.036341190338135, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3088051360100508, + "loss_total_v6": 0.4397713355720043, + "step": 3029 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6988108108108108, + "grad_norm": 28.507017135620117, + "learning_rate": 2.264924539816167e-06, + "loss": 6.821771621704102, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.2936550248414278, + "loss_total_v6": 0.42636072263121605, + "step": 3030 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6990414414414414, + "grad_norm": 26.011507034301758, + "learning_rate": 2.2617460512419378e-06, + "loss": 6.862092018127441, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.3005695976316929, + "loss_total_v6": 0.4288807660341263, + "step": 3031 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6992720720720721, + "grad_norm": 24.840845108032227, + "learning_rate": 2.258569142505098e-06, + "loss": 6.577031135559082, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.2818529959768057, + "loss_total_v6": 0.41106444224715233, + "step": 3032 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6995027027027027, + "grad_norm": 26.583866119384766, + "learning_rate": 2.2553938154385625e-06, + "loss": 6.975137710571289, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3079706486314535, + "loss_total_v6": 0.43594612553715706, + "step": 3033 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.6997333333333333, + "grad_norm": 26.123119354248047, + "learning_rate": 2.252220071874339e-06, + "loss": 7.201849460601807, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.3195156119763851, + "loss_total_v6": 0.4501155912876129, + "step": 3034 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.699963963963964, + "grad_norm": 25.901155471801758, + "learning_rate": 2.249047913643524e-06, + "loss": 7.508065700531006, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.33720456436276436, + "loss_total_v6": 0.46925411000847816, + "step": 3035 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7001945945945945, + "grad_norm": 25.96390724182129, + "learning_rate": 2.2458773425762913e-06, + "loss": 6.930237770080566, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.30614097975194454, + "loss_total_v6": 0.4331398792564869, + "step": 3036 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7004252252252252, + "grad_norm": 25.13576889038086, + "learning_rate": 2.2427083605019085e-06, + "loss": 6.725333213806152, + "loss_alignment": 0.2508544921875, + "loss_alignment_w": 0.12542724609375, + "loss_sft": 0.2945398613810539, + "loss_total_v6": 0.4203333146870136, + "step": 3037 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7006558558558559, + "grad_norm": 26.69791030883789, + "learning_rate": 2.2395409692487174e-06, + "loss": 6.5844950675964355, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.2840895429253578, + "loss_total_v6": 0.4115309529006481, + "step": 3038 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7008864864864864, + "grad_norm": 26.359155654907227, + "learning_rate": 2.236375170644149e-06, + "loss": 6.863422393798828, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.30007290095090866, + "loss_total_v6": 0.42896389216184616, + "step": 3039 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7011171171171171, + "grad_norm": 24.68924331665039, + "learning_rate": 2.233210966514713e-06, + "loss": 7.165501594543457, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3161757364869118, + "loss_total_v6": 0.44784383103251457, + "step": 3040 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7013477477477478, + "grad_norm": 22.827926635742188, + "learning_rate": 2.2300483586859977e-06, + "loss": 6.689923286437988, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.286665728315711, + "loss_total_v6": 0.41812021657824516, + "step": 3041 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7015783783783783, + "grad_norm": 23.311716079711914, + "learning_rate": 2.226887348982675e-06, + "loss": 7.0012593269348145, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.308367308229208, + "loss_total_v6": 0.4375787116587162, + "step": 3042 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.701809009009009, + "grad_norm": 27.54511070251465, + "learning_rate": 2.2237279392284878e-06, + "loss": 6.922720909118652, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3008341062813997, + "loss_total_v6": 0.4326700419187546, + "step": 3043 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7020396396396397, + "grad_norm": 25.52315330505371, + "learning_rate": 2.220570131246262e-06, + "loss": 6.267599582672119, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.2626661378890276, + "loss_total_v6": 0.39172497764229774, + "step": 3044 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7022702702702702, + "grad_norm": 24.5150146484375, + "learning_rate": 2.217413926857894e-06, + "loss": 7.3167829513549805, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3274161033332348, + "loss_total_v6": 0.4572989344596863, + "step": 3045 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7025009009009009, + "grad_norm": 24.59212875366211, + "learning_rate": 2.214259327884362e-06, + "loss": 6.982226371765137, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3050414714962244, + "loss_total_v6": 0.43638912215828896, + "step": 3046 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7027315315315316, + "grad_norm": 25.92963981628418, + "learning_rate": 2.2111063361457104e-06, + "loss": 6.508479118347168, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2769429013133049, + "loss_total_v6": 0.4067799299955368, + "step": 3047 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7029621621621621, + "grad_norm": 26.096128463745117, + "learning_rate": 2.2079549534610606e-06, + "loss": 7.161387920379639, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3161475360393524, + "loss_total_v6": 0.4475867450237274, + "step": 3048 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7031927927927928, + "grad_norm": 25.271739959716797, + "learning_rate": 2.2048051816486054e-06, + "loss": 6.590362548828125, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2817554194480181, + "loss_total_v6": 0.411897636950016, + "step": 3049 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7034234234234235, + "grad_norm": 24.484516143798828, + "learning_rate": 2.201657022525608e-06, + "loss": 6.797628879547119, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.2943128664046526, + "loss_total_v6": 0.42485179379582405, + "step": 3050 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.703654054054054, + "grad_norm": 22.154895782470703, + "learning_rate": 2.198510477908402e-06, + "loss": 7.349128723144531, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.3295445516705513, + "loss_total_v6": 0.4593205563724041, + "step": 3051 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7038846846846847, + "grad_norm": 23.631994247436523, + "learning_rate": 2.1953655496123854e-06, + "loss": 7.13306999206543, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.31739893555641174, + "loss_total_v6": 0.44581690430641174, + "step": 3052 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7041153153153153, + "grad_norm": 26.341007232666016, + "learning_rate": 2.1922222394520304e-06, + "loss": 6.752073287963867, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.2939833663403988, + "loss_total_v6": 0.4220045804977417, + "step": 3053 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7043459459459459, + "grad_norm": 25.47161865234375, + "learning_rate": 2.1890805492408674e-06, + "loss": 6.667606353759766, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.28702568635344505, + "loss_total_v6": 0.41672538220882416, + "step": 3054 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7045765765765766, + "grad_norm": 24.657716751098633, + "learning_rate": 2.1859404807914985e-06, + "loss": 7.369731903076172, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3278415072709322, + "loss_total_v6": 0.46060824394226074, + "step": 3055 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7048072072072072, + "grad_norm": 27.8658504486084, + "learning_rate": 2.1828020359155905e-06, + "loss": 6.71087646484375, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.28918078541755676, + "loss_total_v6": 0.41942980140447617, + "step": 3056 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7050378378378378, + "grad_norm": 23.267969131469727, + "learning_rate": 2.1796652164238655e-06, + "loss": 7.101382255554199, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.31413667276501656, + "loss_total_v6": 0.44383638724684715, + "step": 3057 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7052684684684685, + "grad_norm": 25.847232818603516, + "learning_rate": 2.176530024126121e-06, + "loss": 6.468836784362793, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.27455680817365646, + "loss_total_v6": 0.40430229157209396, + "step": 3058 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7054990990990991, + "grad_norm": 26.805269241333008, + "learning_rate": 2.1733964608312024e-06, + "loss": 7.0950422286987305, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.31027668714523315, + "loss_total_v6": 0.44344014301896095, + "step": 3059 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7057297297297297, + "grad_norm": 23.141477584838867, + "learning_rate": 2.1702645283470238e-06, + "loss": 6.709403038024902, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.28850883431732655, + "loss_total_v6": 0.4193377047777176, + "step": 3060 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7059603603603604, + "grad_norm": 24.769670486450195, + "learning_rate": 2.167134228480556e-06, + "loss": 6.555514335632324, + "loss_alignment": 0.2523193359375, + "loss_alignment_w": 0.12615966796875, + "loss_sft": 0.2833616305142641, + "loss_total_v6": 0.40971966460347176, + "step": 3061 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.706190990990991, + "grad_norm": 23.350692749023438, + "learning_rate": 2.1640055630378254e-06, + "loss": 6.793214321136475, + "loss_alignment": 0.2532958984375, + "loss_alignment_w": 0.12664794921875, + "loss_sft": 0.29805003106594086, + "loss_total_v6": 0.42457589507102966, + "step": 3062 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7064216216216216, + "grad_norm": 27.4858455657959, + "learning_rate": 2.1608785338239203e-06, + "loss": 6.831315517425537, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.296815000474453, + "loss_total_v6": 0.4269572086632252, + "step": 3063 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7066522522522523, + "grad_norm": 27.138132095336914, + "learning_rate": 2.157753142642978e-06, + "loss": 7.0635881423950195, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.3134530298411846, + "loss_total_v6": 0.4414742738008499, + "step": 3064 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7068828828828829, + "grad_norm": 24.324880599975586, + "learning_rate": 2.1546293912981985e-06, + "loss": 6.916900634765625, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.30513953790068626, + "loss_total_v6": 0.43230628594756126, + "step": 3065 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7071135135135135, + "grad_norm": 24.28141975402832, + "learning_rate": 2.151507281591831e-06, + "loss": 6.959546089172363, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.3063400201499462, + "loss_total_v6": 0.4349716268479824, + "step": 3066 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7073441441441441, + "grad_norm": 24.22821807861328, + "learning_rate": 2.148386815325179e-06, + "loss": 7.328300952911377, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.32690001279115677, + "loss_total_v6": 0.45801880955696106, + "step": 3067 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7075747747747748, + "grad_norm": 28.737314224243164, + "learning_rate": 2.145267994298599e-06, + "loss": 6.63049840927124, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.28212770633399487, + "loss_total_v6": 0.414406169205904, + "step": 3068 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7078054054054054, + "grad_norm": 28.188087463378906, + "learning_rate": 2.1421508203114937e-06, + "loss": 7.281857490539551, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3243025168776512, + "loss_total_v6": 0.4551160968840122, + "step": 3069 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.708036036036036, + "grad_norm": 30.452051162719727, + "learning_rate": 2.1390352951623224e-06, + "loss": 6.639837741851807, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.28644982539117336, + "loss_total_v6": 0.4149898625910282, + "step": 3070 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7082666666666667, + "grad_norm": 24.96411895751953, + "learning_rate": 2.1359214206485845e-06, + "loss": 7.063058853149414, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.31126843206584454, + "loss_total_v6": 0.441441185772419, + "step": 3071 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7084972972972973, + "grad_norm": 23.379383087158203, + "learning_rate": 2.132809198566837e-06, + "loss": 6.940898895263672, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.3070209063589573, + "loss_total_v6": 0.4338061697781086, + "step": 3072 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7087279279279279, + "grad_norm": 28.81734275817871, + "learning_rate": 2.1296986307126736e-06, + "loss": 7.414336681365967, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3345203083008528, + "loss_total_v6": 0.4633960425853729, + "step": 3073 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7089585585585586, + "grad_norm": 22.317350387573242, + "learning_rate": 2.1265897188807418e-06, + "loss": 7.285595893859863, + "loss_alignment": 0.25048828125, + "loss_alignment_w": 0.125244140625, + "loss_sft": 0.32992249354720116, + "loss_total_v6": 0.45534972846508026, + "step": 3074 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7091891891891892, + "grad_norm": 26.392187118530273, + "learning_rate": 2.1234824648647283e-06, + "loss": 6.4868974685668945, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.27528889290988445, + "loss_total_v6": 0.4054310992360115, + "step": 3075 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7094198198198198, + "grad_norm": 24.212242126464844, + "learning_rate": 2.1203768704573675e-06, + "loss": 7.241236686706543, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.32426613569259644, + "loss_total_v6": 0.45257728919386864, + "step": 3076 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7096504504504505, + "grad_norm": 31.10893440246582, + "learning_rate": 2.1172729374504344e-06, + "loss": 6.985756874084473, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.31025176867842674, + "loss_total_v6": 0.43660980090498924, + "step": 3077 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7098810810810811, + "grad_norm": 25.099529266357422, + "learning_rate": 2.114170667634741e-06, + "loss": 6.7155656814575195, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.2904503680765629, + "loss_total_v6": 0.4197228290140629, + "step": 3078 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7101117117117117, + "grad_norm": 23.279733657836914, + "learning_rate": 2.1110700628001497e-06, + "loss": 6.556265830993652, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.2793344724923372, + "loss_total_v6": 0.4097666069865227, + "step": 3079 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7103423423423424, + "grad_norm": 22.29045295715332, + "learning_rate": 2.1079711247355505e-06, + "loss": 6.718151569366455, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.28896404430270195, + "loss_total_v6": 0.41988446563482285, + "step": 3080 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.710572972972973, + "grad_norm": 24.820270538330078, + "learning_rate": 2.104873855228881e-06, + "loss": 6.819247245788574, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.2998296767473221, + "loss_total_v6": 0.4262029603123665, + "step": 3081 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7108036036036036, + "grad_norm": 26.19053840637207, + "learning_rate": 2.1017782560671124e-06, + "loss": 6.556998252868652, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.2823251988738775, + "loss_total_v6": 0.409812368452549, + "step": 3082 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7110342342342343, + "grad_norm": 27.212657928466797, + "learning_rate": 2.098684329036253e-06, + "loss": 6.812983512878418, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2955929394811392, + "loss_total_v6": 0.42581145092844963, + "step": 3083 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7112648648648648, + "grad_norm": 28.642736434936523, + "learning_rate": 2.095592075921347e-06, + "loss": 6.659981727600098, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.28598458878695965, + "loss_total_v6": 0.4162488840520382, + "step": 3084 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7114954954954955, + "grad_norm": 26.219614028930664, + "learning_rate": 2.092501498506469e-06, + "loss": 7.098618507385254, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.31355199217796326, + "loss_total_v6": 0.44366367161273956, + "step": 3085 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7117261261261262, + "grad_norm": 23.75528907775879, + "learning_rate": 2.089412598574733e-06, + "loss": 7.001537322998047, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.3080489709973335, + "loss_total_v6": 0.43759608268737793, + "step": 3086 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7119567567567567, + "grad_norm": 25.061054229736328, + "learning_rate": 2.086325377908278e-06, + "loss": 7.048810005187988, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3118274938315153, + "loss_total_v6": 0.44055063277482986, + "step": 3087 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7121873873873874, + "grad_norm": 29.949373245239258, + "learning_rate": 2.08323983828828e-06, + "loss": 6.931754112243652, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.30512188002467155, + "loss_total_v6": 0.43323465064167976, + "step": 3088 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7124180180180181, + "grad_norm": 24.593599319458008, + "learning_rate": 2.0801559814949445e-06, + "loss": 6.889473915100098, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.3041883409023285, + "loss_total_v6": 0.4305921457707882, + "step": 3089 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7126486486486486, + "grad_norm": 26.546213150024414, + "learning_rate": 2.0770738093075015e-06, + "loss": 7.195455551147461, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.3211454004049301, + "loss_total_v6": 0.449715968221426, + "step": 3090 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7128792792792793, + "grad_norm": 23.994441986083984, + "learning_rate": 2.073993323504214e-06, + "loss": 7.2294511795043945, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.32360584661364555, + "loss_total_v6": 0.45184071362018585, + "step": 3091 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7131099099099099, + "grad_norm": 25.403554916381836, + "learning_rate": 2.07091452586237e-06, + "loss": 7.184946060180664, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.3214346393942833, + "loss_total_v6": 0.4490591511130333, + "step": 3092 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7133405405405405, + "grad_norm": 27.103734970092773, + "learning_rate": 2.0678374181582845e-06, + "loss": 6.500260353088379, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.27530008740723133, + "loss_total_v6": 0.406266275793314, + "step": 3093 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7135711711711712, + "grad_norm": 24.79615592956543, + "learning_rate": 2.0647620021672976e-06, + "loss": 6.544267654418945, + "loss_alignment": 0.253662109375, + "loss_alignment_w": 0.1268310546875, + "loss_sft": 0.282185697928071, + "loss_total_v6": 0.4090167433023453, + "step": 3094 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7138018018018017, + "grad_norm": 27.136058807373047, + "learning_rate": 2.0616882796637694e-06, + "loss": 6.751044273376465, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.2931408043950796, + "loss_total_v6": 0.42194024100899696, + "step": 3095 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7140324324324324, + "grad_norm": 21.582399368286133, + "learning_rate": 2.0586162524210895e-06, + "loss": 6.550154685974121, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.28035634383559227, + "loss_total_v6": 0.40938467159867287, + "step": 3096 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7142630630630631, + "grad_norm": 26.411537170410156, + "learning_rate": 2.055545922211663e-06, + "loss": 6.889097213745117, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3019369803369045, + "loss_total_v6": 0.4305685758590698, + "step": 3097 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7144936936936936, + "grad_norm": 22.413043975830078, + "learning_rate": 2.052477290806922e-06, + "loss": 6.743439197540283, + "loss_alignment": 0.253662109375, + "loss_alignment_w": 0.1268310546875, + "loss_sft": 0.29455759562551975, + "loss_total_v6": 0.4214649498462677, + "step": 3098 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7147243243243243, + "grad_norm": 25.16144371032715, + "learning_rate": 2.0494103599773115e-06, + "loss": 6.76145076751709, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.2936844564974308, + "loss_total_v6": 0.4225906953215599, + "step": 3099 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.714954954954955, + "grad_norm": 27.514638900756836, + "learning_rate": 2.0463451314923016e-06, + "loss": 6.877513408660889, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.29987022280693054, + "loss_total_v6": 0.42984458059072495, + "step": 3100 + }, + { + "epoch": 0.714954954954955, + "eval_loss": 0.5707520246505737, + "eval_loss_alignment": 0.2321476348458904, + "eval_loss_alignment_w": 0.1160738174229452, + "eval_loss_sft": 0.45467823311754557, + "eval_loss_total_v6": 0.5707520505319855, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 92.3745, + "eval_samples_per_second": 18.89, + "eval_steps_per_second": 2.371, + "step": 3100 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7151855855855855, + "grad_norm": 24.910306930541992, + "learning_rate": 2.0432816071203776e-06, + "loss": 7.126219272613525, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.31350697949528694, + "loss_total_v6": 0.44538870081305504, + "step": 3101 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7154162162162162, + "grad_norm": 22.826688766479492, + "learning_rate": 2.0402197886290416e-06, + "loss": 7.45893669128418, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.33619391918182373, + "loss_total_v6": 0.46618354693055153, + "step": 3102 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7156468468468469, + "grad_norm": 24.59061050415039, + "learning_rate": 2.037159677784814e-06, + "loss": 6.906772136688232, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3013632223010063, + "loss_total_v6": 0.4316732659935951, + "step": 3103 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7158774774774774, + "grad_norm": 26.42563819885254, + "learning_rate": 2.0341012763532243e-06, + "loss": 6.997034549713135, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.3090340234339237, + "loss_total_v6": 0.4373146556317806, + "step": 3104 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7161081081081081, + "grad_norm": 30.390104293823242, + "learning_rate": 2.031044586098822e-06, + "loss": 7.041591644287109, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.31078122556209564, + "loss_total_v6": 0.44009946659207344, + "step": 3105 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7163387387387388, + "grad_norm": 23.163333892822266, + "learning_rate": 2.0279896087851634e-06, + "loss": 7.082170486450195, + "loss_alignment": 0.2532958984375, + "loss_alignment_w": 0.12664794921875, + "loss_sft": 0.3158809095621109, + "loss_total_v6": 0.442635677754879, + "step": 3106 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7165693693693693, + "grad_norm": 23.39253807067871, + "learning_rate": 2.0249363461748223e-06, + "loss": 7.03364896774292, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.3121006563305855, + "loss_total_v6": 0.4396030679345131, + "step": 3107 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7168, + "grad_norm": 26.941364288330078, + "learning_rate": 2.021884800029379e-06, + "loss": 6.539836883544922, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.28274800442159176, + "loss_total_v6": 0.4087398238480091, + "step": 3108 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7170306306306307, + "grad_norm": 30.15928840637207, + "learning_rate": 2.018834972109426e-06, + "loss": 6.7849225997924805, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2920996677130461, + "loss_total_v6": 0.4240576773881912, + "step": 3109 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7172612612612612, + "grad_norm": 27.629432678222656, + "learning_rate": 2.015786864174564e-06, + "loss": 6.724306106567383, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.2913323640823364, + "loss_total_v6": 0.4202691316604614, + "step": 3110 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7174918918918919, + "grad_norm": 23.578229904174805, + "learning_rate": 2.0127404779833993e-06, + "loss": 7.334839344024658, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3284836243838072, + "loss_total_v6": 0.45842747390270233, + "step": 3111 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7177225225225226, + "grad_norm": 26.164304733276367, + "learning_rate": 2.009695815293548e-06, + "loss": 7.32705020904541, + "loss_alignment": 0.2645263671875, + "loss_alignment_w": 0.13226318359375, + "loss_sft": 0.32546382024884224, + "loss_total_v6": 0.45794064179062843, + "step": 3112 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7179531531531531, + "grad_norm": 23.63446044921875, + "learning_rate": 2.0066528778616284e-06, + "loss": 7.14874267578125, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3157997317612171, + "loss_total_v6": 0.4467964209616184, + "step": 3113 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7181837837837838, + "grad_norm": 22.725566864013672, + "learning_rate": 2.0036116674432653e-06, + "loss": 6.991196632385254, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.30679231137037277, + "loss_total_v6": 0.43694978579878807, + "step": 3114 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7184144144144144, + "grad_norm": 24.732994079589844, + "learning_rate": 2.0005721857930902e-06, + "loss": 6.457385063171387, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.2773048095405102, + "loss_total_v6": 0.4035865403711796, + "step": 3115 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.718645045045045, + "grad_norm": 27.129291534423828, + "learning_rate": 1.9975344346647296e-06, + "loss": 6.836067199707031, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.29702044278383255, + "loss_total_v6": 0.42725419998168945, + "step": 3116 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7188756756756757, + "grad_norm": 26.75857162475586, + "learning_rate": 1.9944984158108185e-06, + "loss": 7.0067973136901855, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30647037364542484, + "loss_total_v6": 0.4379248358309269, + "step": 3117 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7191063063063063, + "grad_norm": 26.22890281677246, + "learning_rate": 1.9914641309829885e-06, + "loss": 6.56202507019043, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.28314293175935745, + "loss_total_v6": 0.41012656316161156, + "step": 3118 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7193369369369369, + "grad_norm": 25.239646911621094, + "learning_rate": 1.988431581931874e-06, + "loss": 6.926095962524414, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3032118231058121, + "loss_total_v6": 0.4328810088336468, + "step": 3119 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7195675675675676, + "grad_norm": 21.89544677734375, + "learning_rate": 1.9854007704071066e-06, + "loss": 6.7730021476745605, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.29277369752526283, + "loss_total_v6": 0.42331263050436974, + "step": 3120 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7197981981981982, + "grad_norm": 24.388729095458984, + "learning_rate": 1.982371698157312e-06, + "loss": 7.117463111877441, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.3155079744756222, + "loss_total_v6": 0.444841455668211, + "step": 3121 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7200288288288288, + "grad_norm": 23.331945419311523, + "learning_rate": 1.979344366930119e-06, + "loss": 6.925832748413086, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.304812790825963, + "loss_total_v6": 0.43286455795168877, + "step": 3122 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7202594594594595, + "grad_norm": 26.085458755493164, + "learning_rate": 1.9763187784721444e-06, + "loss": 6.9037346839904785, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.3039352037012577, + "loss_total_v6": 0.4314834251999855, + "step": 3123 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7204900900900901, + "grad_norm": 29.71018409729004, + "learning_rate": 1.973294934529007e-06, + "loss": 6.93456506729126, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.305923143401742, + "loss_total_v6": 0.43341032788157463, + "step": 3124 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7207207207207207, + "grad_norm": 25.873287200927734, + "learning_rate": 1.9702728368453128e-06, + "loss": 6.567055702209473, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.28119904920458794, + "loss_total_v6": 0.41044098883867264, + "step": 3125 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7209513513513514, + "grad_norm": 25.748821258544922, + "learning_rate": 1.967252487164663e-06, + "loss": 6.636003494262695, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.28387559950351715, + "loss_total_v6": 0.41475023701786995, + "step": 3126 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.721181981981982, + "grad_norm": 25.796770095825195, + "learning_rate": 1.964233887229652e-06, + "loss": 6.823944568634033, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.2988109830766916, + "loss_total_v6": 0.4264965318143368, + "step": 3127 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7214126126126126, + "grad_norm": 22.738628387451172, + "learning_rate": 1.961217038781863e-06, + "loss": 6.864300727844238, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.30191309191286564, + "loss_total_v6": 0.4290187917649746, + "step": 3128 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7216432432432432, + "grad_norm": 21.99645233154297, + "learning_rate": 1.958201943561869e-06, + "loss": 6.889681816101074, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.29931847378611565, + "loss_total_v6": 0.43060509860515594, + "step": 3129 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7218738738738739, + "grad_norm": 23.404993057250977, + "learning_rate": 1.9551886033092286e-06, + "loss": 6.769946098327637, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.29358979873359203, + "loss_total_v6": 0.4231216460466385, + "step": 3130 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7221045045045045, + "grad_norm": 24.485252380371094, + "learning_rate": 1.952177019762494e-06, + "loss": 7.2106614112854, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3198985084891319, + "loss_total_v6": 0.4506663419306278, + "step": 3131 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7223351351351351, + "grad_norm": 29.492111206054688, + "learning_rate": 1.9491671946591963e-06, + "loss": 6.896327495574951, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.30199215188622475, + "loss_total_v6": 0.43102047219872475, + "step": 3132 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7225657657657658, + "grad_norm": 24.661710739135742, + "learning_rate": 1.9461591297358577e-06, + "loss": 7.0550689697265625, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.31110478192567825, + "loss_total_v6": 0.44094182923436165, + "step": 3133 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7227963963963964, + "grad_norm": 25.923004150390625, + "learning_rate": 1.9431528267279842e-06, + "loss": 7.215617656707764, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3219325318932533, + "loss_total_v6": 0.45097610726952553, + "step": 3134 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.723027027027027, + "grad_norm": 23.658935546875, + "learning_rate": 1.940148287370063e-06, + "loss": 6.957344055175781, + "loss_alignment": 0.2503662109375, + "loss_alignment_w": 0.12518310546875, + "loss_sft": 0.30931519716978073, + "loss_total_v6": 0.43483397737145424, + "step": 3135 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7232576576576577, + "grad_norm": 23.805538177490234, + "learning_rate": 1.9371455133955676e-06, + "loss": 6.580194473266602, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.28171506524086, + "loss_total_v6": 0.4112621806561947, + "step": 3136 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7234882882882883, + "grad_norm": 24.212520599365234, + "learning_rate": 1.934144506536946e-06, + "loss": 6.668725967407227, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.287797536700964, + "loss_total_v6": 0.41679535061120987, + "step": 3137 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7237189189189189, + "grad_norm": 25.94756507873535, + "learning_rate": 1.9311452685256364e-06, + "loss": 6.85653018951416, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.296804029494524, + "loss_total_v6": 0.4285331405699253, + "step": 3138 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7239495495495496, + "grad_norm": 27.246644973754883, + "learning_rate": 1.9281478010920457e-06, + "loss": 6.990622520446777, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3088163509964943, + "loss_total_v6": 0.4369138963520527, + "step": 3139 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7241801801801802, + "grad_norm": 24.523914337158203, + "learning_rate": 1.925152105965567e-06, + "loss": 6.604249000549316, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.2823181729763746, + "loss_total_v6": 0.4127655662596226, + "step": 3140 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7244108108108108, + "grad_norm": 25.600833892822266, + "learning_rate": 1.9221581848745708e-06, + "loss": 6.563772201538086, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.2799562141299248, + "loss_total_v6": 0.4102357588708401, + "step": 3141 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7246414414414415, + "grad_norm": 24.473033905029297, + "learning_rate": 1.919166039546398e-06, + "loss": 7.070286273956299, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.31249839439988136, + "loss_total_v6": 0.44189290702342987, + "step": 3142 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.724872072072072, + "grad_norm": 23.859066009521484, + "learning_rate": 1.91617567170737e-06, + "loss": 6.576066493988037, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2809840179979801, + "loss_total_v6": 0.4110041595995426, + "step": 3143 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7251027027027027, + "grad_norm": 26.637725830078125, + "learning_rate": 1.913187083082782e-06, + "loss": 6.848593711853027, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2983221337199211, + "loss_total_v6": 0.4280371069908142, + "step": 3144 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7253333333333334, + "grad_norm": 24.532411575317383, + "learning_rate": 1.9102002753969016e-06, + "loss": 6.807437419891357, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.29677221179008484, + "loss_total_v6": 0.42546483501791954, + "step": 3145 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7255639639639639, + "grad_norm": 27.149747848510742, + "learning_rate": 1.9072152503729723e-06, + "loss": 6.6828813552856445, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.28915530629456043, + "loss_total_v6": 0.4176800921559334, + "step": 3146 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7257945945945946, + "grad_norm": 26.94331169128418, + "learning_rate": 1.9042320097332018e-06, + "loss": 6.810236930847168, + "loss_alignment": 0.2496337890625, + "loss_alignment_w": 0.12481689453125, + "loss_sft": 0.3009449765086174, + "loss_total_v6": 0.4256398044526577, + "step": 3147 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7260252252252253, + "grad_norm": 28.769609451293945, + "learning_rate": 1.9012505551987764e-06, + "loss": 7.338519096374512, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3260432966053486, + "loss_total_v6": 0.4586574323475361, + "step": 3148 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7262558558558558, + "grad_norm": 22.972841262817383, + "learning_rate": 1.8982708884898455e-06, + "loss": 6.446961402893066, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.27364737540483475, + "loss_total_v6": 0.40293510630726814, + "step": 3149 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7264864864864865, + "grad_norm": 27.161651611328125, + "learning_rate": 1.8952930113255313e-06, + "loss": 6.405646800994873, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.2738575581461191, + "loss_total_v6": 0.40035292133688927, + "step": 3150 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7267171171171171, + "grad_norm": 25.596378326416016, + "learning_rate": 1.8923169254239205e-06, + "loss": 7.061069965362549, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.3101828433573246, + "loss_total_v6": 0.4413168802857399, + "step": 3151 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7269477477477477, + "grad_norm": 26.12185287475586, + "learning_rate": 1.8893426325020685e-06, + "loss": 6.95114803314209, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.30724948085844517, + "loss_total_v6": 0.4344467520713806, + "step": 3152 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7271783783783784, + "grad_norm": 28.54520606994629, + "learning_rate": 1.8863701342759954e-06, + "loss": 6.702894687652588, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2886971589177847, + "loss_total_v6": 0.41893091425299644, + "step": 3153 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.727409009009009, + "grad_norm": 28.050811767578125, + "learning_rate": 1.883399432460687e-06, + "loss": 6.991063117980957, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.3055327609181404, + "loss_total_v6": 0.4369414560496807, + "step": 3154 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7276396396396396, + "grad_norm": 25.789031982421875, + "learning_rate": 1.8804305287700919e-06, + "loss": 6.706089973449707, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2891104891896248, + "loss_total_v6": 0.4191306084394455, + "step": 3155 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7278702702702703, + "grad_norm": 24.001298904418945, + "learning_rate": 1.8774634249171187e-06, + "loss": 6.696839332580566, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.28950887359678745, + "loss_total_v6": 0.4185524508357048, + "step": 3156 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7281009009009008, + "grad_norm": 24.95473289489746, + "learning_rate": 1.8744981226136427e-06, + "loss": 6.985772609710693, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.30921516194939613, + "loss_total_v6": 0.43661078438162804, + "step": 3157 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7283315315315315, + "grad_norm": 24.110004425048828, + "learning_rate": 1.871534623570494e-06, + "loss": 6.886171340942383, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.30033504217863083, + "loss_total_v6": 0.43038570135831833, + "step": 3158 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7285621621621622, + "grad_norm": 23.518260955810547, + "learning_rate": 1.8685729294974668e-06, + "loss": 6.769558429718018, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.2949846088886261, + "loss_total_v6": 0.4230974018573761, + "step": 3159 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7287927927927927, + "grad_norm": 25.76371955871582, + "learning_rate": 1.8656130421033125e-06, + "loss": 6.786533355712891, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.29656434059143066, + "loss_total_v6": 0.42415834218263626, + "step": 3160 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7290234234234234, + "grad_norm": 21.13543701171875, + "learning_rate": 1.8626549630957397e-06, + "loss": 6.390251159667969, + "loss_alignment": 0.251220703125, + "loss_alignment_w": 0.1256103515625, + "loss_sft": 0.2739939745515585, + "loss_total_v6": 0.39939069747924805, + "step": 3161 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7292540540540541, + "grad_norm": 25.900545120239258, + "learning_rate": 1.8596986941814161e-06, + "loss": 6.7390336990356445, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.292680099606514, + "loss_total_v6": 0.4211896136403084, + "step": 3162 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7294846846846846, + "grad_norm": 28.065107345581055, + "learning_rate": 1.8567442370659589e-06, + "loss": 6.67446231842041, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.28673706017434597, + "loss_total_v6": 0.4171539209783077, + "step": 3163 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7297153153153153, + "grad_norm": 26.663549423217773, + "learning_rate": 1.8537915934539485e-06, + "loss": 6.934695720672607, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3023759853094816, + "loss_total_v6": 0.43341848254203796, + "step": 3164 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.729945945945946, + "grad_norm": 25.20684242248535, + "learning_rate": 1.8508407650489101e-06, + "loss": 6.790576457977295, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.29480286687612534, + "loss_total_v6": 0.42441101744771004, + "step": 3165 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7301765765765765, + "grad_norm": 26.047222137451172, + "learning_rate": 1.8478917535533291e-06, + "loss": 7.320138454437256, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3269544281065464, + "loss_total_v6": 0.4575086496770382, + "step": 3166 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7304072072072072, + "grad_norm": 24.559667587280273, + "learning_rate": 1.8449445606686395e-06, + "loss": 6.584673881530762, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.2821933701634407, + "loss_total_v6": 0.4115421175956726, + "step": 3167 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7306378378378379, + "grad_norm": 26.360265731811523, + "learning_rate": 1.8419991880952242e-06, + "loss": 6.895212650299072, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.30155626498162746, + "loss_total_v6": 0.4309507869184017, + "step": 3168 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7308684684684684, + "grad_norm": 22.794151306152344, + "learning_rate": 1.8390556375324187e-06, + "loss": 7.12966251373291, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.3158736601471901, + "loss_total_v6": 0.4456038922071457, + "step": 3169 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7310990990990991, + "grad_norm": 26.093355178833008, + "learning_rate": 1.836113910678507e-06, + "loss": 6.8859710693359375, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3010244481265545, + "loss_total_v6": 0.4303731918334961, + "step": 3170 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7313297297297298, + "grad_norm": 27.09285545349121, + "learning_rate": 1.8331740092307215e-06, + "loss": 6.857864856719971, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.300503758713603, + "loss_total_v6": 0.42861655727028847, + "step": 3171 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7315603603603603, + "grad_norm": 25.787437438964844, + "learning_rate": 1.8302359348852373e-06, + "loss": 7.1292643547058105, + "loss_alignment": 0.267578125, + "loss_alignment_w": 0.1337890625, + "loss_sft": 0.31175944209098816, + "loss_total_v6": 0.44557901844382286, + "step": 3172 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.731790990990991, + "grad_norm": 24.7158145904541, + "learning_rate": 1.8272996893371791e-06, + "loss": 7.1741943359375, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.31774142384529114, + "loss_total_v6": 0.44838716089725494, + "step": 3173 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7320216216216217, + "grad_norm": 24.67144775390625, + "learning_rate": 1.8243652742806183e-06, + "loss": 7.118906021118164, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.31408752873539925, + "loss_total_v6": 0.44493163377046585, + "step": 3174 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7322522522522522, + "grad_norm": 27.67255401611328, + "learning_rate": 1.821432691408564e-06, + "loss": 6.6708173751831055, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.28933209367096424, + "loss_total_v6": 0.4169260859489441, + "step": 3175 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7324828828828829, + "grad_norm": 28.242538452148438, + "learning_rate": 1.8185019424129752e-06, + "loss": 6.798077583312988, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.2959888558834791, + "loss_total_v6": 0.4248798340559006, + "step": 3176 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7327135135135135, + "grad_norm": 25.21698760986328, + "learning_rate": 1.8155730289847457e-06, + "loss": 7.422798156738281, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3334469608962536, + "loss_total_v6": 0.463924877345562, + "step": 3177 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7329441441441441, + "grad_norm": 23.944869995117188, + "learning_rate": 1.8126459528137141e-06, + "loss": 7.331433296203613, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.32862166687846184, + "loss_total_v6": 0.45821456238627434, + "step": 3178 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7331747747747748, + "grad_norm": 26.546419143676758, + "learning_rate": 1.8097207155886636e-06, + "loss": 6.633266448974609, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2846505790948868, + "loss_total_v6": 0.4145791679620743, + "step": 3179 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7334054054054054, + "grad_norm": 26.136138916015625, + "learning_rate": 1.8067973189973076e-06, + "loss": 7.176331520080566, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.3189125843346119, + "loss_total_v6": 0.448520727455616, + "step": 3180 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.733636036036036, + "grad_norm": 23.911439895629883, + "learning_rate": 1.8038757647263045e-06, + "loss": 7.229726314544678, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.32482848688960075, + "loss_total_v6": 0.45185789838433266, + "step": 3181 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7338666666666667, + "grad_norm": 28.561601638793945, + "learning_rate": 1.8009560544612436e-06, + "loss": 6.890863418579102, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.3009792473167181, + "loss_total_v6": 0.43067894503474236, + "step": 3182 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7340972972972973, + "grad_norm": 22.932506561279297, + "learning_rate": 1.7980381898866566e-06, + "loss": 7.167299270629883, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.3178750164806843, + "loss_total_v6": 0.4479561932384968, + "step": 3183 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7343279279279279, + "grad_norm": 26.035871505737305, + "learning_rate": 1.7951221726860047e-06, + "loss": 6.839509963989258, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.3009129911661148, + "loss_total_v6": 0.4274693839251995, + "step": 3184 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7345585585585586, + "grad_norm": 23.48263168334961, + "learning_rate": 1.7922080045416867e-06, + "loss": 7.441176414489746, + "loss_alignment": 0.2640380859375, + "loss_alignment_w": 0.13201904296875, + "loss_sft": 0.33299342915415764, + "loss_total_v6": 0.46507351845502853, + "step": 3185 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7347891891891892, + "grad_norm": 25.19296646118164, + "learning_rate": 1.7892956871350347e-06, + "loss": 6.625078201293945, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.2856036424636841, + "loss_total_v6": 0.414067380130291, + "step": 3186 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7350198198198198, + "grad_norm": 26.00220489501953, + "learning_rate": 1.7863852221463117e-06, + "loss": 7.444954872131348, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.33416040800511837, + "loss_total_v6": 0.4653096869587898, + "step": 3187 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7352504504504505, + "grad_norm": 26.49005889892578, + "learning_rate": 1.7834766112547142e-06, + "loss": 6.290240287780762, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.2609073407948017, + "loss_total_v6": 0.3931400142610073, + "step": 3188 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7354810810810811, + "grad_norm": 24.502534866333008, + "learning_rate": 1.780569856138364e-06, + "loss": 6.849946975708008, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.30202305130660534, + "loss_total_v6": 0.4281216897070408, + "step": 3189 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7357117117117117, + "grad_norm": 25.707778930664062, + "learning_rate": 1.7776649584743178e-06, + "loss": 6.8028764724731445, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.2944882698357105, + "loss_total_v6": 0.42517979070544243, + "step": 3190 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7359423423423423, + "grad_norm": 25.1729679107666, + "learning_rate": 1.7747619199385552e-06, + "loss": 6.5700836181640625, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.282776802778244, + "loss_total_v6": 0.4106302037835121, + "step": 3191 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.736172972972973, + "grad_norm": 28.813695907592773, + "learning_rate": 1.771860742205988e-06, + "loss": 7.147172451019287, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.31728848814964294, + "loss_total_v6": 0.44669828563928604, + "step": 3192 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7364036036036036, + "grad_norm": 29.374704360961914, + "learning_rate": 1.7689614269504546e-06, + "loss": 7.273113250732422, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.32343554869294167, + "loss_total_v6": 0.45456958562135696, + "step": 3193 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7366342342342342, + "grad_norm": 27.503637313842773, + "learning_rate": 1.766063975844713e-06, + "loss": 6.779653549194336, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.295005202293396, + "loss_total_v6": 0.4237283356487751, + "step": 3194 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7368648648648649, + "grad_norm": 32.94181442260742, + "learning_rate": 1.7631683905604512e-06, + "loss": 7.414664268493652, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.33211464062333107, + "loss_total_v6": 0.463416513055563, + "step": 3195 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7370954954954955, + "grad_norm": 23.918119430541992, + "learning_rate": 1.7602746727682796e-06, + "loss": 6.8715596199035645, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.30009319074451923, + "loss_total_v6": 0.4294724650681019, + "step": 3196 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7373261261261261, + "grad_norm": 29.320423126220703, + "learning_rate": 1.757382824137731e-06, + "loss": 7.002518653869629, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30641656182706356, + "loss_total_v6": 0.4376574009656906, + "step": 3197 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7375567567567568, + "grad_norm": 24.98562240600586, + "learning_rate": 1.7544928463372567e-06, + "loss": 6.77794075012207, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.2947150245308876, + "loss_total_v6": 0.4236212857067585, + "step": 3198 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7377873873873874, + "grad_norm": 27.04138946533203, + "learning_rate": 1.7516047410342335e-06, + "loss": 6.847508430480957, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.300131157040596, + "loss_total_v6": 0.4279692992568016, + "step": 3199 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.738018018018018, + "grad_norm": 24.878664016723633, + "learning_rate": 1.7487185098949566e-06, + "loss": 6.763529300689697, + "loss_alignment": 0.251220703125, + "loss_alignment_w": 0.1256103515625, + "loss_sft": 0.2970492094755173, + "loss_total_v6": 0.4227205701172352, + "step": 3200 + }, + { + "epoch": 0.738018018018018, + "eval_loss": 0.5683953166007996, + "eval_loss_alignment": 0.2319915632134703, + "eval_loss_alignment_w": 0.11599578160673515, + "eval_loss_sft": 0.4523995570991545, + "eval_loss_total_v6": 0.5683953386633636, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.6962, + "eval_samples_per_second": 19.03, + "eval_steps_per_second": 2.388, + "step": 3200 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7382486486486487, + "grad_norm": 26.01656150817871, + "learning_rate": 1.7458341545846352e-06, + "loss": 6.743073463439941, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.2934819217771292, + "loss_total_v6": 0.42144210636615753, + "step": 3201 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7384792792792793, + "grad_norm": 22.75252914428711, + "learning_rate": 1.7429516767674044e-06, + "loss": 6.744168281555176, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.29498462565243244, + "loss_total_v6": 0.4215105026960373, + "step": 3202 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7387099099099099, + "grad_norm": 26.408382415771484, + "learning_rate": 1.7400710781063073e-06, + "loss": 6.890115737915039, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.3038469459861517, + "loss_total_v6": 0.43063222244381905, + "step": 3203 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7389405405405406, + "grad_norm": 24.59800148010254, + "learning_rate": 1.737192360263308e-06, + "loss": 6.838287353515625, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.29725074023008347, + "loss_total_v6": 0.42739295586943626, + "step": 3204 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7391711711711711, + "grad_norm": 22.618465423583984, + "learning_rate": 1.7343155248992888e-06, + "loss": 6.760838508605957, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.2957976255565882, + "loss_total_v6": 0.4225523956120014, + "step": 3205 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7394018018018018, + "grad_norm": 24.209701538085938, + "learning_rate": 1.731440573674037e-06, + "loss": 7.072441577911377, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.3134112600237131, + "loss_total_v6": 0.44202759861946106, + "step": 3206 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7396324324324325, + "grad_norm": 23.7619686126709, + "learning_rate": 1.7285675082462616e-06, + "loss": 7.06666374206543, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.31193630024790764, + "loss_total_v6": 0.44166650250554085, + "step": 3207 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.739863063063063, + "grad_norm": 24.440744400024414, + "learning_rate": 1.7256963302735752e-06, + "loss": 6.769459247589111, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.29456641897559166, + "loss_total_v6": 0.42309119924902916, + "step": 3208 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7400936936936937, + "grad_norm": 27.282922744750977, + "learning_rate": 1.7228270414125093e-06, + "loss": 6.561431884765625, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.27980994060635567, + "loss_total_v6": 0.41008948162198067, + "step": 3209 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7403243243243244, + "grad_norm": 25.640714645385742, + "learning_rate": 1.7199596433184995e-06, + "loss": 7.718343734741211, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3521627224981785, + "loss_total_v6": 0.4823964796960354, + "step": 3210 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7405549549549549, + "grad_norm": 25.119815826416016, + "learning_rate": 1.7170941376458932e-06, + "loss": 7.024487495422363, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.3099410980939865, + "loss_total_v6": 0.4390304610133171, + "step": 3211 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7407855855855856, + "grad_norm": 26.074573516845703, + "learning_rate": 1.7142305260479476e-06, + "loss": 6.87709903717041, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.30120234191417694, + "loss_total_v6": 0.42981869354844093, + "step": 3212 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7410162162162162, + "grad_norm": 27.36854362487793, + "learning_rate": 1.7113688101768238e-06, + "loss": 6.800459861755371, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2946576289832592, + "loss_total_v6": 0.4250287190079689, + "step": 3213 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7412468468468468, + "grad_norm": 23.40717887878418, + "learning_rate": 1.7085089916835924e-06, + "loss": 6.623687267303467, + "loss_alignment": 0.250244140625, + "loss_alignment_w": 0.1251220703125, + "loss_sft": 0.2885074280202389, + "loss_total_v6": 0.4139804542064667, + "step": 3214 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7414774774774775, + "grad_norm": 25.158302307128906, + "learning_rate": 1.7056510722182251e-06, + "loss": 6.482724189758301, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.27664550207555294, + "loss_total_v6": 0.4051702842116356, + "step": 3215 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.741708108108108, + "grad_norm": 26.4201602935791, + "learning_rate": 1.7027950534296028e-06, + "loss": 6.811124801635742, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2961634173989296, + "loss_total_v6": 0.4256952926516533, + "step": 3216 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7419387387387387, + "grad_norm": 24.259891510009766, + "learning_rate": 1.6999409369655046e-06, + "loss": 7.082118988037109, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.3154351729899645, + "loss_total_v6": 0.44263242557644844, + "step": 3217 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7421693693693694, + "grad_norm": 25.593568801879883, + "learning_rate": 1.6970887244726158e-06, + "loss": 6.28120231628418, + "loss_alignment": 0.2476806640625, + "loss_alignment_w": 0.12384033203125, + "loss_sft": 0.26865852251648903, + "loss_total_v6": 0.39257514476776123, + "step": 3218 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7424, + "grad_norm": 23.439462661743164, + "learning_rate": 1.6942384175965248e-06, + "loss": 6.946566581726074, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.30624594539403915, + "loss_total_v6": 0.43416038900613785, + "step": 3219 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7426306306306306, + "grad_norm": 25.60259246826172, + "learning_rate": 1.6913900179817145e-06, + "loss": 6.814742088317871, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.29790015891194344, + "loss_total_v6": 0.42592139542102814, + "step": 3220 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7428612612612613, + "grad_norm": 24.198665618896484, + "learning_rate": 1.6885435272715721e-06, + "loss": 7.108007431030273, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.3151000812649727, + "loss_total_v6": 0.4442504830658436, + "step": 3221 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7430918918918918, + "grad_norm": 26.832136154174805, + "learning_rate": 1.6856989471083828e-06, + "loss": 6.825473785400391, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.2954428307712078, + "loss_total_v6": 0.426592119038105, + "step": 3222 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7433225225225225, + "grad_norm": 27.051647186279297, + "learning_rate": 1.6828562791333297e-06, + "loss": 6.814212799072266, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.29863000102341175, + "loss_total_v6": 0.4258882999420166, + "step": 3223 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7435531531531532, + "grad_norm": 23.898605346679688, + "learning_rate": 1.6800155249864897e-06, + "loss": 7.043752193450928, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.3127625770866871, + "loss_total_v6": 0.4402345083653927, + "step": 3224 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7437837837837837, + "grad_norm": 22.58234977722168, + "learning_rate": 1.6771766863068389e-06, + "loss": 7.102667808532715, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3113026078790426, + "loss_total_v6": 0.4439167231321335, + "step": 3225 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7440144144144144, + "grad_norm": 30.3823299407959, + "learning_rate": 1.674339764732249e-06, + "loss": 7.214608669281006, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.320526709780097, + "loss_total_v6": 0.45091304928064346, + "step": 3226 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7442450450450451, + "grad_norm": 25.92885398864746, + "learning_rate": 1.67150476189948e-06, + "loss": 7.3208770751953125, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.3260698691010475, + "loss_total_v6": 0.4575548395514488, + "step": 3227 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7444756756756756, + "grad_norm": 24.44870948791504, + "learning_rate": 1.668671679444192e-06, + "loss": 6.857372283935547, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.3007781468331814, + "loss_total_v6": 0.4285857677459717, + "step": 3228 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7447063063063063, + "grad_norm": 23.53548812866211, + "learning_rate": 1.6658405190009291e-06, + "loss": 6.561175346374512, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.281792838126421, + "loss_total_v6": 0.4100734815001488, + "step": 3229 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.744936936936937, + "grad_norm": 27.131399154663086, + "learning_rate": 1.6630112822031374e-06, + "loss": 6.887389183044434, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3002127669751644, + "loss_total_v6": 0.4304618090391159, + "step": 3230 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7451675675675675, + "grad_norm": 24.118114471435547, + "learning_rate": 1.6601839706831423e-06, + "loss": 7.4717116355896, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3349019028246403, + "loss_total_v6": 0.4669819697737694, + "step": 3231 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7453981981981982, + "grad_norm": 24.9405574798584, + "learning_rate": 1.6573585860721648e-06, + "loss": 7.2729902267456055, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.32483167573809624, + "loss_total_v6": 0.45456190034747124, + "step": 3232 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7456288288288289, + "grad_norm": 25.355083465576172, + "learning_rate": 1.6545351300003131e-06, + "loss": 7.141098976135254, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.31782443821430206, + "loss_total_v6": 0.44631869718432426, + "step": 3233 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7458594594594594, + "grad_norm": 23.807640075683594, + "learning_rate": 1.6517136040965798e-06, + "loss": 7.117070198059082, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3151019550859928, + "loss_total_v6": 0.4448169097304344, + "step": 3234 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7460900900900901, + "grad_norm": 25.647647857666016, + "learning_rate": 1.648894009988849e-06, + "loss": 6.938172340393066, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.30309683457016945, + "loss_total_v6": 0.43363577499985695, + "step": 3235 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7463207207207208, + "grad_norm": 26.091157913208008, + "learning_rate": 1.646076349303884e-06, + "loss": 6.77277946472168, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.2932480685412884, + "loss_total_v6": 0.4232987239956856, + "step": 3236 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7465513513513513, + "grad_norm": 25.91142463684082, + "learning_rate": 1.643260623667337e-06, + "loss": 6.667115211486816, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.28560642898082733, + "loss_total_v6": 0.4166946969926357, + "step": 3237 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.746781981981982, + "grad_norm": 24.66263198852539, + "learning_rate": 1.6404468347037438e-06, + "loss": 6.4102582931518555, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.2727419715374708, + "loss_total_v6": 0.40064115449786186, + "step": 3238 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7470126126126126, + "grad_norm": 26.007238388061523, + "learning_rate": 1.6376349840365214e-06, + "loss": 6.982743263244629, + "loss_alignment": 0.2535400390625, + "loss_alignment_w": 0.12677001953125, + "loss_sft": 0.3096056915819645, + "loss_total_v6": 0.4364214763045311, + "step": 3239 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7472432432432432, + "grad_norm": 23.747127532958984, + "learning_rate": 1.63482507328797e-06, + "loss": 6.907940864562988, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.30393869057297707, + "loss_total_v6": 0.43174631148576736, + "step": 3240 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7474738738738739, + "grad_norm": 25.21896743774414, + "learning_rate": 1.6320171040792664e-06, + "loss": 6.739068984985352, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.29410136491060257, + "loss_total_v6": 0.4211918041110039, + "step": 3241 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7477045045045045, + "grad_norm": 26.381010055541992, + "learning_rate": 1.6292110780304743e-06, + "loss": 6.801966667175293, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.296292956918478, + "loss_total_v6": 0.4251229055225849, + "step": 3242 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7479351351351351, + "grad_norm": 25.085599899291992, + "learning_rate": 1.6264069967605278e-06, + "loss": 6.930422306060791, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.30601516366004944, + "loss_total_v6": 0.43315138295292854, + "step": 3243 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7481657657657658, + "grad_norm": 24.088266372680664, + "learning_rate": 1.6236048618872458e-06, + "loss": 6.597095489501953, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.28496863692998886, + "loss_total_v6": 0.41231848299503326, + "step": 3244 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7483963963963964, + "grad_norm": 24.842655181884766, + "learning_rate": 1.6208046750273233e-06, + "loss": 7.077319145202637, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3134872280061245, + "loss_total_v6": 0.4423324353992939, + "step": 3245 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.748627027027027, + "grad_norm": 24.65204429626465, + "learning_rate": 1.6180064377963261e-06, + "loss": 6.818658828735352, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.2988315671682358, + "loss_total_v6": 0.42616618052124977, + "step": 3246 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7488576576576577, + "grad_norm": 22.709251403808594, + "learning_rate": 1.615210151808701e-06, + "loss": 6.872526168823242, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.30268657580018044, + "loss_total_v6": 0.42953288927674294, + "step": 3247 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7490882882882883, + "grad_norm": 21.484981536865234, + "learning_rate": 1.6124158186777678e-06, + "loss": 7.23299503326416, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.31966165266931057, + "loss_total_v6": 0.4520621746778488, + "step": 3248 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7493189189189189, + "grad_norm": 23.461687088012695, + "learning_rate": 1.609623440015719e-06, + "loss": 6.566800117492676, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.28121357038617134, + "loss_total_v6": 0.41042498871684074, + "step": 3249 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7495495495495496, + "grad_norm": 23.5541934967041, + "learning_rate": 1.6068330174336172e-06, + "loss": 6.453468322753906, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.2748169656842947, + "loss_total_v6": 0.40334175154566765, + "step": 3250 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7497801801801802, + "grad_norm": 24.79882049560547, + "learning_rate": 1.6040445525413994e-06, + "loss": 6.717062473297119, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.28852979466319084, + "loss_total_v6": 0.41981639713048935, + "step": 3251 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7500108108108108, + "grad_norm": 27.605249404907227, + "learning_rate": 1.6012580469478744e-06, + "loss": 6.73869514465332, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.294688330963254, + "loss_total_v6": 0.421168427914381, + "step": 3252 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7502414414414414, + "grad_norm": 23.509029388427734, + "learning_rate": 1.598473502260715e-06, + "loss": 6.457107067108154, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.2726335283368826, + "loss_total_v6": 0.40356918796896935, + "step": 3253 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7504720720720721, + "grad_norm": 24.49693489074707, + "learning_rate": 1.5956909200864674e-06, + "loss": 6.990309715270996, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.3088883701711893, + "loss_total_v6": 0.43689435347914696, + "step": 3254 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7507027027027027, + "grad_norm": 26.20893096923828, + "learning_rate": 1.5929103020305441e-06, + "loss": 7.029506683349609, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.3118874989449978, + "loss_total_v6": 0.4393441565334797, + "step": 3255 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7509333333333333, + "grad_norm": 24.33331298828125, + "learning_rate": 1.5901316496972262e-06, + "loss": 6.708399772644043, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2911774553358555, + "loss_total_v6": 0.419274989515543, + "step": 3256 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.751163963963964, + "grad_norm": 26.793132781982422, + "learning_rate": 1.5873549646896563e-06, + "loss": 6.972490310668945, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3064776547253132, + "loss_total_v6": 0.4357806332409382, + "step": 3257 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7513945945945946, + "grad_norm": 23.205150604248047, + "learning_rate": 1.5845802486098461e-06, + "loss": 7.028325080871582, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3088076673448086, + "loss_total_v6": 0.4392703250050545, + "step": 3258 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7516252252252252, + "grad_norm": 24.21146011352539, + "learning_rate": 1.5818075030586704e-06, + "loss": 6.579094886779785, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.281646316871047, + "loss_total_v6": 0.4111934304237366, + "step": 3259 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7518558558558559, + "grad_norm": 27.285646438598633, + "learning_rate": 1.5790367296358644e-06, + "loss": 7.240424156188965, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.32270473800599575, + "loss_total_v6": 0.4525265134871006, + "step": 3260 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7520864864864865, + "grad_norm": 26.117021560668945, + "learning_rate": 1.5762679299400307e-06, + "loss": 7.104720115661621, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3131551332771778, + "loss_total_v6": 0.4440450072288513, + "step": 3261 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7523171171171171, + "grad_norm": 27.00783348083496, + "learning_rate": 1.573501105568626e-06, + "loss": 6.997419834136963, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3071202263236046, + "loss_total_v6": 0.4373387359082699, + "step": 3262 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7525477477477478, + "grad_norm": 26.389049530029297, + "learning_rate": 1.5707362581179746e-06, + "loss": 7.008866310119629, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.3082323633134365, + "loss_total_v6": 0.4380541481077671, + "step": 3263 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7527783783783784, + "grad_norm": 24.10106658935547, + "learning_rate": 1.5679733891832556e-06, + "loss": 6.7958083152771, + "loss_alignment": 0.251708984375, + "loss_alignment_w": 0.1258544921875, + "loss_sft": 0.2987156771123409, + "loss_total_v6": 0.424738023430109, + "step": 3264 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.753009009009009, + "grad_norm": 25.334035873413086, + "learning_rate": 1.5652125003585083e-06, + "loss": 7.1974616050720215, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.32110296189785004, + "loss_total_v6": 0.44984135404229164, + "step": 3265 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7532396396396397, + "grad_norm": 22.608671188354492, + "learning_rate": 1.5624535932366313e-06, + "loss": 6.871740818023682, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.3014320395886898, + "loss_total_v6": 0.4294837936758995, + "step": 3266 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7534702702702702, + "grad_norm": 24.949604034423828, + "learning_rate": 1.5596966694093746e-06, + "loss": 6.911029815673828, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.303246708586812, + "loss_total_v6": 0.43193935230374336, + "step": 3267 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7537009009009009, + "grad_norm": 23.11838722229004, + "learning_rate": 1.55694173046735e-06, + "loss": 6.378731727600098, + "loss_alignment": 0.246826171875, + "loss_alignment_w": 0.1234130859375, + "loss_sft": 0.27522713877260685, + "loss_total_v6": 0.3986707329750061, + "step": 3268 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7539315315315316, + "grad_norm": 28.37371826171875, + "learning_rate": 1.5541887780000187e-06, + "loss": 6.719918251037598, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.29363685473799706, + "loss_total_v6": 0.41999489441514015, + "step": 3269 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7541621621621621, + "grad_norm": 25.772409439086914, + "learning_rate": 1.5514378135956998e-06, + "loss": 6.730254173278809, + "loss_alignment": 0.2520751953125, + "loss_alignment_w": 0.12603759765625, + "loss_sft": 0.29452697932720184, + "loss_total_v6": 0.42064087837934494, + "step": 3270 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7543927927927928, + "grad_norm": 25.27503204345703, + "learning_rate": 1.548688838841565e-06, + "loss": 6.832752227783203, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2974083460867405, + "loss_total_v6": 0.4270470067858696, + "step": 3271 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7546234234234235, + "grad_norm": 27.729841232299805, + "learning_rate": 1.5459418553236343e-06, + "loss": 7.011168479919434, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.3115653470158577, + "loss_total_v6": 0.438198022544384, + "step": 3272 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.754854054054054, + "grad_norm": 23.325685501098633, + "learning_rate": 1.5431968646267836e-06, + "loss": 7.0045166015625, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3072280939668417, + "loss_total_v6": 0.43778228759765625, + "step": 3273 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7550846846846847, + "grad_norm": 24.894895553588867, + "learning_rate": 1.5404538683347365e-06, + "loss": 7.058443069458008, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.3120480813086033, + "loss_total_v6": 0.4411526881158352, + "step": 3274 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7553153153153153, + "grad_norm": 24.778114318847656, + "learning_rate": 1.537712868030068e-06, + "loss": 7.032176971435547, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.31251218169927597, + "loss_total_v6": 0.4395110756158829, + "step": 3275 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7555459459459459, + "grad_norm": 24.345346450805664, + "learning_rate": 1.534973865294197e-06, + "loss": 6.888165473937988, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2993305455893278, + "loss_total_v6": 0.43051035329699516, + "step": 3276 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7557765765765766, + "grad_norm": 28.76556968688965, + "learning_rate": 1.5322368617073952e-06, + "loss": 6.787650108337402, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.29739708453416824, + "loss_total_v6": 0.42422812432050705, + "step": 3277 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7560072072072072, + "grad_norm": 22.956558227539062, + "learning_rate": 1.5295018588487786e-06, + "loss": 6.870156288146973, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.30011231265962124, + "loss_total_v6": 0.429384782910347, + "step": 3278 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7562378378378378, + "grad_norm": 27.374608993530273, + "learning_rate": 1.5267688582963074e-06, + "loss": 6.674947261810303, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.2911008447408676, + "loss_total_v6": 0.4171842224895954, + "step": 3279 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7564684684684685, + "grad_norm": 26.813688278198242, + "learning_rate": 1.5240378616267887e-06, + "loss": 6.7428765296936035, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.292248846963048, + "loss_total_v6": 0.4214297793805599, + "step": 3280 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.756699099099099, + "grad_norm": 25.12472915649414, + "learning_rate": 1.521308870415873e-06, + "loss": 7.11630392074585, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.3128262273967266, + "loss_total_v6": 0.4447689801454544, + "step": 3281 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7569297297297297, + "grad_norm": 22.978681564331055, + "learning_rate": 1.5185818862380547e-06, + "loss": 6.829136848449707, + "loss_alignment": 0.2518310546875, + "loss_alignment_w": 0.12591552734375, + "loss_sft": 0.30105811543762684, + "loss_total_v6": 0.4268210493028164, + "step": 3282 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7571603603603604, + "grad_norm": 26.32447052001953, + "learning_rate": 1.5158569106666665e-06, + "loss": 6.849601745605469, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2961726039648056, + "loss_total_v6": 0.4281000904738903, + "step": 3283 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7573909909909909, + "grad_norm": 26.091520309448242, + "learning_rate": 1.5131339452738864e-06, + "loss": 6.893219947814941, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.3009739499539137, + "loss_total_v6": 0.43082624673843384, + "step": 3284 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7576216216216216, + "grad_norm": 25.443527221679688, + "learning_rate": 1.5104129916307325e-06, + "loss": 6.896010398864746, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.30113310366868973, + "loss_total_v6": 0.43100063875317574, + "step": 3285 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7578522522522523, + "grad_norm": 26.33587074279785, + "learning_rate": 1.507694051307057e-06, + "loss": 6.890962600708008, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3002225123345852, + "loss_total_v6": 0.4306851737201214, + "step": 3286 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7580828828828828, + "grad_norm": 23.58601188659668, + "learning_rate": 1.5049771258715585e-06, + "loss": 7.167810440063477, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.31908190064132214, + "loss_total_v6": 0.4479881599545479, + "step": 3287 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7583135135135135, + "grad_norm": 25.814237594604492, + "learning_rate": 1.502262216891765e-06, + "loss": 6.714786052703857, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.2921716943383217, + "loss_total_v6": 0.4196741208434105, + "step": 3288 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7585441441441442, + "grad_norm": 24.62942123413086, + "learning_rate": 1.4995493259340466e-06, + "loss": 6.8235368728637695, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.3003266528248787, + "loss_total_v6": 0.4264710582792759, + "step": 3289 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7587747747747747, + "grad_norm": 25.714027404785156, + "learning_rate": 1.4968384545636078e-06, + "loss": 7.5456085205078125, + "loss_alignment": 0.2635498046875, + "loss_alignment_w": 0.13177490234375, + "loss_sft": 0.33977984450757504, + "loss_total_v6": 0.4716005362570286, + "step": 3290 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7590054054054054, + "grad_norm": 30.138540267944336, + "learning_rate": 1.4941296043444869e-06, + "loss": 6.89154577255249, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.30309710279107094, + "loss_total_v6": 0.43072161823511124, + "step": 3291 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7592360360360361, + "grad_norm": 23.57328987121582, + "learning_rate": 1.4914227768395594e-06, + "loss": 6.91246223449707, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.306220144033432, + "loss_total_v6": 0.4320288822054863, + "step": 3292 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7594666666666666, + "grad_norm": 27.570051193237305, + "learning_rate": 1.488717973610528e-06, + "loss": 6.893470764160156, + "loss_alignment": 0.249755859375, + "loss_alignment_w": 0.1248779296875, + "loss_sft": 0.30588769540190697, + "loss_total_v6": 0.43084194138646126, + "step": 3293 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7596972972972973, + "grad_norm": 22.633028030395508, + "learning_rate": 1.4860151962179326e-06, + "loss": 6.862497806549072, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3010985106229782, + "loss_total_v6": 0.4289061203598976, + "step": 3294 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.759927927927928, + "grad_norm": 30.74736213684082, + "learning_rate": 1.4833144462211401e-06, + "loss": 6.759258270263672, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.29536315612494946, + "loss_total_v6": 0.4224536120891571, + "step": 3295 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7601585585585585, + "grad_norm": 26.080997467041016, + "learning_rate": 1.4806157251783515e-06, + "loss": 7.262709140777588, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.32724086567759514, + "loss_total_v6": 0.45391932129859924, + "step": 3296 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7603891891891892, + "grad_norm": 22.753177642822266, + "learning_rate": 1.477919034646597e-06, + "loss": 6.755508899688721, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.2943964023143053, + "loss_total_v6": 0.42221929505467415, + "step": 3297 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7606198198198199, + "grad_norm": 31.886180877685547, + "learning_rate": 1.4752243761817313e-06, + "loss": 6.819771766662598, + "loss_alignment": 0.248779296875, + "loss_alignment_w": 0.1243896484375, + "loss_sft": 0.30186133831739426, + "loss_total_v6": 0.42623572051525116, + "step": 3298 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7608504504504504, + "grad_norm": 23.82119369506836, + "learning_rate": 1.4725317513384402e-06, + "loss": 6.682916641235352, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.29001202806830406, + "loss_total_v6": 0.41768230497837067, + "step": 3299 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7610810810810811, + "grad_norm": 26.596969604492188, + "learning_rate": 1.4698411616702358e-06, + "loss": 6.83680534362793, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.2971276231110096, + "loss_total_v6": 0.4273003414273262, + "step": 3300 + }, + { + "epoch": 0.7610810810810811, + "eval_loss": 0.5672104954719543, + "eval_loss_alignment": 0.2325824058219178, + "eval_loss_alignment_w": 0.1162912029109589, + "eval_loss_sft": 0.450919260343188, + "eval_loss_total_v6": 0.5672104627608436, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.2869, + "eval_samples_per_second": 19.116, + "eval_steps_per_second": 2.399, + "step": 3300 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7613117117117117, + "grad_norm": 27.579694747924805, + "learning_rate": 1.4671526087294569e-06, + "loss": 6.85819149017334, + "loss_alignment": 0.252197265625, + "loss_alignment_w": 0.1260986328125, + "loss_sft": 0.30235523357987404, + "loss_total_v6": 0.42863697558641434, + "step": 3301 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7615423423423423, + "grad_norm": 24.605913162231445, + "learning_rate": 1.4644660940672628e-06, + "loss": 6.638450622558594, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2851271592080593, + "loss_total_v6": 0.4149031639099121, + "step": 3302 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.761772972972973, + "grad_norm": 22.456375122070312, + "learning_rate": 1.4617816192336414e-06, + "loss": 6.959887504577637, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.30581206642091274, + "loss_total_v6": 0.4349929764866829, + "step": 3303 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7620036036036036, + "grad_norm": 26.463485717773438, + "learning_rate": 1.459099185777404e-06, + "loss": 6.707391738891602, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.2913127914071083, + "loss_total_v6": 0.4192119836807251, + "step": 3304 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7622342342342342, + "grad_norm": 24.84140396118164, + "learning_rate": 1.456418795246179e-06, + "loss": 6.827231407165527, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.2977346945554018, + "loss_total_v6": 0.42670196294784546, + "step": 3305 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7624648648648649, + "grad_norm": 24.165210723876953, + "learning_rate": 1.453740449186421e-06, + "loss": 6.801239013671875, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.29611016623675823, + "loss_total_v6": 0.4250774420797825, + "step": 3306 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7626954954954955, + "grad_norm": 24.628374099731445, + "learning_rate": 1.4510641491434036e-06, + "loss": 7.13028621673584, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3153633140027523, + "loss_total_v6": 0.4456428699195385, + "step": 3307 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7629261261261261, + "grad_norm": 25.371530532836914, + "learning_rate": 1.448389896661221e-06, + "loss": 6.727447032928467, + "loss_alignment": 0.2659912109375, + "loss_alignment_w": 0.13299560546875, + "loss_sft": 0.2878970764577389, + "loss_total_v6": 0.4204654432833195, + "step": 3308 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7631567567567568, + "grad_norm": 23.414796829223633, + "learning_rate": 1.445717693282782e-06, + "loss": 6.813619136810303, + "loss_alignment": 0.2535400390625, + "loss_alignment_w": 0.12677001953125, + "loss_sft": 0.2989438399672508, + "loss_total_v6": 0.4258511923253536, + "step": 3309 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7633873873873874, + "grad_norm": 26.959829330444336, + "learning_rate": 1.4430475405498178e-06, + "loss": 6.996172904968262, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.30536384880542755, + "loss_total_v6": 0.43726081773638725, + "step": 3310 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.763618018018018, + "grad_norm": 24.2748966217041, + "learning_rate": 1.4403794400028765e-06, + "loss": 7.106956481933594, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3131422698497772, + "loss_total_v6": 0.4441847614943981, + "step": 3311 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7638486486486487, + "grad_norm": 25.802637100219727, + "learning_rate": 1.437713393181317e-06, + "loss": 6.785688877105713, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.29570285603404045, + "loss_total_v6": 0.42410556599497795, + "step": 3312 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7640792792792793, + "grad_norm": 27.230724334716797, + "learning_rate": 1.4350494016233197e-06, + "loss": 6.875950336456299, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.29670549742877483, + "loss_total_v6": 0.4297468885779381, + "step": 3313 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7643099099099099, + "grad_norm": 24.27099609375, + "learning_rate": 1.4323874668658732e-06, + "loss": 6.978809356689453, + "loss_alignment": 0.2532958984375, + "loss_alignment_w": 0.12664794921875, + "loss_sft": 0.3091461695730686, + "loss_total_v6": 0.4361755922436714, + "step": 3314 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7645405405405405, + "grad_norm": 26.220829010009766, + "learning_rate": 1.4297275904447844e-06, + "loss": 6.788154125213623, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.29524656757712364, + "loss_total_v6": 0.42425962164998055, + "step": 3315 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7647711711711712, + "grad_norm": 23.5987491607666, + "learning_rate": 1.4270697738946703e-06, + "loss": 6.578653335571289, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.2803064286708832, + "loss_total_v6": 0.4111658111214638, + "step": 3316 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7650018018018018, + "grad_norm": 23.385509490966797, + "learning_rate": 1.4244140187489602e-06, + "loss": 7.160370349884033, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.321226155385375, + "loss_total_v6": 0.4475231431424618, + "step": 3317 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7652324324324324, + "grad_norm": 26.951452255249023, + "learning_rate": 1.421760326539895e-06, + "loss": 7.503604888916016, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.33660533651709557, + "loss_total_v6": 0.46897532790899277, + "step": 3318 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7654630630630631, + "grad_norm": 22.52552604675293, + "learning_rate": 1.4191086987985214e-06, + "loss": 6.765896797180176, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.2935350500047207, + "loss_total_v6": 0.4228685609996319, + "step": 3319 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7656936936936937, + "grad_norm": 25.455814361572266, + "learning_rate": 1.4164591370547003e-06, + "loss": 6.349359512329102, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.26634179800748825, + "loss_total_v6": 0.39683496579527855, + "step": 3320 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7659243243243243, + "grad_norm": 28.558197021484375, + "learning_rate": 1.4138116428370958e-06, + "loss": 7.239227294921875, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.3203563541173935, + "loss_total_v6": 0.4524516873061657, + "step": 3321 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.766154954954955, + "grad_norm": 24.322843551635742, + "learning_rate": 1.4111662176731823e-06, + "loss": 6.91032600402832, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.303813099861145, + "loss_total_v6": 0.4318953640758991, + "step": 3322 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7663855855855856, + "grad_norm": 23.6044979095459, + "learning_rate": 1.4085228630892406e-06, + "loss": 6.666454315185547, + "loss_alignment": 0.2515869140625, + "loss_alignment_w": 0.12579345703125, + "loss_sft": 0.2910887859761715, + "loss_total_v6": 0.4166533686220646, + "step": 3323 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7666162162162162, + "grad_norm": 24.202775955200195, + "learning_rate": 1.4058815806103542e-06, + "loss": 6.750790596008301, + "loss_alignment": 0.2506103515625, + "loss_alignment_w": 0.12530517578125, + "loss_sft": 0.2966497614979744, + "loss_total_v6": 0.4219244047999382, + "step": 3324 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7668468468468469, + "grad_norm": 26.42208480834961, + "learning_rate": 1.4032423717604133e-06, + "loss": 6.830142974853516, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.2972147911787033, + "loss_total_v6": 0.4268839471042156, + "step": 3325 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7670774774774775, + "grad_norm": 23.874958038330078, + "learning_rate": 1.400605238062111e-06, + "loss": 6.772927284240723, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.29444746673107147, + "loss_total_v6": 0.4233079329133034, + "step": 3326 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7673081081081081, + "grad_norm": 25.557117462158203, + "learning_rate": 1.397970181036945e-06, + "loss": 7.695810317993164, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.34845032170414925, + "loss_total_v6": 0.48098815977573395, + "step": 3327 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7675387387387388, + "grad_norm": 27.256315231323242, + "learning_rate": 1.395337202205211e-06, + "loss": 6.6877899169921875, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.2900266721844673, + "loss_total_v6": 0.4179868847131729, + "step": 3328 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7677693693693693, + "grad_norm": 23.415081024169922, + "learning_rate": 1.3927063030860076e-06, + "loss": 6.568663597106934, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.28099434077739716, + "loss_total_v6": 0.41054147481918335, + "step": 3329 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.768, + "grad_norm": 23.946243286132812, + "learning_rate": 1.3900774851972371e-06, + "loss": 6.978900909423828, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.30739711597561836, + "loss_total_v6": 0.43618129938840866, + "step": 3330 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7682306306306307, + "grad_norm": 26.707191467285156, + "learning_rate": 1.3874507500555939e-06, + "loss": 7.141709327697754, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3151465281844139, + "loss_total_v6": 0.4463568367063999, + "step": 3331 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7684612612612612, + "grad_norm": 23.08054542541504, + "learning_rate": 1.3848260991765754e-06, + "loss": 6.782723426818848, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.2963262256234884, + "loss_total_v6": 0.4239202179014683, + "step": 3332 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7686918918918919, + "grad_norm": 25.61229133605957, + "learning_rate": 1.3822035340744766e-06, + "loss": 7.037534236907959, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.31113799288868904, + "loss_total_v6": 0.43984589353203773, + "step": 3333 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7689225225225226, + "grad_norm": 24.656986236572266, + "learning_rate": 1.379583056262389e-06, + "loss": 7.091940402984619, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3129514791071415, + "loss_total_v6": 0.443246278911829, + "step": 3334 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7691531531531531, + "grad_norm": 26.44728660583496, + "learning_rate": 1.3769646672521964e-06, + "loss": 6.439804553985596, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.2728033512830734, + "loss_total_v6": 0.4024877957999706, + "step": 3335 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7693837837837838, + "grad_norm": 26.205379486083984, + "learning_rate": 1.374348368554581e-06, + "loss": 7.374303817749023, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.33299483731389046, + "loss_total_v6": 0.46089399978518486, + "step": 3336 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7696144144144144, + "grad_norm": 25.240835189819336, + "learning_rate": 1.3717341616790198e-06, + "loss": 7.184267044067383, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3199273608624935, + "loss_total_v6": 0.4490167051553726, + "step": 3337 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.769845045045045, + "grad_norm": 26.692914962768555, + "learning_rate": 1.3691220481337786e-06, + "loss": 7.28939151763916, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3266959451138973, + "loss_total_v6": 0.455586951225996, + "step": 3338 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7700756756756757, + "grad_norm": 25.334932327270508, + "learning_rate": 1.3665120294259204e-06, + "loss": 6.896916389465332, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.30295974016189575, + "loss_total_v6": 0.43105727061629295, + "step": 3339 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7703063063063063, + "grad_norm": 24.102197647094727, + "learning_rate": 1.363904107061294e-06, + "loss": 6.769386291503906, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.2937531303614378, + "loss_total_v6": 0.42308663949370384, + "step": 3340 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7705369369369369, + "grad_norm": 24.443328857421875, + "learning_rate": 1.3612982825445443e-06, + "loss": 6.931029319763184, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.30471030436456203, + "loss_total_v6": 0.4331893138587475, + "step": 3341 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7707675675675676, + "grad_norm": 27.051498413085938, + "learning_rate": 1.3586945573791028e-06, + "loss": 6.5013427734375, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.27729030698537827, + "loss_total_v6": 0.40633390843868256, + "step": 3342 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7709981981981981, + "grad_norm": 25.858104705810547, + "learning_rate": 1.3560929330671908e-06, + "loss": 6.971311569213867, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.3076246790587902, + "loss_total_v6": 0.4357069544494152, + "step": 3343 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7712288288288288, + "grad_norm": 28.4810848236084, + "learning_rate": 1.353493411109818e-06, + "loss": 6.938422203063965, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3027157336473465, + "loss_total_v6": 0.4336513951420784, + "step": 3344 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7714594594594595, + "grad_norm": 28.069965362548828, + "learning_rate": 1.3508959930067778e-06, + "loss": 6.752017021179199, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.29011934250593185, + "loss_total_v6": 0.42200105637311935, + "step": 3345 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.77169009009009, + "grad_norm": 24.786956787109375, + "learning_rate": 1.3483006802566546e-06, + "loss": 6.804934978485107, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.29692098312079906, + "loss_total_v6": 0.4253084361553192, + "step": 3346 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7719207207207207, + "grad_norm": 23.399799346923828, + "learning_rate": 1.3457074743568138e-06, + "loss": 7.120086669921875, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3169689290225506, + "loss_total_v6": 0.4450054243206978, + "step": 3347 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7721513513513514, + "grad_norm": 24.352272033691406, + "learning_rate": 1.3431163768034078e-06, + "loss": 6.994887828826904, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.30935762263834476, + "loss_total_v6": 0.4371804893016815, + "step": 3348 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7723819819819819, + "grad_norm": 23.600692749023438, + "learning_rate": 1.340527389091374e-06, + "loss": 6.811166763305664, + "loss_alignment": 0.25146484375, + "loss_alignment_w": 0.125732421875, + "loss_sft": 0.29988920129835606, + "loss_total_v6": 0.425697922706604, + "step": 3349 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7726126126126126, + "grad_norm": 22.945833206176758, + "learning_rate": 1.3379405127144273e-06, + "loss": 6.664477348327637, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.28902738355100155, + "loss_total_v6": 0.416529830545187, + "step": 3350 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7728432432432433, + "grad_norm": 24.116249084472656, + "learning_rate": 1.3353557491650698e-06, + "loss": 7.054109573364258, + "loss_alignment": 0.2532958984375, + "loss_alignment_w": 0.12664794921875, + "loss_sft": 0.3141423538327217, + "loss_total_v6": 0.4408818446099758, + "step": 3351 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7730738738738738, + "grad_norm": 23.424734115600586, + "learning_rate": 1.3327730999345818e-06, + "loss": 6.950685024261475, + "loss_alignment": 0.2628173828125, + "loss_alignment_w": 0.13140869140625, + "loss_sft": 0.30311593040823936, + "loss_total_v6": 0.43441781029105186, + "step": 3352 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7733045045045045, + "grad_norm": 27.315858840942383, + "learning_rate": 1.3301925665130262e-06, + "loss": 6.637964248657227, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.2871261965483427, + "loss_total_v6": 0.41487278416752815, + "step": 3353 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7735351351351352, + "grad_norm": 28.023794174194336, + "learning_rate": 1.3276141503892399e-06, + "loss": 6.949493408203125, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.3071308061480522, + "loss_total_v6": 0.4343433268368244, + "step": 3354 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7737657657657657, + "grad_norm": 26.443634033203125, + "learning_rate": 1.325037853050844e-06, + "loss": 7.3802642822265625, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.33272648602724075, + "loss_total_v6": 0.46126651018857956, + "step": 3355 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7739963963963964, + "grad_norm": 26.233257293701172, + "learning_rate": 1.3224636759842363e-06, + "loss": 7.000099182128906, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.31020212545990944, + "loss_total_v6": 0.43750619143247604, + "step": 3356 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7742270270270271, + "grad_norm": 25.98595428466797, + "learning_rate": 1.3198916206745871e-06, + "loss": 6.840453147888184, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.3006667383015156, + "loss_total_v6": 0.4275283068418503, + "step": 3357 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7744576576576576, + "grad_norm": 23.041996002197266, + "learning_rate": 1.317321688605846e-06, + "loss": 7.205066680908203, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.32102894969284534, + "loss_total_v6": 0.4503166675567627, + "step": 3358 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7746882882882883, + "grad_norm": 23.59676742553711, + "learning_rate": 1.3147538812607375e-06, + "loss": 7.177192211151123, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.3222469948232174, + "loss_total_v6": 0.4485745131969452, + "step": 3359 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.774918918918919, + "grad_norm": 26.193140029907227, + "learning_rate": 1.3121882001207614e-06, + "loss": 7.081453800201416, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.3099614791572094, + "loss_total_v6": 0.4425908736884594, + "step": 3360 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7751495495495495, + "grad_norm": 23.76968765258789, + "learning_rate": 1.309624646666186e-06, + "loss": 7.043584823608398, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.31223333440721035, + "loss_total_v6": 0.4402240552008152, + "step": 3361 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7753801801801802, + "grad_norm": 25.546669006347656, + "learning_rate": 1.3070632223760565e-06, + "loss": 6.728195667266846, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.29186539351940155, + "loss_total_v6": 0.42051224038004875, + "step": 3362 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7756108108108108, + "grad_norm": 24.683475494384766, + "learning_rate": 1.3045039287281896e-06, + "loss": 7.055802345275879, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.31284432858228683, + "loss_total_v6": 0.44098764657974243, + "step": 3363 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7758414414414414, + "grad_norm": 25.61260986328125, + "learning_rate": 1.3019467671991693e-06, + "loss": 7.023287773132324, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.30866066738963127, + "loss_total_v6": 0.43895547837018967, + "step": 3364 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7760720720720721, + "grad_norm": 25.081722259521484, + "learning_rate": 1.2993917392643535e-06, + "loss": 7.045581340789795, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.31113737262785435, + "loss_total_v6": 0.4403488114476204, + "step": 3365 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7763027027027027, + "grad_norm": 28.532312393188477, + "learning_rate": 1.2968388463978642e-06, + "loss": 7.178394794464111, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.32134558260440826, + "loss_total_v6": 0.44864968210458755, + "step": 3366 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7765333333333333, + "grad_norm": 25.252727508544922, + "learning_rate": 1.2942880900725973e-06, + "loss": 6.971469879150391, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.30655120871961117, + "loss_total_v6": 0.4357168488204479, + "step": 3367 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.776763963963964, + "grad_norm": 24.625141143798828, + "learning_rate": 1.2917394717602123e-06, + "loss": 6.6905364990234375, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.28897762671113014, + "loss_total_v6": 0.41815853118896484, + "step": 3368 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7769945945945946, + "grad_norm": 21.61939239501953, + "learning_rate": 1.2891929929311364e-06, + "loss": 6.725775718688965, + "loss_alignment": 0.2486572265625, + "loss_alignment_w": 0.12432861328125, + "loss_sft": 0.2958492487668991, + "loss_total_v6": 0.4203609712421894, + "step": 3369 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7772252252252252, + "grad_norm": 25.75123405456543, + "learning_rate": 1.2866486550545636e-06, + "loss": 7.20539665222168, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.3198899105191231, + "loss_total_v6": 0.4503373093903065, + "step": 3370 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7774558558558559, + "grad_norm": 24.625818252563477, + "learning_rate": 1.2841064595984493e-06, + "loss": 6.891059398651123, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3000912368297577, + "loss_total_v6": 0.4306912161409855, + "step": 3371 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7776864864864865, + "grad_norm": 28.747596740722656, + "learning_rate": 1.281566408029516e-06, + "loss": 7.398820400238037, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.3351527117192745, + "loss_total_v6": 0.4624262824654579, + "step": 3372 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7779171171171171, + "grad_norm": 26.40867042541504, + "learning_rate": 1.2790285018132464e-06, + "loss": 6.834496974945068, + "loss_alignment": 0.25244140625, + "loss_alignment_w": 0.126220703125, + "loss_sft": 0.30070650205016136, + "loss_total_v6": 0.4271560609340668, + "step": 3373 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7781477477477478, + "grad_norm": 24.747970581054688, + "learning_rate": 1.2764927424138884e-06, + "loss": 6.900991916656494, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.30149020440876484, + "loss_total_v6": 0.43131198361516, + "step": 3374 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7783783783783784, + "grad_norm": 26.213918685913086, + "learning_rate": 1.2739591312944516e-06, + "loss": 6.971109390258789, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3079020082950592, + "loss_total_v6": 0.4356943480670452, + "step": 3375 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.778609009009009, + "grad_norm": 25.137802124023438, + "learning_rate": 1.2714276699166994e-06, + "loss": 6.993792533874512, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3059322014451027, + "loss_total_v6": 0.4371120147407055, + "step": 3376 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7788396396396396, + "grad_norm": 24.198932647705078, + "learning_rate": 1.2688983597411669e-06, + "loss": 7.021635055541992, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3095034621655941, + "loss_total_v6": 0.4388522133231163, + "step": 3377 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7790702702702703, + "grad_norm": 25.4110050201416, + "learning_rate": 1.2663712022271363e-06, + "loss": 7.099994659423828, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.31534695625305176, + "loss_total_v6": 0.44374968111515045, + "step": 3378 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7793009009009009, + "grad_norm": 23.892690658569336, + "learning_rate": 1.2638461988326556e-06, + "loss": 7.073716163635254, + "loss_alignment": 0.250732421875, + "loss_alignment_w": 0.1253662109375, + "loss_sft": 0.31669526919722557, + "loss_total_v6": 0.44210726395249367, + "step": 3379 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7795315315315315, + "grad_norm": 24.267471313476562, + "learning_rate": 1.2613233510145251e-06, + "loss": 7.016968727111816, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3076096139848232, + "loss_total_v6": 0.4385605454444885, + "step": 3380 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7797621621621622, + "grad_norm": 27.08713150024414, + "learning_rate": 1.2588026602283038e-06, + "loss": 6.29673957824707, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.2646246999502182, + "loss_total_v6": 0.3935462161898613, + "step": 3381 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7799927927927928, + "grad_norm": 24.263275146484375, + "learning_rate": 1.2562841279283077e-06, + "loss": 6.709211349487305, + "loss_alignment": 0.2489013671875, + "loss_alignment_w": 0.12445068359375, + "loss_sft": 0.2950123567134142, + "loss_total_v6": 0.41932572796940804, + "step": 3382 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7802234234234234, + "grad_norm": 28.817354202270508, + "learning_rate": 1.2537677555676031e-06, + "loss": 6.8125152587890625, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.2968606688082218, + "loss_total_v6": 0.4257821887731552, + "step": 3383 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7804540540540541, + "grad_norm": 28.665803909301758, + "learning_rate": 1.251253544598014e-06, + "loss": 6.887815475463867, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.30106341280043125, + "loss_total_v6": 0.4304884634912014, + "step": 3384 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7806846846846847, + "grad_norm": 24.841751098632812, + "learning_rate": 1.2487414964701155e-06, + "loss": 6.690324783325195, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2882014252245426, + "loss_total_v6": 0.4181452840566635, + "step": 3385 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7809153153153153, + "grad_norm": 26.227779388427734, + "learning_rate": 1.2462316126332375e-06, + "loss": 6.486898422241211, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.2788900174200535, + "loss_total_v6": 0.4054311476647854, + "step": 3386 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.781145945945946, + "grad_norm": 23.731281280517578, + "learning_rate": 1.2437238945354562e-06, + "loss": 6.312045574188232, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.2682668846100569, + "loss_total_v6": 0.39450284466147423, + "step": 3387 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7813765765765766, + "grad_norm": 24.314594268798828, + "learning_rate": 1.241218343623602e-06, + "loss": 7.028870582580566, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.3121071420609951, + "loss_total_v6": 0.4393044002354145, + "step": 3388 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7816072072072072, + "grad_norm": 23.965219497680664, + "learning_rate": 1.2387149613432563e-06, + "loss": 6.6105241775512695, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.2836411725729704, + "loss_total_v6": 0.41315777599811554, + "step": 3389 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7818378378378379, + "grad_norm": 23.171606063842773, + "learning_rate": 1.2362137491387433e-06, + "loss": 6.993500232696533, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3060818053781986, + "loss_total_v6": 0.4370937682688236, + "step": 3390 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7820684684684684, + "grad_norm": 28.30997657775879, + "learning_rate": 1.233714708453143e-06, + "loss": 7.229310035705566, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3224526308476925, + "loss_total_v6": 0.4518318809568882, + "step": 3391 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7822990990990991, + "grad_norm": 27.5316219329834, + "learning_rate": 1.2312178407282748e-06, + "loss": 7.318889141082764, + "loss_alignment": 0.2672119140625, + "loss_alignment_w": 0.13360595703125, + "loss_sft": 0.32383985444903374, + "loss_total_v6": 0.45743056386709213, + "step": 3392 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7825297297297298, + "grad_norm": 26.075756072998047, + "learning_rate": 1.22872314740471e-06, + "loss": 6.289307594299316, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.26568610966205597, + "loss_total_v6": 0.39308173581957817, + "step": 3393 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7827603603603603, + "grad_norm": 30.17530632019043, + "learning_rate": 1.2262306299217635e-06, + "loss": 6.514533996582031, + "loss_alignment": 0.2486572265625, + "loss_alignment_w": 0.12432861328125, + "loss_sft": 0.28296707570552826, + "loss_total_v6": 0.40715836733579636, + "step": 3394 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.782990990990991, + "grad_norm": 25.826061248779297, + "learning_rate": 1.2237402897174949e-06, + "loss": 6.4311137199401855, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.27102421410381794, + "loss_total_v6": 0.4019446149468422, + "step": 3395 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7832216216216216, + "grad_norm": 27.14560317993164, + "learning_rate": 1.2212521282287093e-06, + "loss": 7.016779899597168, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.310802161693573, + "loss_total_v6": 0.4385487399995327, + "step": 3396 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7834522522522522, + "grad_norm": 27.52853775024414, + "learning_rate": 1.218766146890949e-06, + "loss": 7.089770793914795, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.3128311522305012, + "loss_total_v6": 0.4431106708943844, + "step": 3397 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7836828828828829, + "grad_norm": 29.126148223876953, + "learning_rate": 1.2162823471385072e-06, + "loss": 7.062004089355469, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.31170608289539814, + "loss_total_v6": 0.4413752667605877, + "step": 3398 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7839135135135135, + "grad_norm": 24.36956024169922, + "learning_rate": 1.2138007304044091e-06, + "loss": 7.163885116577148, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3182109110057354, + "loss_total_v6": 0.4477427862584591, + "step": 3399 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7841441441441441, + "grad_norm": 21.371540069580078, + "learning_rate": 1.2113212981204292e-06, + "loss": 7.180553436279297, + "loss_alignment": 0.2630615234375, + "loss_alignment_w": 0.13153076171875, + "loss_sft": 0.31691809743642807, + "loss_total_v6": 0.44878456369042397, + "step": 3400 + }, + { + "epoch": 0.7841441441441441, + "eval_loss": 0.5669552683830261, + "eval_loss_alignment": 0.23138957263127855, + "eval_loss_alignment_w": 0.11569478631563927, + "eval_loss_sft": 0.45126048855687656, + "eval_loss_total_v6": 0.5669552749660733, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.5344, + "eval_samples_per_second": 19.064, + "eval_steps_per_second": 2.393, + "step": 3400 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7843747747747748, + "grad_norm": 23.230566024780273, + "learning_rate": 1.2088440517170729e-06, + "loss": 6.713983535766602, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.28894768469035625, + "loss_total_v6": 0.4196239486336708, + "step": 3401 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7846054054054054, + "grad_norm": 26.619895935058594, + "learning_rate": 1.2063689926235932e-06, + "loss": 6.954512119293213, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.305323526263237, + "loss_total_v6": 0.434657022356987, + "step": 3402 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.784836036036036, + "grad_norm": 27.064510345458984, + "learning_rate": 1.2038961222679774e-06, + "loss": 6.891364574432373, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.30268901214003563, + "loss_total_v6": 0.4307102784514427, + "step": 3403 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7850666666666667, + "grad_norm": 19.754972457885742, + "learning_rate": 1.2014254420769466e-06, + "loss": 6.645565986633301, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.2872351035475731, + "loss_total_v6": 0.4153478816151619, + "step": 3404 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7852972972972972, + "grad_norm": 30.944690704345703, + "learning_rate": 1.1989569534759638e-06, + "loss": 7.10075569152832, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.3133498281240463, + "loss_total_v6": 0.4437972120940685, + "step": 3405 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7855279279279279, + "grad_norm": 25.046276092529297, + "learning_rate": 1.1964906578892232e-06, + "loss": 6.818538188934326, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.29859515465795994, + "loss_total_v6": 0.4261586330831051, + "step": 3406 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7857585585585586, + "grad_norm": 26.766149520874023, + "learning_rate": 1.1940265567396564e-06, + "loss": 6.408078193664551, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2722395081073046, + "loss_total_v6": 0.4005048871040344, + "step": 3407 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7859891891891891, + "grad_norm": 26.108003616333008, + "learning_rate": 1.1915646514489293e-06, + "loss": 7.019683837890625, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.31255529448390007, + "loss_total_v6": 0.43873023614287376, + "step": 3408 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7862198198198198, + "grad_norm": 27.680828094482422, + "learning_rate": 1.1891049434374375e-06, + "loss": 6.9026288986206055, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3018671665340662, + "loss_total_v6": 0.43141429126262665, + "step": 3409 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7864504504504505, + "grad_norm": 32.22554016113281, + "learning_rate": 1.1866474341243124e-06, + "loss": 6.976936340332031, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3075947631150484, + "loss_total_v6": 0.43605851754546165, + "step": 3410 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.786681081081081, + "grad_norm": 25.412220001220703, + "learning_rate": 1.1841921249274158e-06, + "loss": 6.6465301513671875, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.28728009946644306, + "loss_total_v6": 0.4154081456363201, + "step": 3411 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7869117117117117, + "grad_norm": 22.807491302490234, + "learning_rate": 1.1817390172633402e-06, + "loss": 6.77998161315918, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.2945526912808418, + "loss_total_v6": 0.4237488619983196, + "step": 3412 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7871423423423424, + "grad_norm": 27.050296783447266, + "learning_rate": 1.1792881125474064e-06, + "loss": 7.571438312530518, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.34504107758402824, + "loss_total_v6": 0.47321490198373795, + "step": 3413 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7873729729729729, + "grad_norm": 23.02647590637207, + "learning_rate": 1.176839412193666e-06, + "loss": 6.960391044616699, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3024560511112213, + "loss_total_v6": 0.4350244291126728, + "step": 3414 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7876036036036036, + "grad_norm": 34.89702606201172, + "learning_rate": 1.1743929176149e-06, + "loss": 6.736944198608398, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.293251384049654, + "loss_total_v6": 0.4210589937865734, + "step": 3415 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7878342342342343, + "grad_norm": 22.605289459228516, + "learning_rate": 1.1719486302226119e-06, + "loss": 6.6486358642578125, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.28777787275612354, + "loss_total_v6": 0.4155397154390812, + "step": 3416 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7880648648648648, + "grad_norm": 28.234878540039062, + "learning_rate": 1.1695065514270375e-06, + "loss": 7.441417694091797, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.3365943320095539, + "loss_total_v6": 0.4650885947048664, + "step": 3417 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7882954954954955, + "grad_norm": 24.399168014526367, + "learning_rate": 1.1670666826371336e-06, + "loss": 7.0785675048828125, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.31277178041636944, + "loss_total_v6": 0.4424104504287243, + "step": 3418 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7885261261261262, + "grad_norm": 24.3345947265625, + "learning_rate": 1.1646290252605852e-06, + "loss": 6.90297794342041, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.30204159021377563, + "loss_total_v6": 0.43143612518906593, + "step": 3419 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7887567567567567, + "grad_norm": 24.73907470703125, + "learning_rate": 1.1621935807038004e-06, + "loss": 6.7236199378967285, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.29408182948827744, + "loss_total_v6": 0.42022623121738434, + "step": 3420 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7889873873873874, + "grad_norm": 24.310840606689453, + "learning_rate": 1.1597603503719102e-06, + "loss": 6.709390640258789, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.2915292698889971, + "loss_total_v6": 0.4193368963897228, + "step": 3421 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.789218018018018, + "grad_norm": 25.779319763183594, + "learning_rate": 1.15732933566877e-06, + "loss": 6.873428821563721, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.3017053920775652, + "loss_total_v6": 0.42958930879831314, + "step": 3422 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7894486486486486, + "grad_norm": 32.24707794189453, + "learning_rate": 1.154900537996952e-06, + "loss": 6.63125467300415, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.2839907817542553, + "loss_total_v6": 0.4144534207880497, + "step": 3423 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7896792792792793, + "grad_norm": 25.441953659057617, + "learning_rate": 1.152473958757756e-06, + "loss": 6.81606388092041, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.2984252870082855, + "loss_total_v6": 0.42600400373339653, + "step": 3424 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.78990990990991, + "grad_norm": 25.403221130371094, + "learning_rate": 1.1500495993511962e-06, + "loss": 7.070080757141113, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.31262283585965633, + "loss_total_v6": 0.4418800361454487, + "step": 3425 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7901405405405405, + "grad_norm": 22.377634048461914, + "learning_rate": 1.1476274611760096e-06, + "loss": 7.220751762390137, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.32144469767808914, + "loss_total_v6": 0.45129700005054474, + "step": 3426 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7903711711711712, + "grad_norm": 23.94342613220215, + "learning_rate": 1.1452075456296475e-06, + "loss": 6.830993175506592, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.2965659648180008, + "loss_total_v6": 0.4269370697438717, + "step": 3427 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7906018018018018, + "grad_norm": 22.6904239654541, + "learning_rate": 1.1427898541082855e-06, + "loss": 7.002684593200684, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3078002445399761, + "loss_total_v6": 0.4376677833497524, + "step": 3428 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7908324324324324, + "grad_norm": 24.559690475463867, + "learning_rate": 1.1403743880068124e-06, + "loss": 6.813285827636719, + "loss_alignment": 0.24951171875, + "loss_alignment_w": 0.124755859375, + "loss_sft": 0.3011660762131214, + "loss_total_v6": 0.4258303865790367, + "step": 3429 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7910630630630631, + "grad_norm": 26.69942283630371, + "learning_rate": 1.1379611487188297e-06, + "loss": 6.656367778778076, + "loss_alignment": 0.2501220703125, + "loss_alignment_w": 0.12506103515625, + "loss_sft": 0.2907788585871458, + "loss_total_v6": 0.41602298244833946, + "step": 3430 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7912936936936937, + "grad_norm": 28.275938034057617, + "learning_rate": 1.13555013763666e-06, + "loss": 6.783046245574951, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.29564449936151505, + "loss_total_v6": 0.42394039407372475, + "step": 3431 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7915243243243243, + "grad_norm": 26.668664932250977, + "learning_rate": 1.133141356151336e-06, + "loss": 6.816166877746582, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2951205261051655, + "loss_total_v6": 0.4260104112327099, + "step": 3432 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.791754954954955, + "grad_norm": 28.025957107543945, + "learning_rate": 1.1307348056526069e-06, + "loss": 6.805776119232178, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.2966531291604042, + "loss_total_v6": 0.425361018627882, + "step": 3433 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7919855855855856, + "grad_norm": 24.170928955078125, + "learning_rate": 1.1283304875289335e-06, + "loss": 6.737058162689209, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.2938536051660776, + "loss_total_v6": 0.42106615006923676, + "step": 3434 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7922162162162162, + "grad_norm": 25.620746612548828, + "learning_rate": 1.1259284031674877e-06, + "loss": 7.409142971038818, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.33485183864831924, + "loss_total_v6": 0.46307143568992615, + "step": 3435 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7924468468468469, + "grad_norm": 23.826513290405273, + "learning_rate": 1.123528553954154e-06, + "loss": 7.120260715484619, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.3164152204990387, + "loss_total_v6": 0.4450163021683693, + "step": 3436 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7926774774774775, + "grad_norm": 25.074735641479492, + "learning_rate": 1.1211309412735278e-06, + "loss": 6.944548606872559, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.3074015863239765, + "loss_total_v6": 0.4340342730283737, + "step": 3437 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7929081081081081, + "grad_norm": 26.35997200012207, + "learning_rate": 1.1187355665089133e-06, + "loss": 6.418225288391113, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.2737281993031502, + "loss_total_v6": 0.4011390767991543, + "step": 3438 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7931387387387387, + "grad_norm": 27.357887268066406, + "learning_rate": 1.116342431042322e-06, + "loss": 7.132712364196777, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3148741126060486, + "loss_total_v6": 0.4457945190370083, + "step": 3439 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7933693693693694, + "grad_norm": 23.891679763793945, + "learning_rate": 1.1139515362544756e-06, + "loss": 6.619233131408691, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.2848415821790695, + "loss_total_v6": 0.4137020483613014, + "step": 3440 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7936, + "grad_norm": 25.586626052856445, + "learning_rate": 1.1115628835248043e-06, + "loss": 6.986587047576904, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3082894869148731, + "loss_total_v6": 0.4366616867482662, + "step": 3441 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7938306306306306, + "grad_norm": 23.596717834472656, + "learning_rate": 1.1091764742314398e-06, + "loss": 6.680755615234375, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.2894496787339449, + "loss_total_v6": 0.41754720360040665, + "step": 3442 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7940612612612613, + "grad_norm": 24.86579132080078, + "learning_rate": 1.1067923097512256e-06, + "loss": 6.979292392730713, + "loss_alignment": 0.2493896484375, + "loss_alignment_w": 0.12469482421875, + "loss_sft": 0.31141942366957664, + "loss_total_v6": 0.43620578944683075, + "step": 3443 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7942918918918919, + "grad_norm": 28.186782836914062, + "learning_rate": 1.1044103914597033e-06, + "loss": 7.367231369018555, + "loss_alignment": 0.2657470703125, + "loss_alignment_w": 0.13287353515625, + "loss_sft": 0.3274410907179117, + "loss_total_v6": 0.4604519493877888, + "step": 3444 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7945225225225225, + "grad_norm": 23.959089279174805, + "learning_rate": 1.1020307207311244e-06, + "loss": 6.5935564041137695, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.28421333245933056, + "loss_total_v6": 0.4120972603559494, + "step": 3445 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7947531531531532, + "grad_norm": 26.038423538208008, + "learning_rate": 1.0996532989384412e-06, + "loss": 6.825834274291992, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.29876123555004597, + "loss_total_v6": 0.4266146346926689, + "step": 3446 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7949837837837838, + "grad_norm": 25.62110710144043, + "learning_rate": 1.097278127453309e-06, + "loss": 6.782046318054199, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.29436127096414566, + "loss_total_v6": 0.42387786135077477, + "step": 3447 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7952144144144144, + "grad_norm": 27.998306274414062, + "learning_rate": 1.0949052076460854e-06, + "loss": 6.834927082061768, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.299146443605423, + "loss_total_v6": 0.4271829389035702, + "step": 3448 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7954450450450451, + "grad_norm": 27.962989807128906, + "learning_rate": 1.0925345408858256e-06, + "loss": 6.659743309020996, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.28462690487504005, + "loss_total_v6": 0.41623396798968315, + "step": 3449 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7956756756756757, + "grad_norm": 25.944604873657227, + "learning_rate": 1.0901661285402898e-06, + "loss": 6.949878692626953, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.3041946832090616, + "loss_total_v6": 0.43436741828918457, + "step": 3450 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7959063063063063, + "grad_norm": 27.48289680480957, + "learning_rate": 1.087799971975933e-06, + "loss": 7.041930198669434, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3068503439426422, + "loss_total_v6": 0.4401206225156784, + "step": 3451 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.796136936936937, + "grad_norm": 21.795846939086914, + "learning_rate": 1.085436072557911e-06, + "loss": 6.85099458694458, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.2982432935386896, + "loss_total_v6": 0.42818714305758476, + "step": 3452 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7963675675675675, + "grad_norm": 24.928842544555664, + "learning_rate": 1.083074431650078e-06, + "loss": 7.01202392578125, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3088569547981024, + "loss_total_v6": 0.43825148418545723, + "step": 3453 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7965981981981982, + "grad_norm": 25.34624671936035, + "learning_rate": 1.0807150506149839e-06, + "loss": 6.909552574157715, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.3016437888145447, + "loss_total_v6": 0.4318470321595669, + "step": 3454 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7968288288288289, + "grad_norm": 27.19706153869629, + "learning_rate": 1.0783579308138765e-06, + "loss": 6.6753973960876465, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2872837521135807, + "loss_total_v6": 0.4172123335301876, + "step": 3455 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7970594594594594, + "grad_norm": 23.60416030883789, + "learning_rate": 1.0760030736066952e-06, + "loss": 7.039258003234863, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.31199341639876366, + "loss_total_v6": 0.43995363637804985, + "step": 3456 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7972900900900901, + "grad_norm": 24.459169387817383, + "learning_rate": 1.0736504803520787e-06, + "loss": 6.67195987701416, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.2881217710673809, + "loss_total_v6": 0.4169975034892559, + "step": 3457 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7975207207207207, + "grad_norm": 25.353910446166992, + "learning_rate": 1.0713001524073542e-06, + "loss": 7.016459941864014, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.3094393964856863, + "loss_total_v6": 0.43852874636650085, + "step": 3458 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7977513513513513, + "grad_norm": 24.72015380859375, + "learning_rate": 1.0689520911285483e-06, + "loss": 6.9906721115112305, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3066679909825325, + "loss_total_v6": 0.4369170069694519, + "step": 3459 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.797981981981982, + "grad_norm": 26.005741119384766, + "learning_rate": 1.0666062978703733e-06, + "loss": 6.533473491668701, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.2814805619418621, + "loss_total_v6": 0.4083421155810356, + "step": 3460 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7982126126126126, + "grad_norm": 23.864845275878906, + "learning_rate": 1.0642627739862382e-06, + "loss": 7.207679748535156, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.32084131985902786, + "loss_total_v6": 0.45048000290989876, + "step": 3461 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7984432432432432, + "grad_norm": 29.6826171875, + "learning_rate": 1.0619215208282397e-06, + "loss": 7.057623386383057, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.31245461106300354, + "loss_total_v6": 0.44110145419836044, + "step": 3462 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7986738738738739, + "grad_norm": 24.438339233398438, + "learning_rate": 1.059582539747166e-06, + "loss": 6.8291916847229, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.2987574599683285, + "loss_total_v6": 0.426824476569891, + "step": 3463 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7989045045045045, + "grad_norm": 26.997777938842773, + "learning_rate": 1.0572458320924943e-06, + "loss": 6.64979887008667, + "loss_alignment": 0.2467041015625, + "loss_alignment_w": 0.12335205078125, + "loss_sft": 0.29256555065512657, + "loss_total_v6": 0.4156124219298363, + "step": 3464 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7991351351351351, + "grad_norm": 23.84813117980957, + "learning_rate": 1.0549113992123878e-06, + "loss": 6.522146701812744, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.27833120711147785, + "loss_total_v6": 0.4076341837644577, + "step": 3465 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7993657657657658, + "grad_norm": 26.075923919677734, + "learning_rate": 1.0525792424537e-06, + "loss": 6.878881454467773, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.3012222349643707, + "loss_total_v6": 0.42993010953068733, + "step": 3466 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.7995963963963963, + "grad_norm": 26.22519302368164, + "learning_rate": 1.0502493631619715e-06, + "loss": 6.531098365783691, + "loss_alignment": 0.250732421875, + "loss_alignment_w": 0.1253662109375, + "loss_sft": 0.28275116346776485, + "loss_total_v6": 0.4081936627626419, + "step": 3467 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.799827027027027, + "grad_norm": 23.23867416381836, + "learning_rate": 1.0479217626814252e-06, + "loss": 6.775137901306152, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.29415843449532986, + "loss_total_v6": 0.4234461449086666, + "step": 3468 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8000576576576577, + "grad_norm": 24.00800895690918, + "learning_rate": 1.0455964423549741e-06, + "loss": 7.168270111083984, + "loss_alignment": 0.251708984375, + "loss_alignment_w": 0.1258544921875, + "loss_sft": 0.3222997151315212, + "loss_total_v6": 0.4480168893933296, + "step": 3469 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8002882882882882, + "grad_norm": 25.331174850463867, + "learning_rate": 1.0432734035242115e-06, + "loss": 7.077507495880127, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.3150706849992275, + "loss_total_v6": 0.44234422966837883, + "step": 3470 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8005189189189189, + "grad_norm": 24.068870544433594, + "learning_rate": 1.0409526475294156e-06, + "loss": 6.602653503417969, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.2817607205361128, + "loss_total_v6": 0.41266586259007454, + "step": 3471 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8007495495495496, + "grad_norm": 26.11575698852539, + "learning_rate": 1.0386341757095503e-06, + "loss": 6.567121505737305, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.28098952397704124, + "loss_total_v6": 0.41044510528445244, + "step": 3472 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8009801801801801, + "grad_norm": 24.045244216918945, + "learning_rate": 1.036317989402258e-06, + "loss": 7.304308891296387, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.32717055827379227, + "loss_total_v6": 0.4565192982554436, + "step": 3473 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8012108108108108, + "grad_norm": 28.90226936340332, + "learning_rate": 1.0340040899438652e-06, + "loss": 6.919393539428711, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.30518851801753044, + "loss_total_v6": 0.43246208503842354, + "step": 3474 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8014414414414415, + "grad_norm": 24.058074951171875, + "learning_rate": 1.0316924786693754e-06, + "loss": 6.931236743927002, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.301213763654232, + "loss_total_v6": 0.433202289044857, + "step": 3475 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.801672072072072, + "grad_norm": 24.934776306152344, + "learning_rate": 1.0293831569124774e-06, + "loss": 6.838468551635742, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.29666693136096, + "loss_total_v6": 0.4274042546749115, + "step": 3476 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8019027027027027, + "grad_norm": 26.68366050720215, + "learning_rate": 1.0270761260055323e-06, + "loss": 7.464673042297363, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.3363846242427826, + "loss_total_v6": 0.4665420986711979, + "step": 3477 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8021333333333334, + "grad_norm": 26.22174835205078, + "learning_rate": 1.024771387279585e-06, + "loss": 7.042380332946777, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3116087578237057, + "loss_total_v6": 0.4401487894356251, + "step": 3478 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8023639639639639, + "grad_norm": 25.737852096557617, + "learning_rate": 1.0224689420643558e-06, + "loss": 6.741210460662842, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.2929839603602886, + "loss_total_v6": 0.4213256537914276, + "step": 3479 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8025945945945946, + "grad_norm": 22.032230377197266, + "learning_rate": 1.020168791688242e-06, + "loss": 6.952971935272217, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3070278000086546, + "loss_total_v6": 0.43456074595451355, + "step": 3480 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8028252252252253, + "grad_norm": 24.80775260925293, + "learning_rate": 1.017870937478318e-06, + "loss": 7.46380615234375, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.336040498688817, + "loss_total_v6": 0.4664878770709038, + "step": 3481 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8030558558558558, + "grad_norm": 23.051361083984375, + "learning_rate": 1.0155753807603297e-06, + "loss": 6.94194221496582, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.30331716127693653, + "loss_total_v6": 0.433871366083622, + "step": 3482 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8032864864864865, + "grad_norm": 23.01740264892578, + "learning_rate": 1.013282122858703e-06, + "loss": 7.166531085968018, + "loss_alignment": 0.2508544921875, + "loss_alignment_w": 0.12542724609375, + "loss_sft": 0.3226335532963276, + "loss_total_v6": 0.447908204048872, + "step": 3483 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8035171171171172, + "grad_norm": 24.046573638916016, + "learning_rate": 1.0109911650965314e-06, + "loss": 6.624848365783691, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.28532982990145683, + "loss_total_v6": 0.414052989333868, + "step": 3484 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8037477477477477, + "grad_norm": 23.22113609313965, + "learning_rate": 1.008702508795587e-06, + "loss": 6.765101432800293, + "loss_alignment": 0.2520751953125, + "loss_alignment_w": 0.12603759765625, + "loss_sft": 0.29691858030855656, + "loss_total_v6": 0.4228188470005989, + "step": 3485 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8039783783783784, + "grad_norm": 22.64310646057129, + "learning_rate": 1.0064161552763096e-06, + "loss": 7.038968563079834, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3103273641318083, + "loss_total_v6": 0.4399355351924896, + "step": 3486 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.804209009009009, + "grad_norm": 25.79750633239746, + "learning_rate": 1.0041321058578135e-06, + "loss": 6.948821067810059, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.30475420504808426, + "loss_total_v6": 0.43430132418870926, + "step": 3487 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8044396396396396, + "grad_norm": 24.173545837402344, + "learning_rate": 1.0018503618578817e-06, + "loss": 6.726197719573975, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.2926255390048027, + "loss_total_v6": 0.420387364923954, + "step": 3488 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8046702702702703, + "grad_norm": 24.91179656982422, + "learning_rate": 9.995709245929691e-07, + "loss": 7.253885746002197, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.32427848502993584, + "loss_total_v6": 0.45336784794926643, + "step": 3489 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8049009009009009, + "grad_norm": 24.38319206237793, + "learning_rate": 9.972937953781985e-07, + "loss": 7.005218029022217, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.30666157230734825, + "loss_total_v6": 0.43782613053917885, + "step": 3490 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8051315315315315, + "grad_norm": 25.695859909057617, + "learning_rate": 9.950189755273592e-07, + "loss": 6.644609451293945, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2869769185781479, + "loss_total_v6": 0.415288083255291, + "step": 3491 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8053621621621622, + "grad_norm": 22.975942611694336, + "learning_rate": 9.927464663529119e-07, + "loss": 7.027493953704834, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.3118227422237396, + "loss_total_v6": 0.439218383282423, + "step": 3492 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8055927927927928, + "grad_norm": 23.524124145507812, + "learning_rate": 9.90476269165982e-07, + "loss": 7.269067287445068, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.32441866025328636, + "loss_total_v6": 0.45431672409176826, + "step": 3493 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8058234234234234, + "grad_norm": 25.541671752929688, + "learning_rate": 9.882083852763597e-07, + "loss": 6.634605407714844, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.28595495969057083, + "loss_total_v6": 0.41466283798217773, + "step": 3494 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8060540540540541, + "grad_norm": 25.316282272338867, + "learning_rate": 9.859428159925044e-07, + "loss": 7.036489009857178, + "loss_alignment": 0.27197265625, + "loss_alignment_w": 0.135986328125, + "loss_sft": 0.30362639017403126, + "loss_total_v6": 0.4397805519402027, + "step": 3495 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8062846846846847, + "grad_norm": 24.827199935913086, + "learning_rate": 9.836795626215356e-07, + "loss": 6.35577392578125, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.2680854834616184, + "loss_total_v6": 0.3972358703613281, + "step": 3496 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8065153153153153, + "grad_norm": 24.524791717529297, + "learning_rate": 9.814186264692394e-07, + "loss": 6.869132041931152, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.302123473957181, + "loss_total_v6": 0.4293207414448261, + "step": 3497 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.806745945945946, + "grad_norm": 25.920513153076172, + "learning_rate": 9.791600088400643e-07, + "loss": 6.718386650085449, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2899095304310322, + "loss_total_v6": 0.41989915817976, + "step": 3498 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8069765765765766, + "grad_norm": 27.597684860229492, + "learning_rate": 9.769037110371222e-07, + "loss": 6.769489288330078, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.29238629899919033, + "loss_total_v6": 0.423093069344759, + "step": 3499 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8072072072072072, + "grad_norm": 25.138774871826172, + "learning_rate": 9.746497343621857e-07, + "loss": 6.714563369750977, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.29208143800497055, + "loss_total_v6": 0.41966018825769424, + "step": 3500 + }, + { + "epoch": 0.8072072072072072, + "eval_loss": 0.56589275598526, + "eval_loss_alignment": 0.2319068386130137, + "eval_loss_alignment_w": 0.11595341930650685, + "eval_loss_sft": 0.4499392959982429, + "eval_loss_total_v6": 0.5658927152622236, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.9129, + "eval_samples_per_second": 18.985, + "eval_steps_per_second": 2.383, + "step": 3500 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8074378378378378, + "grad_norm": 22.099773406982422, + "learning_rate": 9.723980801156863e-07, + "loss": 7.098269462585449, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3124772999435663, + "loss_total_v6": 0.4436418488621712, + "step": 3501 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8076684684684685, + "grad_norm": 25.417091369628906, + "learning_rate": 9.701487495967198e-07, + "loss": 7.129385948181152, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.3164820075035095, + "loss_total_v6": 0.4455866180360317, + "step": 3502 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8078990990990991, + "grad_norm": 28.005849838256836, + "learning_rate": 9.679017441030363e-07, + "loss": 6.714847564697266, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.2890016846358776, + "loss_total_v6": 0.4196779541671276, + "step": 3503 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8081297297297297, + "grad_norm": 24.228065490722656, + "learning_rate": 9.656570649310483e-07, + "loss": 6.567734718322754, + "loss_alignment": 0.2501220703125, + "loss_alignment_w": 0.12506103515625, + "loss_sft": 0.2854681573808193, + "loss_total_v6": 0.4104834161698818, + "step": 3504 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8083603603603604, + "grad_norm": 24.6109561920166, + "learning_rate": 9.634147133758248e-07, + "loss": 6.774574279785156, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.29279562272131443, + "loss_total_v6": 0.4234108738601208, + "step": 3505 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.808590990990991, + "grad_norm": 25.995574951171875, + "learning_rate": 9.61174690731092e-07, + "loss": 6.919001579284668, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.30290575698018074, + "loss_total_v6": 0.43243760988116264, + "step": 3506 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8088216216216216, + "grad_norm": 23.60700225830078, + "learning_rate": 9.589369982892338e-07, + "loss": 7.121640205383301, + "loss_alignment": 0.26513671875, + "loss_alignment_w": 0.132568359375, + "loss_sft": 0.31253413297235966, + "loss_total_v6": 0.4451024942100048, + "step": 3507 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8090522522522523, + "grad_norm": 28.406768798828125, + "learning_rate": 9.567016373412857e-07, + "loss": 6.79278039932251, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.29587141796946526, + "loss_total_v6": 0.42454877868294716, + "step": 3508 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8092828828828829, + "grad_norm": 25.5246639251709, + "learning_rate": 9.544686091769434e-07, + "loss": 6.964575290679932, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.303816232830286, + "loss_total_v6": 0.4352859631180763, + "step": 3509 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8095135135135135, + "grad_norm": 27.55365753173828, + "learning_rate": 9.522379150845512e-07, + "loss": 7.069645881652832, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.3124583698809147, + "loss_total_v6": 0.4418528899550438, + "step": 3510 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8097441441441442, + "grad_norm": 26.208969116210938, + "learning_rate": 9.500095563511119e-07, + "loss": 6.784219741821289, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.29370366781949997, + "loss_total_v6": 0.4240137152373791, + "step": 3511 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8099747747747748, + "grad_norm": 25.374380111694336, + "learning_rate": 9.47783534262276e-07, + "loss": 7.20320987701416, + "loss_alignment": 0.2515869140625, + "loss_alignment_w": 0.12579345703125, + "loss_sft": 0.3242851048707962, + "loss_total_v6": 0.4502006322145462, + "step": 3512 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8102054054054054, + "grad_norm": 24.78176498413086, + "learning_rate": 9.4555985010235e-07, + "loss": 7.091712951660156, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.31394432485103607, + "loss_total_v6": 0.4432320408523083, + "step": 3513 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8104360360360361, + "grad_norm": 26.979930877685547, + "learning_rate": 9.433385051542892e-07, + "loss": 6.9637064933776855, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.30481477454304695, + "loss_total_v6": 0.43523165583610535, + "step": 3514 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8106666666666666, + "grad_norm": 26.14642333984375, + "learning_rate": 9.41119500699701e-07, + "loss": 7.034607410430908, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3077965136617422, + "loss_total_v6": 0.43966295942664146, + "step": 3515 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8108972972972973, + "grad_norm": 29.68674659729004, + "learning_rate": 9.38902838018842e-07, + "loss": 7.297330379486084, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.32479654252529144, + "loss_total_v6": 0.45608315244317055, + "step": 3516 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.811127927927928, + "grad_norm": 26.51676368713379, + "learning_rate": 9.366885183906149e-07, + "loss": 6.784639358520508, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.29549992457032204, + "loss_total_v6": 0.42403996363282204, + "step": 3517 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8113585585585585, + "grad_norm": 23.49327278137207, + "learning_rate": 9.34476543092574e-07, + "loss": 6.690473556518555, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.28799712657928467, + "loss_total_v6": 0.41815459355711937, + "step": 3518 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8115891891891892, + "grad_norm": 22.915454864501953, + "learning_rate": 9.322669134009215e-07, + "loss": 6.729767799377441, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.2929707281291485, + "loss_total_v6": 0.4206104911863804, + "step": 3519 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8118198198198198, + "grad_norm": 25.813047409057617, + "learning_rate": 9.300596305905013e-07, + "loss": 6.738324165344238, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.2933834008872509, + "loss_total_v6": 0.4211452454328537, + "step": 3520 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8120504504504504, + "grad_norm": 27.430341720581055, + "learning_rate": 9.2785469593481e-07, + "loss": 6.846066474914551, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.3001325875520706, + "loss_total_v6": 0.4278791770339012, + "step": 3521 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8122810810810811, + "grad_norm": 27.344646453857422, + "learning_rate": 9.256521107059834e-07, + "loss": 7.19869327545166, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.3198981750756502, + "loss_total_v6": 0.44991831108927727, + "step": 3522 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8125117117117117, + "grad_norm": 24.3715877532959, + "learning_rate": 9.234518761748057e-07, + "loss": 7.383890628814697, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3310915306210518, + "loss_total_v6": 0.4614931605756283, + "step": 3523 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8127423423423423, + "grad_norm": 25.074132919311523, + "learning_rate": 9.212539936107029e-07, + "loss": 6.744569778442383, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.2937432490289211, + "loss_total_v6": 0.4215356074273586, + "step": 3524 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.812972972972973, + "grad_norm": 25.488903045654297, + "learning_rate": 9.190584642817463e-07, + "loss": 7.198189735412598, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.32040081545710564, + "loss_total_v6": 0.44988688826560974, + "step": 3525 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8132036036036036, + "grad_norm": 25.240819931030273, + "learning_rate": 9.168652894546476e-07, + "loss": 6.880825996398926, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.30216772109270096, + "loss_total_v6": 0.43005163595080376, + "step": 3526 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8134342342342342, + "grad_norm": 22.964757919311523, + "learning_rate": 9.146744703947591e-07, + "loss": 6.917941570281982, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3037702813744545, + "loss_total_v6": 0.4323713518679142, + "step": 3527 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8136648648648649, + "grad_norm": 26.583229064941406, + "learning_rate": 9.12486008366077e-07, + "loss": 6.932280540466309, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.3064975291490555, + "loss_total_v6": 0.4332675412297249, + "step": 3528 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8138954954954954, + "grad_norm": 28.873668670654297, + "learning_rate": 9.102999046312339e-07, + "loss": 7.343510627746582, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.33090241253376007, + "loss_total_v6": 0.45896943286061287, + "step": 3529 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8141261261261261, + "grad_norm": 22.01143455505371, + "learning_rate": 9.081161604515043e-07, + "loss": 7.248075008392334, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.32313714921474457, + "loss_total_v6": 0.4530046917498112, + "step": 3530 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8143567567567568, + "grad_norm": 26.842105865478516, + "learning_rate": 9.059347770868021e-07, + "loss": 7.233206272125244, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.32153647020459175, + "loss_total_v6": 0.45207539945840836, + "step": 3531 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8145873873873873, + "grad_norm": 26.9864444732666, + "learning_rate": 9.037557557956766e-07, + "loss": 6.886252403259277, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.2991652023047209, + "loss_total_v6": 0.4303907938301563, + "step": 3532 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.814818018018018, + "grad_norm": 24.570953369140625, + "learning_rate": 9.015790978353173e-07, + "loss": 6.679776191711426, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.28787786327302456, + "loss_total_v6": 0.4174860082566738, + "step": 3533 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8150486486486487, + "grad_norm": 24.599573135375977, + "learning_rate": 8.994048044615461e-07, + "loss": 6.970211029052734, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3049161396920681, + "loss_total_v6": 0.4356381967663765, + "step": 3534 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8152792792792792, + "grad_norm": 24.99618911743164, + "learning_rate": 8.972328769288252e-07, + "loss": 6.733549118041992, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2919405773282051, + "loss_total_v6": 0.4208468347787857, + "step": 3535 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8155099099099099, + "grad_norm": 27.066362380981445, + "learning_rate": 8.950633164902467e-07, + "loss": 6.991349220275879, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.3095331937074661, + "loss_total_v6": 0.4369593411684036, + "step": 3536 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8157405405405406, + "grad_norm": 22.38064193725586, + "learning_rate": 8.928961243975437e-07, + "loss": 6.897361755371094, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3022856656461954, + "loss_total_v6": 0.43108509853482246, + "step": 3537 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8159711711711711, + "grad_norm": 22.37954330444336, + "learning_rate": 8.90731301901075e-07, + "loss": 7.2241668701171875, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.32202432304620743, + "loss_total_v6": 0.4515104256570339, + "step": 3538 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8162018018018018, + "grad_norm": 21.799806594848633, + "learning_rate": 8.885688502498385e-07, + "loss": 6.886532783508301, + "loss_alignment": 0.2503662109375, + "loss_alignment_w": 0.12518310546875, + "loss_sft": 0.30502684228122234, + "loss_total_v6": 0.43040831387043, + "step": 3539 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8164324324324325, + "grad_norm": 26.486867904663086, + "learning_rate": 8.86408770691462e-07, + "loss": 6.724325180053711, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2920812126249075, + "loss_total_v6": 0.42027030140161514, + "step": 3540 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.816663063063063, + "grad_norm": 23.751785278320312, + "learning_rate": 8.842510644722047e-07, + "loss": 6.644399642944336, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.28552949987351894, + "loss_total_v6": 0.4152749814093113, + "step": 3541 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8168936936936937, + "grad_norm": 23.886951446533203, + "learning_rate": 8.820957328369578e-07, + "loss": 6.860321998596191, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.29958921670913696, + "loss_total_v6": 0.42877013608813286, + "step": 3542 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8171243243243244, + "grad_norm": 25.68241310119629, + "learning_rate": 8.79942777029239e-07, + "loss": 6.6855621337890625, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.29124546609818935, + "loss_total_v6": 0.4178476370871067, + "step": 3543 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8173549549549549, + "grad_norm": 24.794015884399414, + "learning_rate": 8.777921982911996e-07, + "loss": 7.007515907287598, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.30704935640096664, + "loss_total_v6": 0.43796976655721664, + "step": 3544 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8175855855855856, + "grad_norm": 23.55233383178711, + "learning_rate": 8.75643997863615e-07, + "loss": 6.984802722930908, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3087425697594881, + "loss_total_v6": 0.43655018135905266, + "step": 3545 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8178162162162163, + "grad_norm": 25.53053855895996, + "learning_rate": 8.734981769858925e-07, + "loss": 7.0361833572387695, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3117402084171772, + "loss_total_v6": 0.4397614523768425, + "step": 3546 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8180468468468468, + "grad_norm": 29.858556747436523, + "learning_rate": 8.713547368960656e-07, + "loss": 6.901159763336182, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.30166855826973915, + "loss_total_v6": 0.43132248893380165, + "step": 3547 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8182774774774775, + "grad_norm": 25.229108810424805, + "learning_rate": 8.692136788307903e-07, + "loss": 7.239195823669434, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.3252372145652771, + "loss_total_v6": 0.4524497352540493, + "step": 3548 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8185081081081081, + "grad_norm": 24.271453857421875, + "learning_rate": 8.670750040253561e-07, + "loss": 6.701560020446777, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.28984969295561314, + "loss_total_v6": 0.4188474975526333, + "step": 3549 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8187387387387387, + "grad_norm": 24.67306900024414, + "learning_rate": 8.649387137136689e-07, + "loss": 7.17255163192749, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.31719620153307915, + "loss_total_v6": 0.44828445836901665, + "step": 3550 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8189693693693694, + "grad_norm": 22.274227142333984, + "learning_rate": 8.628048091282637e-07, + "loss": 6.965875625610352, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.30591166764497757, + "loss_total_v6": 0.43536723405122757, + "step": 3551 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8192, + "grad_norm": 23.817903518676758, + "learning_rate": 8.606732915003002e-07, + "loss": 7.023225784301758, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.30887046456336975, + "loss_total_v6": 0.43895163759589195, + "step": 3552 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8194306306306306, + "grad_norm": 24.5721435546875, + "learning_rate": 8.585441620595559e-07, + "loss": 6.980978965759277, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.30792373046278954, + "loss_total_v6": 0.43631119653582573, + "step": 3553 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8196612612612613, + "grad_norm": 23.30042266845703, + "learning_rate": 8.564174220344362e-07, + "loss": 6.902216911315918, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.29988835752010345, + "loss_total_v6": 0.43138857930898666, + "step": 3554 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8198918918918919, + "grad_norm": 27.874191284179688, + "learning_rate": 8.542930726519622e-07, + "loss": 7.189518928527832, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3210642971098423, + "loss_total_v6": 0.4493449404835701, + "step": 3555 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8201225225225225, + "grad_norm": 26.048128128051758, + "learning_rate": 8.521711151377804e-07, + "loss": 6.849564552307129, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3006105851382017, + "loss_total_v6": 0.42809776589274406, + "step": 3556 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8203531531531532, + "grad_norm": 23.06169319152832, + "learning_rate": 8.50051550716155e-07, + "loss": 6.860598564147949, + "loss_alignment": 0.2501220703125, + "loss_alignment_w": 0.12506103515625, + "loss_sft": 0.3039705231785774, + "loss_total_v6": 0.4287874288856983, + "step": 3557 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8205837837837838, + "grad_norm": 28.38227081298828, + "learning_rate": 8.479343806099705e-07, + "loss": 6.60593318939209, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.28512422554194927, + "loss_total_v6": 0.4128708131611347, + "step": 3558 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8208144144144144, + "grad_norm": 27.90231704711914, + "learning_rate": 8.458196060407308e-07, + "loss": 7.350487232208252, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.3307280819863081, + "loss_total_v6": 0.45940544456243515, + "step": 3559 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.821045045045045, + "grad_norm": 24.20928382873535, + "learning_rate": 8.437072282285536e-07, + "loss": 7.26827335357666, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.32589486986398697, + "loss_total_v6": 0.45426706969738007, + "step": 3560 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8212756756756757, + "grad_norm": 25.46185874938965, + "learning_rate": 8.415972483921792e-07, + "loss": 6.745429039001465, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.290363734588027, + "loss_total_v6": 0.42158931493759155, + "step": 3561 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8215063063063063, + "grad_norm": 26.345754623413086, + "learning_rate": 8.3948966774896e-07, + "loss": 6.371013164520264, + "loss_alignment": 0.2515869140625, + "loss_alignment_w": 0.12579345703125, + "loss_sft": 0.27247113548219204, + "loss_total_v6": 0.3981883153319359, + "step": 3562 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.821736936936937, + "grad_norm": 29.463138580322266, + "learning_rate": 8.373844875148679e-07, + "loss": 6.611767768859863, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.28475645929574966, + "loss_total_v6": 0.41323547065258026, + "step": 3563 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8219675675675676, + "grad_norm": 26.570730209350586, + "learning_rate": 8.352817089044851e-07, + "loss": 7.069597244262695, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3112345989793539, + "loss_total_v6": 0.44184985011816025, + "step": 3564 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8221981981981982, + "grad_norm": 26.972949981689453, + "learning_rate": 8.331813331310134e-07, + "loss": 6.870495319366455, + "loss_alignment": 0.2501220703125, + "loss_alignment_w": 0.12506103515625, + "loss_sft": 0.30463483557105064, + "loss_total_v6": 0.42940595373511314, + "step": 3565 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8224288288288288, + "grad_norm": 25.232376098632812, + "learning_rate": 8.310833614062652e-07, + "loss": 6.247714996337891, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.26090456172823906, + "loss_total_v6": 0.39048220589756966, + "step": 3566 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8226594594594595, + "grad_norm": 24.904293060302734, + "learning_rate": 8.289877949406672e-07, + "loss": 6.619387626647949, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.2864076551049948, + "loss_total_v6": 0.41371170803904533, + "step": 3567 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8228900900900901, + "grad_norm": 27.15870475769043, + "learning_rate": 8.268946349432583e-07, + "loss": 6.8018999099731445, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.29523590952157974, + "loss_total_v6": 0.42511873692274094, + "step": 3568 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8231207207207207, + "grad_norm": 27.466367721557617, + "learning_rate": 8.248038826216864e-07, + "loss": 6.964132308959961, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.30650459975004196, + "loss_total_v6": 0.43525825068354607, + "step": 3569 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8233513513513514, + "grad_norm": 22.146860122680664, + "learning_rate": 8.22715539182215e-07, + "loss": 7.1536431312561035, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3186389524489641, + "loss_total_v6": 0.44710269197821617, + "step": 3570 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.823581981981982, + "grad_norm": 28.81201171875, + "learning_rate": 8.206296058297126e-07, + "loss": 6.261664390563965, + "loss_alignment": 0.2510986328125, + "loss_alignment_w": 0.12554931640625, + "loss_sft": 0.2659573070704937, + "loss_total_v6": 0.3913540244102478, + "step": 3571 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8238126126126126, + "grad_norm": 25.735567092895508, + "learning_rate": 8.185460837676612e-07, + "loss": 6.934994220733643, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.30511070787906647, + "loss_total_v6": 0.43343713507056236, + "step": 3572 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8240432432432433, + "grad_norm": 26.212350845336914, + "learning_rate": 8.164649741981517e-07, + "loss": 7.204170227050781, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3197827525436878, + "loss_total_v6": 0.4502606615424156, + "step": 3573 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8242738738738739, + "grad_norm": 23.56644630432129, + "learning_rate": 8.14386278321877e-07, + "loss": 7.516668319702148, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3393291346728802, + "loss_total_v6": 0.4697917774319649, + "step": 3574 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8245045045045045, + "grad_norm": 25.296518325805664, + "learning_rate": 8.123099973381476e-07, + "loss": 6.920825481414795, + "loss_alignment": 0.2520751953125, + "loss_alignment_w": 0.12603759765625, + "loss_sft": 0.3063003644347191, + "loss_total_v6": 0.432551596313715, + "step": 3575 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8247351351351352, + "grad_norm": 28.272830963134766, + "learning_rate": 8.102361324448715e-07, + "loss": 6.827523231506348, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.29883630760014057, + "loss_total_v6": 0.4267202205955982, + "step": 3576 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8249657657657657, + "grad_norm": 23.15123176574707, + "learning_rate": 8.081646848385671e-07, + "loss": 7.338486671447754, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.3303595073521137, + "loss_total_v6": 0.4586554169654846, + "step": 3577 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8251963963963964, + "grad_norm": 23.635372161865234, + "learning_rate": 8.060956557143595e-07, + "loss": 7.3741865158081055, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.3296915888786316, + "loss_total_v6": 0.4608866535127163, + "step": 3578 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.825427027027027, + "grad_norm": 23.3763427734375, + "learning_rate": 8.040290462659733e-07, + "loss": 7.134360313415527, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.31555694714188576, + "loss_total_v6": 0.44589753076434135, + "step": 3579 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8256576576576576, + "grad_norm": 28.105836868286133, + "learning_rate": 8.019648576857425e-07, + "loss": 7.13007116317749, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.31681475415825844, + "loss_total_v6": 0.44562945142388344, + "step": 3580 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8258882882882883, + "grad_norm": 25.2263240814209, + "learning_rate": 7.999030911646005e-07, + "loss": 7.41963005065918, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.3362092077732086, + "loss_total_v6": 0.4637269005179405, + "step": 3581 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8261189189189189, + "grad_norm": 24.73691749572754, + "learning_rate": 7.978437478920847e-07, + "loss": 7.141920566558838, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.31840984895825386, + "loss_total_v6": 0.44637005031108856, + "step": 3582 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8263495495495495, + "grad_norm": 25.633350372314453, + "learning_rate": 7.957868290563359e-07, + "loss": 7.247547149658203, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3244316540658474, + "loss_total_v6": 0.4529717043042183, + "step": 3583 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8265801801801802, + "grad_norm": 27.421911239624023, + "learning_rate": 7.937323358440935e-07, + "loss": 6.814112663269043, + "loss_alignment": 0.2535400390625, + "loss_alignment_w": 0.12677001953125, + "loss_sft": 0.2993561774492264, + "loss_total_v6": 0.4258820600807667, + "step": 3584 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8268108108108108, + "grad_norm": 24.032020568847656, + "learning_rate": 7.916802694407011e-07, + "loss": 6.324572563171387, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.26810377836227417, + "loss_total_v6": 0.39528578892350197, + "step": 3585 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8270414414414414, + "grad_norm": 23.43284797668457, + "learning_rate": 7.896306310300961e-07, + "loss": 7.254100322723389, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.322247251868248, + "loss_total_v6": 0.4533812776207924, + "step": 3586 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8272720720720721, + "grad_norm": 26.24119758605957, + "learning_rate": 7.875834217948219e-07, + "loss": 6.714853286743164, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.2938543651252985, + "loss_total_v6": 0.41967833042144775, + "step": 3587 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8275027027027027, + "grad_norm": 25.26097297668457, + "learning_rate": 7.85538642916015e-07, + "loss": 7.007227420806885, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3079773634672165, + "loss_total_v6": 0.4379517175257206, + "step": 3588 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8277333333333333, + "grad_norm": 23.674354553222656, + "learning_rate": 7.834962955734138e-07, + "loss": 6.665524482727051, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2875516824424267, + "loss_total_v6": 0.4165952615439892, + "step": 3589 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.827963963963964, + "grad_norm": 23.4420108795166, + "learning_rate": 7.814563809453496e-07, + "loss": 6.7228169441223145, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.29262784123420715, + "loss_total_v6": 0.42017605528235435, + "step": 3590 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8281945945945945, + "grad_norm": 21.975597381591797, + "learning_rate": 7.794189002087538e-07, + "loss": 6.888303279876709, + "loss_alignment": 0.2508544921875, + "loss_alignment_w": 0.12542724609375, + "loss_sft": 0.3051069565117359, + "loss_total_v6": 0.4305189587175846, + "step": 3591 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8284252252252252, + "grad_norm": 26.789573669433594, + "learning_rate": 7.773838545391516e-07, + "loss": 7.218719482421875, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.3211345784366131, + "loss_total_v6": 0.4511699788272381, + "step": 3592 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8286558558558559, + "grad_norm": 27.310077667236328, + "learning_rate": 7.753512451106648e-07, + "loss": 7.217113494873047, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.32121728733181953, + "loss_total_v6": 0.45106959342956543, + "step": 3593 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8288864864864864, + "grad_norm": 30.28376007080078, + "learning_rate": 7.7332107309601e-07, + "loss": 6.839767932891846, + "loss_alignment": 0.251708984375, + "loss_alignment_w": 0.1258544921875, + "loss_sft": 0.30167678371071815, + "loss_total_v6": 0.42748550325632095, + "step": 3594 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8291171171171171, + "grad_norm": 24.90283203125, + "learning_rate": 7.712933396664929e-07, + "loss": 7.029283046722412, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.31082067266106606, + "loss_total_v6": 0.43933017924427986, + "step": 3595 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8293477477477478, + "grad_norm": 26.436731338500977, + "learning_rate": 7.692680459920187e-07, + "loss": 6.5400166511535645, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.28219461999833584, + "loss_total_v6": 0.4087510295212269, + "step": 3596 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8295783783783783, + "grad_norm": 27.445594787597656, + "learning_rate": 7.67245193241079e-07, + "loss": 6.910370826721191, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.30502135679125786, + "loss_total_v6": 0.43189819157123566, + "step": 3597 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.829809009009009, + "grad_norm": 44.71440505981445, + "learning_rate": 7.652247825807624e-07, + "loss": 6.819392681121826, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.29800768569111824, + "loss_total_v6": 0.42621204257011414, + "step": 3598 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8300396396396397, + "grad_norm": 22.03183364868164, + "learning_rate": 7.632068151767447e-07, + "loss": 6.730815887451172, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.29149507358670235, + "loss_total_v6": 0.42067598551511765, + "step": 3599 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8302702702702702, + "grad_norm": 23.581398010253906, + "learning_rate": 7.611912921932941e-07, + "loss": 7.151060104370117, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.3180197663605213, + "loss_total_v6": 0.4469412751495838, + "step": 3600 + }, + { + "epoch": 0.8302702702702702, + "eval_loss": 0.5648824572563171, + "eval_loss_alignment": 0.23085224029680365, + "eval_loss_alignment_w": 0.11542612014840183, + "eval_loss_sft": 0.4494563753474249, + "eval_loss_total_v6": 0.5648824958020149, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.4377, + "eval_samples_per_second": 19.084, + "eval_steps_per_second": 2.395, + "step": 3600 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8305009009009009, + "grad_norm": 21.681459426879883, + "learning_rate": 7.591782147932692e-07, + "loss": 7.120366096496582, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.31752045452594757, + "loss_total_v6": 0.445022888481617, + "step": 3601 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8307315315315316, + "grad_norm": 26.557628631591797, + "learning_rate": 7.571675841381132e-07, + "loss": 6.814563751220703, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.29735491052269936, + "loss_total_v6": 0.42591020464897156, + "step": 3602 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8309621621621621, + "grad_norm": 25.621475219726562, + "learning_rate": 7.551594013878643e-07, + "loss": 6.805476188659668, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.29542893916368484, + "loss_total_v6": 0.42534226179122925, + "step": 3603 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8311927927927928, + "grad_norm": 21.337594985961914, + "learning_rate": 7.531536677011419e-07, + "loss": 7.035090446472168, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3101460561156273, + "loss_total_v6": 0.4396931678056717, + "step": 3604 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8314234234234235, + "grad_norm": 25.70911407470703, + "learning_rate": 7.511503842351564e-07, + "loss": 7.366268634796143, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.33064630068838596, + "loss_total_v6": 0.4603917822241783, + "step": 3605 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.831654054054054, + "grad_norm": 26.148029327392578, + "learning_rate": 7.491495521457043e-07, + "loss": 7.173871040344238, + "loss_alignment": 0.264404296875, + "loss_alignment_w": 0.1322021484375, + "loss_sft": 0.31634788773953915, + "loss_total_v6": 0.4483669362962246, + "step": 3606 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8318846846846847, + "grad_norm": 29.129289627075195, + "learning_rate": 7.471511725871655e-07, + "loss": 6.866501331329346, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.3017606809735298, + "loss_total_v6": 0.4291563294827938, + "step": 3607 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8321153153153154, + "grad_norm": 25.521026611328125, + "learning_rate": 7.451552467125072e-07, + "loss": 7.092624664306641, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.31485581398010254, + "loss_total_v6": 0.44328904896974564, + "step": 3608 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8323459459459459, + "grad_norm": 24.13743019104004, + "learning_rate": 7.431617756732801e-07, + "loss": 6.922097682952881, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.30346545204520226, + "loss_total_v6": 0.43263112381100655, + "step": 3609 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8325765765765766, + "grad_norm": 25.859806060791016, + "learning_rate": 7.411707606196189e-07, + "loss": 7.349845886230469, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.3312525451183319, + "loss_total_v6": 0.4593653418123722, + "step": 3610 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8328072072072072, + "grad_norm": 52.4101676940918, + "learning_rate": 7.391822027002409e-07, + "loss": 6.772553443908691, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.2950039245188236, + "loss_total_v6": 0.4232845790684223, + "step": 3611 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8330378378378378, + "grad_norm": 25.13279914855957, + "learning_rate": 7.371961030624452e-07, + "loss": 7.1761322021484375, + "loss_alignment": 0.2535400390625, + "loss_alignment_w": 0.12677001953125, + "loss_sft": 0.32170769199728966, + "loss_total_v6": 0.44850824028253555, + "step": 3612 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8332684684684685, + "grad_norm": 24.717832565307617, + "learning_rate": 7.352124628521151e-07, + "loss": 6.6018524169921875, + "loss_alignment": 0.2496337890625, + "loss_alignment_w": 0.12481689453125, + "loss_sft": 0.2880430147051811, + "loss_total_v6": 0.4126157835125923, + "step": 3613 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8334990990990991, + "grad_norm": 25.92465591430664, + "learning_rate": 7.3323128321371e-07, + "loss": 6.845890998840332, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.2970545943826437, + "loss_total_v6": 0.42786817625164986, + "step": 3614 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8337297297297297, + "grad_norm": 26.679000854492188, + "learning_rate": 7.312525652902764e-07, + "loss": 7.084807395935059, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.31599991768598557, + "loss_total_v6": 0.44280046224594116, + "step": 3615 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8339603603603604, + "grad_norm": 23.906085968017578, + "learning_rate": 7.292763102234329e-07, + "loss": 7.173482894897461, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.3191617615520954, + "loss_total_v6": 0.448342677205801, + "step": 3616 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.834190990990991, + "grad_norm": 26.55779266357422, + "learning_rate": 7.273025191533839e-07, + "loss": 6.319185733795166, + "loss_alignment": 0.2532958984375, + "loss_alignment_w": 0.12664794921875, + "loss_sft": 0.26856057345867157, + "loss_total_v6": 0.3949491120874882, + "step": 3617 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8344216216216216, + "grad_norm": 23.957744598388672, + "learning_rate": 7.253311932189089e-07, + "loss": 7.727944374084473, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.35447177290916443, + "loss_total_v6": 0.48299654945731163, + "step": 3618 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8346522522522523, + "grad_norm": 30.64579963684082, + "learning_rate": 7.233623335573653e-07, + "loss": 6.703930854797363, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2891128435730934, + "loss_total_v6": 0.418995663523674, + "step": 3619 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8348828828828829, + "grad_norm": 23.65808868408203, + "learning_rate": 7.213959413046895e-07, + "loss": 7.004631996154785, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.30848652869462967, + "loss_total_v6": 0.43778951093554497, + "step": 3620 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8351135135135135, + "grad_norm": 22.38115882873535, + "learning_rate": 7.194320175953901e-07, + "loss": 6.717006683349609, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.2922036964446306, + "loss_total_v6": 0.4198129363358021, + "step": 3621 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8353441441441442, + "grad_norm": 20.660049438476562, + "learning_rate": 7.174705635625568e-07, + "loss": 7.183005332946777, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.31998581625521183, + "loss_total_v6": 0.4489378333091736, + "step": 3622 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8355747747747748, + "grad_norm": 23.837005615234375, + "learning_rate": 7.155115803378493e-07, + "loss": 6.759174346923828, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.29433560371398926, + "loss_total_v6": 0.42244839668273926, + "step": 3623 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8358054054054054, + "grad_norm": 28.58538246154785, + "learning_rate": 7.135550690515053e-07, + "loss": 7.055060386657715, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.312569085508585, + "loss_total_v6": 0.4409412555396557, + "step": 3624 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.836036036036036, + "grad_norm": 22.781301498413086, + "learning_rate": 7.116010308323357e-07, + "loss": 6.546496391296387, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.281775638461113, + "loss_total_v6": 0.409156009554863, + "step": 3625 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8362666666666667, + "grad_norm": 24.834239959716797, + "learning_rate": 7.096494668077237e-07, + "loss": 6.725299835205078, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2920048087835312, + "loss_total_v6": 0.4203312285244465, + "step": 3626 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8364972972972973, + "grad_norm": 24.014921188354492, + "learning_rate": 7.077003781036262e-07, + "loss": 7.300601482391357, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.32637425884604454, + "loss_total_v6": 0.45628758892416954, + "step": 3627 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8367279279279279, + "grad_norm": 24.585105895996094, + "learning_rate": 7.057537658445701e-07, + "loss": 6.975411415100098, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.3090558610856533, + "loss_total_v6": 0.4359632134437561, + "step": 3628 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8369585585585586, + "grad_norm": 27.691946029663086, + "learning_rate": 7.038096311536557e-07, + "loss": 7.021397590637207, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.30878669023513794, + "loss_total_v6": 0.43883734568953514, + "step": 3629 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8371891891891892, + "grad_norm": 25.62799644470215, + "learning_rate": 7.018679751525515e-07, + "loss": 6.485455513000488, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.27840312011539936, + "loss_total_v6": 0.4053409807384014, + "step": 3630 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8374198198198198, + "grad_norm": 27.859149932861328, + "learning_rate": 6.999287989614972e-07, + "loss": 6.843788146972656, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2978386767208576, + "loss_total_v6": 0.4277367554605007, + "step": 3631 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8376504504504505, + "grad_norm": 23.971935272216797, + "learning_rate": 6.979921036993042e-07, + "loss": 6.850063323974609, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.30036710947752, + "loss_total_v6": 0.4281289502978325, + "step": 3632 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8378810810810811, + "grad_norm": 24.69729232788086, + "learning_rate": 6.960578904833465e-07, + "loss": 6.673830986022949, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.28868119791150093, + "loss_total_v6": 0.41711442917585373, + "step": 3633 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8381117117117117, + "grad_norm": 26.95261001586914, + "learning_rate": 6.941261604295713e-07, + "loss": 6.917039394378662, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.30131828412413597, + "loss_total_v6": 0.432314969599247, + "step": 3634 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8383423423423424, + "grad_norm": 28.320018768310547, + "learning_rate": 6.921969146524921e-07, + "loss": 6.986429214477539, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.30675377510488033, + "loss_total_v6": 0.4366518259048462, + "step": 3635 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.838572972972973, + "grad_norm": 25.888280868530273, + "learning_rate": 6.902701542651874e-07, + "loss": 7.074650764465332, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31065015867352486, + "loss_total_v6": 0.44216566532850266, + "step": 3636 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8388036036036036, + "grad_norm": 26.402408599853516, + "learning_rate": 6.883458803793053e-07, + "loss": 6.560521125793457, + "loss_alignment": 0.250244140625, + "loss_alignment_w": 0.1251220703125, + "loss_sft": 0.28483419492840767, + "loss_total_v6": 0.41003255546092987, + "step": 3637 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8390342342342343, + "grad_norm": 26.126930236816406, + "learning_rate": 6.864240941050532e-07, + "loss": 7.144646644592285, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.3189616799354553, + "loss_total_v6": 0.446540430188179, + "step": 3638 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8392648648648648, + "grad_norm": 24.759899139404297, + "learning_rate": 6.845047965512103e-07, + "loss": 6.953133583068848, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.3059697337448597, + "loss_total_v6": 0.4345708377659321, + "step": 3639 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8394954954954955, + "grad_norm": 25.04482650756836, + "learning_rate": 6.825879888251135e-07, + "loss": 7.018496513366699, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.31063477136194706, + "loss_total_v6": 0.4386560134589672, + "step": 3640 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8397261261261261, + "grad_norm": 24.63793182373047, + "learning_rate": 6.80673672032669e-07, + "loss": 7.0157976150512695, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.31005413830280304, + "loss_total_v6": 0.43848735839128494, + "step": 3641 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8399567567567567, + "grad_norm": 24.578800201416016, + "learning_rate": 6.787618472783408e-07, + "loss": 6.419637203216553, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.27247367054224014, + "loss_total_v6": 0.40122732147574425, + "step": 3642 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8401873873873874, + "grad_norm": 23.60378074645996, + "learning_rate": 6.768525156651589e-07, + "loss": 7.008973121643066, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3101311270147562, + "loss_total_v6": 0.43806080892682076, + "step": 3643 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.840418018018018, + "grad_norm": 28.411306381225586, + "learning_rate": 6.749456782947123e-07, + "loss": 6.896005630493164, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.29924074187874794, + "loss_total_v6": 0.43100038170814514, + "step": 3644 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8406486486486486, + "grad_norm": 23.52851676940918, + "learning_rate": 6.730413362671534e-07, + "loss": 7.0999627113342285, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3142005614936352, + "loss_total_v6": 0.4437476731836796, + "step": 3645 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8408792792792793, + "grad_norm": 26.209402084350586, + "learning_rate": 6.711394906811935e-07, + "loss": 7.047774791717529, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.3119306303560734, + "loss_total_v6": 0.4404859282076359, + "step": 3646 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8411099099099099, + "grad_norm": 25.57771873474121, + "learning_rate": 6.692401426341022e-07, + "loss": 7.326681137084961, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.32861459627747536, + "loss_total_v6": 0.45791758969426155, + "step": 3647 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8413405405405405, + "grad_norm": 24.557231903076172, + "learning_rate": 6.673432932217117e-07, + "loss": 7.1425065994262695, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.31827861070632935, + "loss_total_v6": 0.44640666991472244, + "step": 3648 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8415711711711712, + "grad_norm": 26.25345802307129, + "learning_rate": 6.654489435384081e-07, + "loss": 7.362911701202393, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.33023813739418983, + "loss_total_v6": 0.46018198505043983, + "step": 3649 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8418018018018018, + "grad_norm": 22.707674026489258, + "learning_rate": 6.635570946771386e-07, + "loss": 7.171980857849121, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.31969350576400757, + "loss_total_v6": 0.44824880734086037, + "step": 3650 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8420324324324324, + "grad_norm": 22.77420997619629, + "learning_rate": 6.61667747729407e-07, + "loss": 7.2071967124938965, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.32023130916059017, + "loss_total_v6": 0.45044979453086853, + "step": 3651 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8422630630630631, + "grad_norm": 30.615888595581055, + "learning_rate": 6.597809037852726e-07, + "loss": 7.253087997436523, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.32201611809432507, + "loss_total_v6": 0.4533179961144924, + "step": 3652 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8424936936936936, + "grad_norm": 24.701248168945312, + "learning_rate": 6.578965639333534e-07, + "loss": 7.0851664543151855, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.31266545690596104, + "loss_total_v6": 0.4428229108452797, + "step": 3653 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8427243243243243, + "grad_norm": 27.383806228637695, + "learning_rate": 6.560147292608177e-07, + "loss": 6.843727111816406, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.3006577529013157, + "loss_total_v6": 0.4277329482138157, + "step": 3654 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.842954954954955, + "grad_norm": 27.260217666625977, + "learning_rate": 6.541354008533934e-07, + "loss": 7.104913711547852, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.3138386309146881, + "loss_total_v6": 0.4440571293234825, + "step": 3655 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8431855855855855, + "grad_norm": 23.93442726135254, + "learning_rate": 6.522585797953579e-07, + "loss": 6.669936656951904, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2876901365816593, + "loss_total_v6": 0.4168710485100746, + "step": 3656 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8434162162162162, + "grad_norm": 26.203575134277344, + "learning_rate": 6.503842671695453e-07, + "loss": 7.029385566711426, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.31297856010496616, + "loss_total_v6": 0.4393365867435932, + "step": 3657 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8436468468468469, + "grad_norm": 25.474483489990234, + "learning_rate": 6.485124640573437e-07, + "loss": 7.046411991119385, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.31256260722875595, + "loss_total_v6": 0.44040074571967125, + "step": 3658 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8438774774774774, + "grad_norm": 29.7011775970459, + "learning_rate": 6.466431715386873e-07, + "loss": 6.732958793640137, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.28958435729146004, + "loss_total_v6": 0.42080993577837944, + "step": 3659 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8441081081081081, + "grad_norm": 24.66541290283203, + "learning_rate": 6.447763906920679e-07, + "loss": 7.067816734313965, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3137630708515644, + "loss_total_v6": 0.4417385384440422, + "step": 3660 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8443387387387388, + "grad_norm": 29.04910659790039, + "learning_rate": 6.429121225945256e-07, + "loss": 6.438215255737305, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.27285661175847054, + "loss_total_v6": 0.40238846838474274, + "step": 3661 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8445693693693693, + "grad_norm": 28.60841941833496, + "learning_rate": 6.410503683216513e-07, + "loss": 6.760163307189941, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.2949619758874178, + "loss_total_v6": 0.42251018807291985, + "step": 3662 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8448, + "grad_norm": 24.17282485961914, + "learning_rate": 6.391911289475855e-07, + "loss": 7.491486549377441, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3395558334887028, + "loss_total_v6": 0.4682179391384125, + "step": 3663 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8450306306306307, + "grad_norm": 28.99001121520996, + "learning_rate": 6.373344055450165e-07, + "loss": 6.951651573181152, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3067468963563442, + "loss_total_v6": 0.4344782195985317, + "step": 3664 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8452612612612612, + "grad_norm": 23.776430130004883, + "learning_rate": 6.354801991851839e-07, + "loss": 6.728631973266602, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.292380940169096, + "loss_total_v6": 0.4205395244061947, + "step": 3665 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8454918918918919, + "grad_norm": 25.14780044555664, + "learning_rate": 6.336285109378704e-07, + "loss": 6.994056224822998, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.309504009783268, + "loss_total_v6": 0.4371285140514374, + "step": 3666 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8457225225225226, + "grad_norm": 23.735855102539062, + "learning_rate": 6.317793418714119e-07, + "loss": 7.163869380950928, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3167756497859955, + "loss_total_v6": 0.4477418325841427, + "step": 3667 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8459531531531531, + "grad_norm": 24.979827880859375, + "learning_rate": 6.29932693052685e-07, + "loss": 7.450014114379883, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3348122797906399, + "loss_total_v6": 0.4656258821487427, + "step": 3668 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8461837837837838, + "grad_norm": 24.430307388305664, + "learning_rate": 6.280885655471158e-07, + "loss": 6.931056499481201, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.304971419274807, + "loss_total_v6": 0.4331910163164139, + "step": 3669 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8464144144144145, + "grad_norm": 26.150604248046875, + "learning_rate": 6.262469604186755e-07, + "loss": 6.95242166519165, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.30606262385845184, + "loss_total_v6": 0.43452636152505875, + "step": 3670 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.846645045045045, + "grad_norm": 25.426918029785156, + "learning_rate": 6.244078787298791e-07, + "loss": 7.051987648010254, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.31269746273756027, + "loss_total_v6": 0.4407492093741894, + "step": 3671 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8468756756756757, + "grad_norm": 23.031827926635742, + "learning_rate": 6.225713215417866e-07, + "loss": 7.0518903732299805, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.3136526755988598, + "loss_total_v6": 0.4407431334257126, + "step": 3672 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8471063063063063, + "grad_norm": 25.376989364624023, + "learning_rate": 6.20737289913999e-07, + "loss": 6.895979881286621, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.30189413018524647, + "loss_total_v6": 0.4309987351298332, + "step": 3673 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8473369369369369, + "grad_norm": 25.101139068603516, + "learning_rate": 6.189057849046648e-07, + "loss": 7.103586196899414, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.3167921043932438, + "loss_total_v6": 0.4439741149544716, + "step": 3674 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8475675675675676, + "grad_norm": 23.978282928466797, + "learning_rate": 6.170768075704692e-07, + "loss": 6.547319412231445, + "loss_alignment": 0.2510986328125, + "loss_alignment_w": 0.12554931640625, + "loss_sft": 0.2834139931946993, + "loss_total_v6": 0.40920744091272354, + "step": 3675 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8477981981981982, + "grad_norm": 26.567707061767578, + "learning_rate": 6.152503589666426e-07, + "loss": 6.475890159606934, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2760352548211813, + "loss_total_v6": 0.40474313497543335, + "step": 3676 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8480288288288288, + "grad_norm": 26.11138343811035, + "learning_rate": 6.134264401469564e-07, + "loss": 6.7169599533081055, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2877451963722706, + "loss_total_v6": 0.4198100082576275, + "step": 3677 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8482594594594595, + "grad_norm": 23.2789249420166, + "learning_rate": 6.116050521637218e-07, + "loss": 6.678368091583252, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.28656917810440063, + "loss_total_v6": 0.41739802062511444, + "step": 3678 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8484900900900901, + "grad_norm": 27.08448600769043, + "learning_rate": 6.097861960677898e-07, + "loss": 7.204598903656006, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.32293757423758507, + "loss_total_v6": 0.45028743892908096, + "step": 3679 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8487207207207207, + "grad_norm": 25.131393432617188, + "learning_rate": 6.079698729085498e-07, + "loss": 7.079307556152344, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.31496952287852764, + "loss_total_v6": 0.4424567259848118, + "step": 3680 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8489513513513514, + "grad_norm": 24.311481475830078, + "learning_rate": 6.06156083733932e-07, + "loss": 7.485112190246582, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3374789319932461, + "loss_total_v6": 0.4678195230662823, + "step": 3681 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.849181981981982, + "grad_norm": 21.96382713317871, + "learning_rate": 6.043448295904009e-07, + "loss": 6.955979824066162, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.30686483159661293, + "loss_total_v6": 0.43474873527884483, + "step": 3682 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8494126126126126, + "grad_norm": 22.453977584838867, + "learning_rate": 6.025361115229628e-07, + "loss": 7.325284957885742, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3272913880646229, + "loss_total_v6": 0.4578303247690201, + "step": 3683 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8496432432432433, + "grad_norm": 24.716493606567383, + "learning_rate": 6.007299305751585e-07, + "loss": 6.778315544128418, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2952572740614414, + "loss_total_v6": 0.423644732683897, + "step": 3684 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8498738738738739, + "grad_norm": 25.670467376708984, + "learning_rate": 5.989262877890645e-07, + "loss": 7.015841960906982, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.3084242269396782, + "loss_total_v6": 0.4384901374578476, + "step": 3685 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8501045045045045, + "grad_norm": 23.65933609008789, + "learning_rate": 5.971251842052933e-07, + "loss": 7.12101936340332, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.31673727184534073, + "loss_total_v6": 0.4450636953115463, + "step": 3686 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8503351351351351, + "grad_norm": 26.067825317382812, + "learning_rate": 5.953266208629943e-07, + "loss": 6.391852378845215, + "loss_alignment": 0.248779296875, + "loss_alignment_w": 0.1243896484375, + "loss_sft": 0.27507059648633003, + "loss_total_v6": 0.39949075505137444, + "step": 3687 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8505657657657658, + "grad_norm": 25.202632904052734, + "learning_rate": 5.935305987998497e-07, + "loss": 6.9316301345825195, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.3061821945011616, + "loss_total_v6": 0.43322688341140747, + "step": 3688 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8507963963963964, + "grad_norm": 24.054147720336914, + "learning_rate": 5.917371190520737e-07, + "loss": 6.933030128479004, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.30539995059370995, + "loss_total_v6": 0.43331439048051834, + "step": 3689 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.851027027027027, + "grad_norm": 24.614608764648438, + "learning_rate": 5.899461826544173e-07, + "loss": 7.2823166847229, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.32732193544507027, + "loss_total_v6": 0.4551447965204716, + "step": 3690 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8512576576576577, + "grad_norm": 24.39618682861328, + "learning_rate": 5.881577906401625e-07, + "loss": 6.915874481201172, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3048312570899725, + "loss_total_v6": 0.43224216625094414, + "step": 3691 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8514882882882883, + "grad_norm": 25.41626739501953, + "learning_rate": 5.863719440411214e-07, + "loss": 7.417600154876709, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3311690017580986, + "loss_total_v6": 0.4636000208556652, + "step": 3692 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8517189189189189, + "grad_norm": 23.777490615844727, + "learning_rate": 5.845886438876424e-07, + "loss": 6.754664421081543, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.2932145204395056, + "loss_total_v6": 0.4221665523946285, + "step": 3693 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8519495495495496, + "grad_norm": 22.9741153717041, + "learning_rate": 5.828078912085983e-07, + "loss": 6.6507978439331055, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.2875621188431978, + "loss_total_v6": 0.4156748987734318, + "step": 3694 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8521801801801802, + "grad_norm": 26.466100692749023, + "learning_rate": 5.810296870313964e-07, + "loss": 6.7678046226501465, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.2962635587900877, + "loss_total_v6": 0.42298778519034386, + "step": 3695 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8524108108108108, + "grad_norm": 23.670915603637695, + "learning_rate": 5.792540323819751e-07, + "loss": 6.484309673309326, + "loss_alignment": 0.250244140625, + "loss_alignment_w": 0.1251220703125, + "loss_sft": 0.2801778018474579, + "loss_total_v6": 0.4052693508565426, + "step": 3696 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8526414414414415, + "grad_norm": 23.41657829284668, + "learning_rate": 5.774809282847971e-07, + "loss": 7.084945201873779, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3143148235976696, + "loss_total_v6": 0.4428090788424015, + "step": 3697 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.852872072072072, + "grad_norm": 25.140445709228516, + "learning_rate": 5.757103757628573e-07, + "loss": 7.288825988769531, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3251195028424263, + "loss_total_v6": 0.4555516317486763, + "step": 3698 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8531027027027027, + "grad_norm": 22.03139877319336, + "learning_rate": 5.739423758376761e-07, + "loss": 7.2159504890441895, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.32103780284523964, + "loss_total_v6": 0.45099690183997154, + "step": 3699 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8533333333333334, + "grad_norm": 25.505624771118164, + "learning_rate": 5.721769295293033e-07, + "loss": 7.093979835510254, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.31367404013872147, + "loss_total_v6": 0.44337374716997147, + "step": 3700 + }, + { + "epoch": 0.8533333333333334, + "eval_loss": 0.5646238923072815, + "eval_loss_alignment": 0.2316013841324201, + "eval_loss_alignment_w": 0.11580069206621005, + "eval_loss_sft": 0.4488231975023877, + "eval_loss_total_v6": 0.5646238898407651, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 92.14, + "eval_samples_per_second": 18.939, + "eval_steps_per_second": 2.377, + "step": 3700 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.853563963963964, + "grad_norm": 26.5705509185791, + "learning_rate": 5.704140378563128e-07, + "loss": 6.78133487701416, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.2955375388264656, + "loss_total_v6": 0.4238334372639656, + "step": 3701 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8537945945945946, + "grad_norm": 23.417078018188477, + "learning_rate": 5.686537018358063e-07, + "loss": 6.905818939208984, + "loss_alignment": 0.2523193359375, + "loss_alignment_w": 0.12615966796875, + "loss_sft": 0.3054387830197811, + "loss_total_v6": 0.4316136911511421, + "step": 3702 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8540252252252252, + "grad_norm": 24.222850799560547, + "learning_rate": 5.668959224834119e-07, + "loss": 7.4760942459106445, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.3348706290125847, + "loss_total_v6": 0.4672558940947056, + "step": 3703 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8542558558558558, + "grad_norm": 25.092817306518555, + "learning_rate": 5.651407008132808e-07, + "loss": 7.0050368309021, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.31055648624897003, + "loss_total_v6": 0.43781479075551033, + "step": 3704 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8544864864864865, + "grad_norm": 26.921302795410156, + "learning_rate": 5.633880378380907e-07, + "loss": 7.631189346313477, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.3454338423907757, + "loss_total_v6": 0.4769493229687214, + "step": 3705 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8547171171171171, + "grad_norm": 23.966873168945312, + "learning_rate": 5.616379345690398e-07, + "loss": 7.022566795349121, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.3084630202502012, + "loss_total_v6": 0.4389104023575783, + "step": 3706 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8549477477477477, + "grad_norm": 26.101646423339844, + "learning_rate": 5.598903920158521e-07, + "loss": 6.5112833976745605, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.27862877398729324, + "loss_total_v6": 0.40695521235466003, + "step": 3707 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8551783783783784, + "grad_norm": 27.741819381713867, + "learning_rate": 5.581454111867729e-07, + "loss": 7.566237449645996, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3423966560512781, + "loss_total_v6": 0.47288982570171356, + "step": 3708 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.855409009009009, + "grad_norm": 28.35756492614746, + "learning_rate": 5.5640299308857e-07, + "loss": 6.768070220947266, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.29727196879684925, + "loss_total_v6": 0.4230043813586235, + "step": 3709 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8556396396396396, + "grad_norm": 27.40414047241211, + "learning_rate": 5.546631387265344e-07, + "loss": 6.567948341369629, + "loss_alignment": 0.252197265625, + "loss_alignment_w": 0.1260986328125, + "loss_sft": 0.28444389067590237, + "loss_total_v6": 0.4104967527091503, + "step": 3710 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8558702702702703, + "grad_norm": 27.748788833618164, + "learning_rate": 5.529258491044737e-07, + "loss": 7.234583854675293, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.322217658162117, + "loss_total_v6": 0.4521614946424961, + "step": 3711 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8561009009009009, + "grad_norm": 26.3800048828125, + "learning_rate": 5.511911252247193e-07, + "loss": 6.718326091766357, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2917520608752966, + "loss_total_v6": 0.41989537701010704, + "step": 3712 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8563315315315315, + "grad_norm": 25.031089782714844, + "learning_rate": 5.494589680881213e-07, + "loss": 6.776578903198242, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.2970713395625353, + "loss_total_v6": 0.42353618517518044, + "step": 3713 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8565621621621622, + "grad_norm": 24.08017921447754, + "learning_rate": 5.477293786940502e-07, + "loss": 7.269590854644775, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.32547369226813316, + "loss_total_v6": 0.45434943214058876, + "step": 3714 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8567927927927927, + "grad_norm": 23.001855850219727, + "learning_rate": 5.460023580403911e-07, + "loss": 6.426743984222412, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.2752524595707655, + "loss_total_v6": 0.40167151391506195, + "step": 3715 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8570234234234234, + "grad_norm": 26.819412231445312, + "learning_rate": 5.442779071235515e-07, + "loss": 7.084920406341553, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.31461843103170395, + "loss_total_v6": 0.44280752912163734, + "step": 3716 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8572540540540541, + "grad_norm": 23.128602981567383, + "learning_rate": 5.425560269384544e-07, + "loss": 6.689974308013916, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.29043788090348244, + "loss_total_v6": 0.41812340915203094, + "step": 3717 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8574846846846846, + "grad_norm": 25.29380989074707, + "learning_rate": 5.408367184785391e-07, + "loss": 7.178973197937012, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.3204662147909403, + "loss_total_v6": 0.44868582487106323, + "step": 3718 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8577153153153153, + "grad_norm": 22.551122665405273, + "learning_rate": 5.391199827357624e-07, + "loss": 7.3132548332214355, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3264326713979244, + "loss_total_v6": 0.4570784270763397, + "step": 3719 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.857945945945946, + "grad_norm": 24.93827247619629, + "learning_rate": 5.374058207005945e-07, + "loss": 7.034459114074707, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3084281086921692, + "loss_total_v6": 0.4396537020802498, + "step": 3720 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8581765765765765, + "grad_norm": 25.508134841918945, + "learning_rate": 5.35694233362023e-07, + "loss": 6.5675482749938965, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.2842510659247637, + "loss_total_v6": 0.41047175973653793, + "step": 3721 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8584072072072072, + "grad_norm": 27.484899520874023, + "learning_rate": 5.339852217075508e-07, + "loss": 6.735455513000488, + "loss_alignment": 0.2518310546875, + "loss_alignment_w": 0.12591552734375, + "loss_sft": 0.29500467516481876, + "loss_total_v6": 0.4209659770131111, + "step": 3722 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8586378378378379, + "grad_norm": 27.55447006225586, + "learning_rate": 5.322787867231921e-07, + "loss": 6.777713775634766, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.2954485546797514, + "loss_total_v6": 0.42360712587833405, + "step": 3723 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8588684684684684, + "grad_norm": 26.923446655273438, + "learning_rate": 5.305749293934764e-07, + "loss": 7.207718372344971, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3224459085613489, + "loss_total_v6": 0.45048239827156067, + "step": 3724 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8590990990990991, + "grad_norm": 23.682950973510742, + "learning_rate": 5.288736507014436e-07, + "loss": 6.828892707824707, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.3002188839018345, + "loss_total_v6": 0.4268057942390442, + "step": 3725 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8593297297297298, + "grad_norm": 24.810585021972656, + "learning_rate": 5.271749516286501e-07, + "loss": 6.9312825202941895, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3046651128679514, + "loss_total_v6": 0.43320515751838684, + "step": 3726 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8595603603603603, + "grad_norm": 23.27100944519043, + "learning_rate": 5.254788331551586e-07, + "loss": 6.565178871154785, + "loss_alignment": 0.249755859375, + "loss_alignment_w": 0.1248779296875, + "loss_sft": 0.2854457385838032, + "loss_total_v6": 0.4103236682713032, + "step": 3727 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.859790990990991, + "grad_norm": 25.012968063354492, + "learning_rate": 5.237852962595469e-07, + "loss": 6.807279586791992, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.298974871635437, + "loss_total_v6": 0.4254549667239189, + "step": 3728 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8600216216216217, + "grad_norm": 27.806997299194336, + "learning_rate": 5.220943419189028e-07, + "loss": 6.638164520263672, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.28527713380753994, + "loss_total_v6": 0.4148852676153183, + "step": 3729 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8602522522522522, + "grad_norm": 28.266387939453125, + "learning_rate": 5.204059711088228e-07, + "loss": 6.925800323486328, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.3028881512582302, + "loss_total_v6": 0.4328625313937664, + "step": 3730 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8604828828828829, + "grad_norm": 23.803743362426758, + "learning_rate": 5.187201848034146e-07, + "loss": 6.602242469787598, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.28467997163534164, + "loss_total_v6": 0.41264016926288605, + "step": 3731 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8607135135135136, + "grad_norm": 26.52587127685547, + "learning_rate": 5.170369839752926e-07, + "loss": 6.570229530334473, + "loss_alignment": 0.2509765625, + "loss_alignment_w": 0.12548828125, + "loss_sft": 0.28524261340498924, + "loss_total_v6": 0.41063934564590454, + "step": 3732 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8609441441441441, + "grad_norm": 25.412763595581055, + "learning_rate": 5.153563695955816e-07, + "loss": 6.927801609039307, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3049358185380697, + "loss_total_v6": 0.43298759683966637, + "step": 3733 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8611747747747748, + "grad_norm": 25.986392974853516, + "learning_rate": 5.136783426339115e-07, + "loss": 7.019804000854492, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.31146417185664177, + "loss_total_v6": 0.43873773887753487, + "step": 3734 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8614054054054054, + "grad_norm": 25.700481414794922, + "learning_rate": 5.120029040584217e-07, + "loss": 6.834593296051025, + "loss_alignment": 0.2646484375, + "loss_alignment_w": 0.13232421875, + "loss_sft": 0.2948531210422516, + "loss_total_v6": 0.4271620810031891, + "step": 3735 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.861636036036036, + "grad_norm": 22.091581344604492, + "learning_rate": 5.103300548357592e-07, + "loss": 6.9863786697387695, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.30655223317444324, + "loss_total_v6": 0.4366486519575119, + "step": 3736 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8618666666666667, + "grad_norm": 23.206153869628906, + "learning_rate": 5.086597959310735e-07, + "loss": 6.881099700927734, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.2999265305697918, + "loss_total_v6": 0.4300687536597252, + "step": 3737 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8620972972972973, + "grad_norm": 27.077655792236328, + "learning_rate": 5.069921283080214e-07, + "loss": 6.675267696380615, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.2899306695908308, + "loss_total_v6": 0.41720422729849815, + "step": 3738 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8623279279279279, + "grad_norm": 22.704177856445312, + "learning_rate": 5.053270529287663e-07, + "loss": 7.376739501953125, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.3329334519803524, + "loss_total_v6": 0.4610462300479412, + "step": 3739 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8625585585585586, + "grad_norm": 27.13896369934082, + "learning_rate": 5.036645707539745e-07, + "loss": 7.452963829040527, + "loss_alignment": 0.26611328125, + "loss_alignment_w": 0.133056640625, + "loss_sft": 0.33272305130958557, + "loss_total_v6": 0.46581022068858147, + "step": 3740 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8627891891891892, + "grad_norm": 23.69502067565918, + "learning_rate": 5.020046827428149e-07, + "loss": 7.1852827072143555, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3194415047764778, + "loss_total_v6": 0.4490801952779293, + "step": 3741 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8630198198198198, + "grad_norm": 23.764089584350586, + "learning_rate": 5.00347389852961e-07, + "loss": 7.431838035583496, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3367280475795269, + "loss_total_v6": 0.4644898846745491, + "step": 3742 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8632504504504505, + "grad_norm": 24.05132484436035, + "learning_rate": 4.986926930405905e-07, + "loss": 6.959986686706543, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3047501277178526, + "loss_total_v6": 0.43499916046857834, + "step": 3743 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8634810810810811, + "grad_norm": 24.842802047729492, + "learning_rate": 4.970405932603806e-07, + "loss": 6.976841926574707, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.3065360374748707, + "loss_total_v6": 0.4360526390373707, + "step": 3744 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8637117117117117, + "grad_norm": 26.63590431213379, + "learning_rate": 4.953910914655119e-07, + "loss": 6.635533332824707, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.28663857094943523, + "loss_total_v6": 0.4147208444774151, + "step": 3745 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8639423423423424, + "grad_norm": 24.605375289916992, + "learning_rate": 4.937441886076633e-07, + "loss": 6.592228412628174, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.2849238235503435, + "loss_total_v6": 0.41201428323984146, + "step": 3746 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.864172972972973, + "grad_norm": 23.842884063720703, + "learning_rate": 4.920998856370207e-07, + "loss": 6.676315784454346, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.2863645851612091, + "loss_total_v6": 0.4172697514295578, + "step": 3747 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8644036036036036, + "grad_norm": 23.929407119750977, + "learning_rate": 4.90458183502262e-07, + "loss": 6.742599964141846, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.28997329995036125, + "loss_total_v6": 0.42141250148415565, + "step": 3748 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8646342342342342, + "grad_norm": 25.324684143066406, + "learning_rate": 4.888190831505701e-07, + "loss": 6.99046516418457, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3072348553687334, + "loss_total_v6": 0.43690406158566475, + "step": 3749 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8648648648648649, + "grad_norm": 24.515625, + "learning_rate": 4.871825855276252e-07, + "loss": 7.006234169006348, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3076101131737232, + "loss_total_v6": 0.4378896541893482, + "step": 3750 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8650954954954955, + "grad_norm": 23.431137084960938, + "learning_rate": 4.855486915776048e-07, + "loss": 7.117703437805176, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3133867643773556, + "loss_total_v6": 0.4448564909398556, + "step": 3751 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8653261261261261, + "grad_norm": 27.03569984436035, + "learning_rate": 4.839174022431859e-07, + "loss": 7.357032299041748, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.32999273017048836, + "loss_total_v6": 0.45981451869010925, + "step": 3752 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8655567567567568, + "grad_norm": 24.042583465576172, + "learning_rate": 4.822887184655406e-07, + "loss": 6.8658881187438965, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3006390053778887, + "loss_total_v6": 0.42911800742149353, + "step": 3753 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8657873873873874, + "grad_norm": 25.424341201782227, + "learning_rate": 4.806626411843401e-07, + "loss": 7.192703723907471, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.3221941143274307, + "loss_total_v6": 0.4495439752936363, + "step": 3754 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.866018018018018, + "grad_norm": 23.116331100463867, + "learning_rate": 4.790391713377501e-07, + "loss": 7.10693359375, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.3131561242043972, + "loss_total_v6": 0.4441833645105362, + "step": 3755 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8662486486486487, + "grad_norm": 22.668262481689453, + "learning_rate": 4.774183098624335e-07, + "loss": 6.992043495178223, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.30628063902258873, + "loss_total_v6": 0.4370026960968971, + "step": 3756 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8664792792792793, + "grad_norm": 26.169233322143555, + "learning_rate": 4.758000576935473e-07, + "loss": 7.337345600128174, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.32871654629707336, + "loss_total_v6": 0.45858411490917206, + "step": 3757 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8667099099099099, + "grad_norm": 21.69467544555664, + "learning_rate": 4.7418441576474185e-07, + "loss": 7.036355972290039, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.31196461245417595, + "loss_total_v6": 0.43977221846580505, + "step": 3758 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8669405405405406, + "grad_norm": 24.935203552246094, + "learning_rate": 4.725713850081648e-07, + "loss": 7.08836555480957, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.31312477216124535, + "loss_total_v6": 0.44302285090088844, + "step": 3759 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8671711711711712, + "grad_norm": 23.28136444091797, + "learning_rate": 4.709609663544534e-07, + "loss": 6.904941558837891, + "loss_alignment": 0.2510986328125, + "loss_alignment_w": 0.12554931640625, + "loss_sft": 0.3062231782823801, + "loss_total_v6": 0.43155886977910995, + "step": 3760 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8674018018018018, + "grad_norm": 23.744850158691406, + "learning_rate": 4.6935316073274007e-07, + "loss": 6.726157188415527, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2918142694979906, + "loss_total_v6": 0.42038482427597046, + "step": 3761 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8676324324324325, + "grad_norm": 25.882314682006836, + "learning_rate": 4.67747969070651e-07, + "loss": 7.007680892944336, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3072427175939083, + "loss_total_v6": 0.4379800409078598, + "step": 3762 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.867863063063063, + "grad_norm": 27.602283477783203, + "learning_rate": 4.661453922943004e-07, + "loss": 6.923540115356445, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3037234600633383, + "loss_total_v6": 0.43272126093506813, + "step": 3763 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8680936936936937, + "grad_norm": 21.86226463317871, + "learning_rate": 4.6454543132829653e-07, + "loss": 7.038724899291992, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.3118075132369995, + "loss_total_v6": 0.4399203136563301, + "step": 3764 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8683243243243243, + "grad_norm": 25.84218978881836, + "learning_rate": 4.629480870957381e-07, + "loss": 6.964109420776367, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.30561818927526474, + "loss_total_v6": 0.43525686115026474, + "step": 3765 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8685549549549549, + "grad_norm": 25.173540115356445, + "learning_rate": 4.6135336051821523e-07, + "loss": 6.923623561859131, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.30400334671139717, + "loss_total_v6": 0.432726476341486, + "step": 3766 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8687855855855856, + "grad_norm": 24.980241775512695, + "learning_rate": 4.597612525158046e-07, + "loss": 6.463523864746094, + "loss_alignment": 0.2474365234375, + "loss_alignment_w": 0.12371826171875, + "loss_sft": 0.28014514222741127, + "loss_total_v6": 0.40397022664546967, + "step": 3767 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8690162162162162, + "grad_norm": 22.82219123840332, + "learning_rate": 4.581717640070743e-07, + "loss": 6.947015762329102, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.3071590829640627, + "loss_total_v6": 0.43418847396969795, + "step": 3768 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8692468468468468, + "grad_norm": 23.90606689453125, + "learning_rate": 4.565848959090818e-07, + "loss": 6.7513017654418945, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.29269915260374546, + "loss_total_v6": 0.4219563603401184, + "step": 3769 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8694774774774775, + "grad_norm": 26.370098114013672, + "learning_rate": 4.550006491373704e-07, + "loss": 6.888486862182617, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.3042486645281315, + "loss_total_v6": 0.4305304065346718, + "step": 3770 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8697081081081081, + "grad_norm": 24.402772903442383, + "learning_rate": 4.5341902460597277e-07, + "loss": 6.8425798416137695, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.30067760683596134, + "loss_total_v6": 0.4276612587273121, + "step": 3771 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8699387387387387, + "grad_norm": 25.917530059814453, + "learning_rate": 4.5184002322740784e-07, + "loss": 6.8653364181518555, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.3001162149012089, + "loss_total_v6": 0.4290835075080395, + "step": 3772 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8701693693693694, + "grad_norm": 23.789491653442383, + "learning_rate": 4.5026364591268303e-07, + "loss": 6.787283897399902, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.2972521260380745, + "loss_total_v6": 0.4242052547633648, + "step": 3773 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8704, + "grad_norm": 24.7706356048584, + "learning_rate": 4.486898935712886e-07, + "loss": 6.838214874267578, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.29866527020931244, + "loss_total_v6": 0.42738841474056244, + "step": 3774 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8706306306306306, + "grad_norm": 22.890174865722656, + "learning_rate": 4.4711876711120206e-07, + "loss": 6.702202796936035, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2907748557627201, + "loss_total_v6": 0.418887659907341, + "step": 3775 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8708612612612613, + "grad_norm": 27.214948654174805, + "learning_rate": 4.455502674388873e-07, + "loss": 7.01826286315918, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.3093231990933418, + "loss_total_v6": 0.4386414401233196, + "step": 3776 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8710918918918918, + "grad_norm": 27.604339599609375, + "learning_rate": 4.4398439545928926e-07, + "loss": 7.139754772186279, + "loss_alignment": 0.2506103515625, + "loss_alignment_w": 0.12530517578125, + "loss_sft": 0.32080743461847305, + "loss_total_v6": 0.44623467326164246, + "step": 3777 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8713225225225225, + "grad_norm": 26.35548210144043, + "learning_rate": 4.424211520758404e-07, + "loss": 7.034242153167725, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.31254967674613, + "loss_total_v6": 0.4396401308476925, + "step": 3778 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8715531531531532, + "grad_norm": 26.559568405151367, + "learning_rate": 4.4086053819045407e-07, + "loss": 7.021899223327637, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3069107010960579, + "loss_total_v6": 0.4388687126338482, + "step": 3779 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8717837837837837, + "grad_norm": 25.03171730041504, + "learning_rate": 4.393025547035273e-07, + "loss": 6.884341239929199, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.30253998935222626, + "loss_total_v6": 0.43027131259441376, + "step": 3780 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8720144144144144, + "grad_norm": 21.6591854095459, + "learning_rate": 4.3774720251394007e-07, + "loss": 6.806836128234863, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.2954224105924368, + "loss_total_v6": 0.42542728781700134, + "step": 3781 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8722450450450451, + "grad_norm": 29.422775268554688, + "learning_rate": 4.3619448251905395e-07, + "loss": 7.271848201751709, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3240736424922943, + "loss_total_v6": 0.4544905163347721, + "step": 3782 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8724756756756756, + "grad_norm": 28.054149627685547, + "learning_rate": 4.346443956147123e-07, + "loss": 6.947972774505615, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3060592021793127, + "loss_total_v6": 0.43424829840660095, + "step": 3783 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8727063063063063, + "grad_norm": 25.332744598388672, + "learning_rate": 4.330969426952375e-07, + "loss": 6.80726957321167, + "loss_alignment": 0.25, + "loss_alignment_w": 0.125, + "loss_sft": 0.3002864718437195, + "loss_total_v6": 0.4254543334245682, + "step": 3784 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.872936936936937, + "grad_norm": 27.494150161743164, + "learning_rate": 4.315521246534349e-07, + "loss": 7.03122615814209, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.3125137146562338, + "loss_total_v6": 0.4394516199827194, + "step": 3785 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8731675675675675, + "grad_norm": 24.2012996673584, + "learning_rate": 4.300099423805865e-07, + "loss": 6.719102382659912, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2919226549565792, + "loss_total_v6": 0.4199438914656639, + "step": 3786 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8733981981981982, + "grad_norm": 27.11763572692871, + "learning_rate": 4.2847039676645664e-07, + "loss": 7.151797771453857, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3181421421468258, + "loss_total_v6": 0.4469873607158661, + "step": 3787 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8736288288288289, + "grad_norm": 24.826791763305664, + "learning_rate": 4.269334886992876e-07, + "loss": 6.854162693023682, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.2988533116877079, + "loss_total_v6": 0.4283851720392704, + "step": 3788 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8738594594594594, + "grad_norm": 25.887434005737305, + "learning_rate": 4.2539921906579795e-07, + "loss": 6.630545616149902, + "loss_alignment": 0.2503662109375, + "loss_alignment_w": 0.12518310546875, + "loss_sft": 0.28910390101373196, + "loss_total_v6": 0.4144090935587883, + "step": 3789 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8740900900900901, + "grad_norm": 25.83568572998047, + "learning_rate": 4.238675887511862e-07, + "loss": 7.050273895263672, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.30934025533497334, + "loss_total_v6": 0.4406421259045601, + "step": 3790 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8743207207207208, + "grad_norm": 25.27288055419922, + "learning_rate": 4.223385986391276e-07, + "loss": 7.098462104797363, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.3153732605278492, + "loss_total_v6": 0.4436538890004158, + "step": 3791 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8745513513513513, + "grad_norm": 27.44247817993164, + "learning_rate": 4.208122496117745e-07, + "loss": 6.8998332023620605, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.3021349385380745, + "loss_total_v6": 0.4312395565211773, + "step": 3792 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.874781981981982, + "grad_norm": 22.653038024902344, + "learning_rate": 4.192885425497539e-07, + "loss": 7.034085273742676, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.30996115505695343, + "loss_total_v6": 0.43963034078478813, + "step": 3793 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8750126126126127, + "grad_norm": 24.45285987854004, + "learning_rate": 4.1776747833217056e-07, + "loss": 7.160961627960205, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.3172042928636074, + "loss_total_v6": 0.4475601091980934, + "step": 3794 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8752432432432432, + "grad_norm": 25.157920837402344, + "learning_rate": 4.1624905783660354e-07, + "loss": 7.217431545257568, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.3220001086592674, + "loss_total_v6": 0.4510894790291786, + "step": 3795 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8754738738738739, + "grad_norm": 27.600269317626953, + "learning_rate": 4.147332819391048e-07, + "loss": 7.109746932983398, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.3180621825158596, + "loss_total_v6": 0.4443591795861721, + "step": 3796 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8757045045045045, + "grad_norm": 25.95212745666504, + "learning_rate": 4.132201515142037e-07, + "loss": 6.963653564453125, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3024311065673828, + "loss_total_v6": 0.4352283589541912, + "step": 3797 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8759351351351351, + "grad_norm": 24.989381790161133, + "learning_rate": 4.1170966743490137e-07, + "loss": 6.664792060852051, + "loss_alignment": 0.2503662109375, + "loss_alignment_w": 0.12518310546875, + "loss_sft": 0.2916563078761101, + "loss_total_v6": 0.416549488902092, + "step": 3798 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8761657657657658, + "grad_norm": 29.654029846191406, + "learning_rate": 4.1020183057267403e-07, + "loss": 6.519004821777344, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.28046945482492447, + "loss_total_v6": 0.4074378199875355, + "step": 3799 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8763963963963964, + "grad_norm": 23.820825576782227, + "learning_rate": 4.0869664179746695e-07, + "loss": 6.823360919952393, + "loss_alignment": 0.254150390625, + "loss_alignment_w": 0.1270751953125, + "loss_sft": 0.2994001414626837, + "loss_total_v6": 0.42646007612347603, + "step": 3800 + }, + { + "epoch": 0.8763963963963964, + "eval_loss": 0.5626396536827087, + "eval_loss_alignment": 0.23047097959474885, + "eval_loss_alignment_w": 0.11523548979737443, + "eval_loss_sft": 0.4474041412359095, + "eval_loss_total_v6": 0.5626396308206532, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 91.8551, + "eval_samples_per_second": 18.997, + "eval_steps_per_second": 2.384, + "step": 3800 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.876627027027027, + "grad_norm": 24.04084587097168, + "learning_rate": 4.0719410197770036e-07, + "loss": 7.061979293823242, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.31255899742245674, + "loss_total_v6": 0.44137371331453323, + "step": 3801 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8768576576576577, + "grad_norm": 23.169687271118164, + "learning_rate": 4.0569421198026647e-07, + "loss": 6.942881107330322, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.30503904446959496, + "loss_total_v6": 0.43393005803227425, + "step": 3802 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8770882882882883, + "grad_norm": 24.363887786865234, + "learning_rate": 4.0419697267052696e-07, + "loss": 6.697976112365723, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2904496919363737, + "loss_total_v6": 0.41862352564930916, + "step": 3803 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8773189189189189, + "grad_norm": 27.55424690246582, + "learning_rate": 4.0270238491231573e-07, + "loss": 7.187554359436035, + "loss_alignment": 0.253662109375, + "loss_alignment_w": 0.1268310546875, + "loss_sft": 0.3220401108264923, + "loss_total_v6": 0.4492221213877201, + "step": 3804 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8775495495495496, + "grad_norm": 19.9143123626709, + "learning_rate": 4.0121044956793476e-07, + "loss": 7.097606658935547, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3112762067466974, + "loss_total_v6": 0.443600419908762, + "step": 3805 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8777801801801802, + "grad_norm": 24.936281204223633, + "learning_rate": 3.9972116749815827e-07, + "loss": 6.575281143188477, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.28516160510480404, + "loss_total_v6": 0.4109550639986992, + "step": 3806 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8780108108108108, + "grad_norm": 27.238502502441406, + "learning_rate": 3.982345395622284e-07, + "loss": 7.308200359344482, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.32799360156059265, + "loss_total_v6": 0.45676253363490105, + "step": 3807 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8782414414414415, + "grad_norm": 26.278409957885742, + "learning_rate": 3.9675056661785563e-07, + "loss": 6.543217658996582, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.2823794186115265, + "loss_total_v6": 0.4089510850608349, + "step": 3808 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8784720720720721, + "grad_norm": 25.978363037109375, + "learning_rate": 3.952692495212207e-07, + "loss": 7.063138961791992, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.31225001625716686, + "loss_total_v6": 0.4414462000131607, + "step": 3809 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8787027027027027, + "grad_norm": 23.53573226928711, + "learning_rate": 3.9379058912696833e-07, + "loss": 7.334744453430176, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3281114734709263, + "loss_total_v6": 0.4584215171635151, + "step": 3810 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8789333333333333, + "grad_norm": 24.763105392456055, + "learning_rate": 3.923145862882138e-07, + "loss": 6.379570960998535, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.2723651509732008, + "loss_total_v6": 0.39872318878769875, + "step": 3811 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.879163963963964, + "grad_norm": 23.313125610351562, + "learning_rate": 3.9084124185653716e-07, + "loss": 6.573878765106201, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.2846925016492605, + "loss_total_v6": 0.41086742654442787, + "step": 3812 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8793945945945946, + "grad_norm": 23.543476104736328, + "learning_rate": 3.893705566819861e-07, + "loss": 7.199141502380371, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3213452659547329, + "loss_total_v6": 0.4499463364481926, + "step": 3813 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8796252252252252, + "grad_norm": 25.70889663696289, + "learning_rate": 3.8790253161307355e-07, + "loss": 6.909544467926025, + "loss_alignment": 0.2535400390625, + "loss_alignment_w": 0.12677001953125, + "loss_sft": 0.30484760366380215, + "loss_total_v6": 0.4318465255200863, + "step": 3814 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8798558558558559, + "grad_norm": 23.461658477783203, + "learning_rate": 3.864371674967765e-07, + "loss": 7.063327789306641, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.311178470030427, + "loss_total_v6": 0.44145799800753593, + "step": 3815 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8800864864864865, + "grad_norm": 22.553133010864258, + "learning_rate": 3.849744651785381e-07, + "loss": 6.8257927894592285, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.2972175292670727, + "loss_total_v6": 0.4266120530664921, + "step": 3816 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8803171171171171, + "grad_norm": 22.477130889892578, + "learning_rate": 3.83514425502266e-07, + "loss": 6.998620510101318, + "loss_alignment": 0.2523193359375, + "loss_alignment_w": 0.12615966796875, + "loss_sft": 0.3110405020415783, + "loss_total_v6": 0.4374137856066227, + "step": 3817 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8805477477477478, + "grad_norm": 24.739316940307617, + "learning_rate": 3.8205704931033097e-07, + "loss": 6.665313243865967, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.28605837374925613, + "loss_total_v6": 0.41658205911517143, + "step": 3818 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8807783783783784, + "grad_norm": 23.4628849029541, + "learning_rate": 3.8060233744356634e-07, + "loss": 7.236181735992432, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.324621569365263, + "loss_total_v6": 0.4522613435983658, + "step": 3819 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.881009009009009, + "grad_norm": 24.289703369140625, + "learning_rate": 3.7915029074126976e-07, + "loss": 6.8765363693237305, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.30142661929130554, + "loss_total_v6": 0.42978353798389435, + "step": 3820 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8812396396396397, + "grad_norm": 25.187442779541016, + "learning_rate": 3.77700910041201e-07, + "loss": 7.356449127197266, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.3299715742468834, + "loss_total_v6": 0.4597780928015709, + "step": 3821 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8814702702702703, + "grad_norm": 23.767070770263672, + "learning_rate": 3.762541961795796e-07, + "loss": 7.3151140213012695, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.32833415642380714, + "loss_total_v6": 0.45719461888074875, + "step": 3822 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8817009009009009, + "grad_norm": 24.97907829284668, + "learning_rate": 3.7481014999108886e-07, + "loss": 6.7631516456604, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2937296871095896, + "loss_total_v6": 0.422696977853775, + "step": 3823 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8819315315315315, + "grad_norm": 24.447874069213867, + "learning_rate": 3.733687723088725e-07, + "loss": 6.727500915527344, + "loss_alignment": 0.2509765625, + "loss_alignment_w": 0.12548828125, + "loss_sft": 0.2949957922101021, + "loss_total_v6": 0.420468807220459, + "step": 3824 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8821621621621621, + "grad_norm": 23.553129196166992, + "learning_rate": 3.7193006396453456e-07, + "loss": 6.934333324432373, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.30505415238440037, + "loss_total_v6": 0.433395829051733, + "step": 3825 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8823927927927928, + "grad_norm": 25.646427154541016, + "learning_rate": 3.7049402578813787e-07, + "loss": 6.783952713012695, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.29278672486543655, + "loss_total_v6": 0.42399704828858376, + "step": 3826 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8826234234234234, + "grad_norm": 24.849956512451172, + "learning_rate": 3.690606586082057e-07, + "loss": 7.435132026672363, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3348892331123352, + "loss_total_v6": 0.464695755392313, + "step": 3827 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.882854054054054, + "grad_norm": 25.434783935546875, + "learning_rate": 3.676299632517216e-07, + "loss": 6.689944267272949, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2880860846489668, + "loss_total_v6": 0.41812148690223694, + "step": 3828 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8830846846846847, + "grad_norm": 24.19113540649414, + "learning_rate": 3.662019405441236e-07, + "loss": 7.062163829803467, + "loss_alignment": 0.2506103515625, + "loss_alignment_w": 0.12530517578125, + "loss_sft": 0.3161563612520695, + "loss_total_v6": 0.4413852393627167, + "step": 3829 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8833153153153153, + "grad_norm": 24.929489135742188, + "learning_rate": 3.647765913093132e-07, + "loss": 7.485444068908691, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.3359432891011238, + "loss_total_v6": 0.4678402543067932, + "step": 3830 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8835459459459459, + "grad_norm": 28.311864852905273, + "learning_rate": 3.633539163696442e-07, + "loss": 6.687777042388916, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.2876760121434927, + "loss_total_v6": 0.41798607632517815, + "step": 3831 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8837765765765766, + "grad_norm": 23.840669631958008, + "learning_rate": 3.619339165459307e-07, + "loss": 7.436881065368652, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.33519691601395607, + "loss_total_v6": 0.4648050628602505, + "step": 3832 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8840072072072072, + "grad_norm": 23.52416229248047, + "learning_rate": 3.605165926574422e-07, + "loss": 6.884485244750977, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.30286945030093193, + "loss_total_v6": 0.4302803464233875, + "step": 3833 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8842378378378378, + "grad_norm": 25.943607330322266, + "learning_rate": 3.591019455219053e-07, + "loss": 7.13709831237793, + "loss_alignment": 0.265380859375, + "loss_alignment_w": 0.1326904296875, + "loss_sft": 0.3133782111108303, + "loss_total_v6": 0.4460686258971691, + "step": 3834 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8844684684684685, + "grad_norm": 25.20363426208496, + "learning_rate": 3.576899759555019e-07, + "loss": 6.516183376312256, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.28102548979222775, + "loss_total_v6": 0.4072614647448063, + "step": 3835 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.884699099099099, + "grad_norm": 22.695966720581055, + "learning_rate": 3.562806847728678e-07, + "loss": 7.405315399169922, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.3327815420925617, + "loss_total_v6": 0.4628321938216686, + "step": 3836 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8849297297297297, + "grad_norm": 24.453243255615234, + "learning_rate": 3.548740727870953e-07, + "loss": 6.59619140625, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.2855682373046875, + "loss_total_v6": 0.4122619517147541, + "step": 3837 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8851603603603604, + "grad_norm": 23.07328987121582, + "learning_rate": 3.5347014080972854e-07, + "loss": 6.721035003662109, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.2923485804349184, + "loss_total_v6": 0.42006466165184975, + "step": 3838 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.885390990990991, + "grad_norm": 25.09476661682129, + "learning_rate": 3.520688896507685e-07, + "loss": 6.77762508392334, + "loss_alignment": 0.2523193359375, + "loss_alignment_w": 0.12615966796875, + "loss_sft": 0.2975029591470957, + "loss_total_v6": 0.42360157892107964, + "step": 3839 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8856216216216216, + "grad_norm": 23.93675994873047, + "learning_rate": 3.506703201186679e-07, + "loss": 7.12809944152832, + "loss_alignment": 0.248046875, + "loss_alignment_w": 0.1240234375, + "loss_sft": 0.3215285651385784, + "loss_total_v6": 0.4455062337219715, + "step": 3840 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8858522522522523, + "grad_norm": 25.525283813476562, + "learning_rate": 3.4927443302033127e-07, + "loss": 6.487393379211426, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.2764948159456253, + "loss_total_v6": 0.4054621122777462, + "step": 3841 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8860828828828828, + "grad_norm": 23.930139541625977, + "learning_rate": 3.4788122916111723e-07, + "loss": 7.144420146942139, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.3180777579545975, + "loss_total_v6": 0.44652625545859337, + "step": 3842 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8863135135135135, + "grad_norm": 24.36656379699707, + "learning_rate": 3.4649070934483554e-07, + "loss": 6.604159355163574, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.28255671821534634, + "loss_total_v6": 0.4127599410712719, + "step": 3843 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8865441441441442, + "grad_norm": 27.928218841552734, + "learning_rate": 3.4510287437374835e-07, + "loss": 7.39473819732666, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.33341750502586365, + "loss_total_v6": 0.46217115595936775, + "step": 3844 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8867747747747747, + "grad_norm": 24.158794403076172, + "learning_rate": 3.4371772504856624e-07, + "loss": 6.650064468383789, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2858530320227146, + "loss_total_v6": 0.4156290180981159, + "step": 3845 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8870054054054054, + "grad_norm": 26.69586944580078, + "learning_rate": 3.4233526216845336e-07, + "loss": 6.411923408508301, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.27208310551941395, + "loss_total_v6": 0.4007452204823494, + "step": 3846 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8872360360360361, + "grad_norm": 24.11471176147461, + "learning_rate": 3.4095548653102273e-07, + "loss": 7.086827278137207, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.31493596732616425, + "loss_total_v6": 0.44292669370770454, + "step": 3847 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8874666666666666, + "grad_norm": 21.66419792175293, + "learning_rate": 3.3957839893233535e-07, + "loss": 6.554874897003174, + "loss_alignment": 0.247802734375, + "loss_alignment_w": 0.1239013671875, + "loss_sft": 0.28609875589609146, + "loss_total_v6": 0.40967968106269836, + "step": 3848 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8876972972972973, + "grad_norm": 25.88050079345703, + "learning_rate": 3.3820400016690345e-07, + "loss": 7.237098217010498, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.321093063801527, + "loss_total_v6": 0.4523186460137367, + "step": 3849 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.887927927927928, + "grad_norm": 25.083742141723633, + "learning_rate": 3.368322910276883e-07, + "loss": 6.6390790939331055, + "loss_alignment": 0.2518310546875, + "loss_alignment_w": 0.12591552734375, + "loss_sft": 0.28949992544949055, + "loss_total_v6": 0.4149424433708191, + "step": 3850 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8881585585585585, + "grad_norm": 26.738616943359375, + "learning_rate": 3.3546327230609746e-07, + "loss": 6.732181549072266, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.2909243106842041, + "loss_total_v6": 0.4207613468170166, + "step": 3851 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8883891891891892, + "grad_norm": 22.8088321685791, + "learning_rate": 3.340969447919873e-07, + "loss": 6.767388343811035, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.2941470704972744, + "loss_total_v6": 0.4229617528617382, + "step": 3852 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8886198198198199, + "grad_norm": 24.628746032714844, + "learning_rate": 3.3273330927366065e-07, + "loss": 7.209447860717773, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3228591475635767, + "loss_total_v6": 0.45059048384428024, + "step": 3853 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8888504504504504, + "grad_norm": 24.721908569335938, + "learning_rate": 3.313723665378699e-07, + "loss": 6.862705230712891, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.3016760144382715, + "loss_total_v6": 0.42891906946897507, + "step": 3854 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8890810810810811, + "grad_norm": 23.15623664855957, + "learning_rate": 3.300141173698096e-07, + "loss": 7.149912357330322, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.31625428795814514, + "loss_total_v6": 0.44686952233314514, + "step": 3855 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8893117117117117, + "grad_norm": 21.872705459594727, + "learning_rate": 3.286585625531241e-07, + "loss": 7.098265647888184, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3120650537312031, + "loss_total_v6": 0.44364158436656, + "step": 3856 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8895423423423423, + "grad_norm": 25.359989166259766, + "learning_rate": 3.273057028699006e-07, + "loss": 7.196817398071289, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.320788050070405, + "loss_total_v6": 0.44980111345648766, + "step": 3857 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.889772972972973, + "grad_norm": 28.741920471191406, + "learning_rate": 3.2595553910067247e-07, + "loss": 6.83168888092041, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.29802851378917694, + "loss_total_v6": 0.42698055505752563, + "step": 3858 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8900036036036036, + "grad_norm": 26.493003845214844, + "learning_rate": 3.2460807202441767e-07, + "loss": 6.9542083740234375, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.30504509434103966, + "loss_total_v6": 0.43463800475001335, + "step": 3859 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8902342342342342, + "grad_norm": 25.096752166748047, + "learning_rate": 3.232633024185583e-07, + "loss": 7.1571502685546875, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.3156385477632284, + "loss_total_v6": 0.44732189178466797, + "step": 3860 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8904648648648649, + "grad_norm": 23.090839385986328, + "learning_rate": 3.2192123105896147e-07, + "loss": 6.693784713745117, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.2916372921317816, + "loss_total_v6": 0.4183615408837795, + "step": 3861 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8906954954954955, + "grad_norm": 25.331165313720703, + "learning_rate": 3.205818587199333e-07, + "loss": 6.7843852043151855, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.2943091243505478, + "loss_total_v6": 0.4240240789949894, + "step": 3862 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8909261261261261, + "grad_norm": 25.377931594848633, + "learning_rate": 3.1924518617422796e-07, + "loss": 6.261448860168457, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.26254112645983696, + "loss_total_v6": 0.39134055748581886, + "step": 3863 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8911567567567568, + "grad_norm": 25.678852081298828, + "learning_rate": 3.1791121419303794e-07, + "loss": 7.35869836807251, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.3293949291110039, + "loss_total_v6": 0.45991864055395126, + "step": 3864 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8913873873873874, + "grad_norm": 22.90373992919922, + "learning_rate": 3.1657994354599987e-07, + "loss": 6.6786932945251465, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.29084667190909386, + "loss_total_v6": 0.41741833090782166, + "step": 3865 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.891618018018018, + "grad_norm": 27.441513061523438, + "learning_rate": 3.1525137500119207e-07, + "loss": 7.081650733947754, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.31549743190407753, + "loss_total_v6": 0.4426031559705734, + "step": 3866 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8918486486486487, + "grad_norm": 26.063474655151367, + "learning_rate": 3.1392550932513086e-07, + "loss": 7.154536247253418, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3194730021059513, + "loss_total_v6": 0.4471585266292095, + "step": 3867 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8920792792792793, + "grad_norm": 25.954557418823242, + "learning_rate": 3.1260234728277716e-07, + "loss": 6.826472282409668, + "loss_alignment": 0.26416015625, + "loss_alignment_w": 0.132080078125, + "loss_sft": 0.2946659903973341, + "loss_total_v6": 0.42665452137589455, + "step": 3868 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8923099099099099, + "grad_norm": 28.569528579711914, + "learning_rate": 3.112818896375297e-07, + "loss": 6.806670188903809, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.29916566237807274, + "loss_total_v6": 0.42541688308119774, + "step": 3869 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8925405405405406, + "grad_norm": 23.814165115356445, + "learning_rate": 3.0996413715122744e-07, + "loss": 6.459655284881592, + "loss_alignment": 0.247314453125, + "loss_alignment_w": 0.1236572265625, + "loss_sft": 0.28010172955691814, + "loss_total_v6": 0.4037284329533577, + "step": 3870 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8927711711711712, + "grad_norm": 25.775480270385742, + "learning_rate": 3.0864909058414826e-07, + "loss": 6.836159706115723, + "loss_alignment": 0.2532958984375, + "loss_alignment_w": 0.12664794921875, + "loss_sft": 0.3006425593048334, + "loss_total_v6": 0.42725999653339386, + "step": 3871 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8930018018018018, + "grad_norm": 24.86655616760254, + "learning_rate": 3.0733675069500865e-07, + "loss": 7.175161838531494, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.3183816894888878, + "loss_total_v6": 0.4484476111829281, + "step": 3872 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8932324324324324, + "grad_norm": 22.958438873291016, + "learning_rate": 3.0602711824096576e-07, + "loss": 6.996350288391113, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.30796893686056137, + "loss_total_v6": 0.4372719116508961, + "step": 3873 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8934630630630631, + "grad_norm": 25.51674461364746, + "learning_rate": 3.0472019397761065e-07, + "loss": 6.858759880065918, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.29827089980244637, + "loss_total_v6": 0.42867251485586166, + "step": 3874 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8936936936936937, + "grad_norm": 21.74680519104004, + "learning_rate": 3.034159786589752e-07, + "loss": 6.729142189025879, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.29261115193367004, + "loss_total_v6": 0.42057136818766594, + "step": 3875 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8939243243243243, + "grad_norm": 26.051471710205078, + "learning_rate": 3.0211447303752697e-07, + "loss": 6.640337944030762, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.28571816347539425, + "loss_total_v6": 0.4150211401283741, + "step": 3876 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.894154954954955, + "grad_norm": 24.886953353881836, + "learning_rate": 3.0081567786417187e-07, + "loss": 6.933570384979248, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3056931160390377, + "loss_total_v6": 0.4333481416106224, + "step": 3877 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8943855855855856, + "grad_norm": 27.253849029541016, + "learning_rate": 2.995195938882484e-07, + "loss": 6.795540809631348, + "loss_alignment": 0.2523193359375, + "loss_alignment_w": 0.12615966796875, + "loss_sft": 0.29845481365919113, + "loss_total_v6": 0.42472128942608833, + "step": 3878 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8946162162162162, + "grad_norm": 28.18679428100586, + "learning_rate": 2.982262218575338e-07, + "loss": 6.457900524139404, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.27549075335264206, + "loss_total_v6": 0.40361879765987396, + "step": 3879 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8948468468468469, + "grad_norm": 25.40758514404297, + "learning_rate": 2.9693556251824187e-07, + "loss": 7.245041847229004, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.32299333438277245, + "loss_total_v6": 0.45281510427594185, + "step": 3880 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8950774774774775, + "grad_norm": 26.861051559448242, + "learning_rate": 2.956476166150163e-07, + "loss": 6.651227951049805, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.28830611333251, + "loss_total_v6": 0.4157017320394516, + "step": 3881 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8953081081081081, + "grad_norm": 26.10268783569336, + "learning_rate": 2.943623848909405e-07, + "loss": 7.225095748901367, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.32313528284430504, + "loss_total_v6": 0.45156849920749664, + "step": 3882 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8955387387387388, + "grad_norm": 25.823514938354492, + "learning_rate": 2.9307986808752796e-07, + "loss": 6.932271957397461, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.3067563660442829, + "loss_total_v6": 0.4332669898867607, + "step": 3883 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8957693693693694, + "grad_norm": 23.136903762817383, + "learning_rate": 2.9180006694472907e-07, + "loss": 6.959779739379883, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3065835162997246, + "loss_total_v6": 0.4349862299859524, + "step": 3884 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.896, + "grad_norm": 24.298080444335938, + "learning_rate": 2.905229822009253e-07, + "loss": 6.802700042724609, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.2958962842822075, + "loss_total_v6": 0.4251687526702881, + "step": 3885 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8962306306306306, + "grad_norm": 23.631324768066406, + "learning_rate": 2.892486145929313e-07, + "loss": 7.089542388916016, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.31213023886084557, + "loss_total_v6": 0.44309642910957336, + "step": 3886 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8964612612612612, + "grad_norm": 25.03489112854004, + "learning_rate": 2.8797696485599547e-07, + "loss": 7.476803779602051, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.33807355910539627, + "loss_total_v6": 0.46730025485157967, + "step": 3887 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8966918918918919, + "grad_norm": 25.926488876342773, + "learning_rate": 2.867080337237954e-07, + "loss": 7.2038397789001465, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.32318005710840225, + "loss_total_v6": 0.45023998990654945, + "step": 3888 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8969225225225225, + "grad_norm": 27.459674835205078, + "learning_rate": 2.854418219284433e-07, + "loss": 6.741927146911621, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.2938069812953472, + "loss_total_v6": 0.4213704653084278, + "step": 3889 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8971531531531531, + "grad_norm": 23.70281219482422, + "learning_rate": 2.841783302004791e-07, + "loss": 6.976343631744385, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.30502476915717125, + "loss_total_v6": 0.43602147698402405, + "step": 3890 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8973837837837838, + "grad_norm": 29.329137802124023, + "learning_rate": 2.82917559268876e-07, + "loss": 6.681015968322754, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.29151064716279507, + "loss_total_v6": 0.4175635054707527, + "step": 3891 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8976144144144144, + "grad_norm": 22.790618896484375, + "learning_rate": 2.8165950986103807e-07, + "loss": 7.064831256866455, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.31165387108922005, + "loss_total_v6": 0.44155194237828255, + "step": 3892 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.897845045045045, + "grad_norm": 22.97252082824707, + "learning_rate": 2.804041827027942e-07, + "loss": 6.707708358764648, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.2902492228895426, + "loss_total_v6": 0.4192317798733711, + "step": 3893 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8980756756756757, + "grad_norm": 25.699825286865234, + "learning_rate": 2.791515785184101e-07, + "loss": 7.2442474365234375, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.32186031341552734, + "loss_total_v6": 0.45276545360684395, + "step": 3894 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8983063063063063, + "grad_norm": 25.62598991394043, + "learning_rate": 2.7790169803057375e-07, + "loss": 7.136237144470215, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31518595665693283, + "loss_total_v6": 0.44601481780409813, + "step": 3895 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8985369369369369, + "grad_norm": 21.32936668395996, + "learning_rate": 2.7665454196040665e-07, + "loss": 6.987024307250977, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.30441056564450264, + "loss_total_v6": 0.43668902665376663, + "step": 3896 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8987675675675676, + "grad_norm": 25.016128540039062, + "learning_rate": 2.754101110274543e-07, + "loss": 6.732314586639404, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.29338930547237396, + "loss_total_v6": 0.42076966911554337, + "step": 3897 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8989981981981982, + "grad_norm": 24.364471435546875, + "learning_rate": 2.741684059496924e-07, + "loss": 7.239570617675781, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3230938632041216, + "loss_total_v6": 0.45247314125299454, + "step": 3898 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8992288288288288, + "grad_norm": 25.01824378967285, + "learning_rate": 2.729294274435251e-07, + "loss": 7.083296775817871, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.31123635172843933, + "loss_total_v6": 0.44270606338977814, + "step": 3899 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8994594594594595, + "grad_norm": 25.914796829223633, + "learning_rate": 2.716931762237801e-07, + "loss": 6.676930904388428, + "loss_alignment": 0.2626953125, + "loss_alignment_w": 0.13134765625, + "loss_sft": 0.28593001887202263, + "loss_total_v6": 0.41730818524956703, + "step": 3900 + }, + { + "epoch": 0.8994594594594595, + "eval_loss": 0.5632458925247192, + "eval_loss_alignment": 0.23136727668378995, + "eval_loss_alignment_w": 0.11568363834189498, + "eval_loss_sft": 0.44756226812289457, + "eval_loss_total_v6": 0.5632459073578386, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 93.7794, + "eval_samples_per_second": 18.607, + "eval_steps_per_second": 2.335, + "step": 3900 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.89969009009009, + "grad_norm": 24.908903121948242, + "learning_rate": 2.7045965300371404e-07, + "loss": 7.35515022277832, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.3310957998037338, + "loss_total_v6": 0.4596968777477741, + "step": 3901 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.8999207207207207, + "grad_norm": 25.566932678222656, + "learning_rate": 2.6922885849500867e-07, + "loss": 6.792300224304199, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.2952310014516115, + "loss_total_v6": 0.42451874911785126, + "step": 3902 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9001513513513514, + "grad_norm": 24.381092071533203, + "learning_rate": 2.680007934077722e-07, + "loss": 6.95654821395874, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.30455049499869347, + "loss_total_v6": 0.43478426337242126, + "step": 3903 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9003819819819819, + "grad_norm": 26.921031951904297, + "learning_rate": 2.667754584505372e-07, + "loss": 6.862865447998047, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.30055689066648483, + "loss_total_v6": 0.42892907187342644, + "step": 3904 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9006126126126126, + "grad_norm": 24.549222946166992, + "learning_rate": 2.6555285433026103e-07, + "loss": 7.295281410217285, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3278117775917053, + "loss_total_v6": 0.45595508441329, + "step": 3905 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9008432432432433, + "grad_norm": 28.523624420166016, + "learning_rate": 2.6433298175232715e-07, + "loss": 7.222455024719238, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.323061753064394, + "loss_total_v6": 0.4514034278690815, + "step": 3906 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9010738738738738, + "grad_norm": 27.834226608276367, + "learning_rate": 2.6311584142054036e-07, + "loss": 7.410823822021484, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.33286643773317337, + "loss_total_v6": 0.46317649260163307, + "step": 3907 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9013045045045045, + "grad_norm": 24.289575576782227, + "learning_rate": 2.6190143403713175e-07, + "loss": 6.569639205932617, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.28250489570200443, + "loss_total_v6": 0.4106024466454983, + "step": 3908 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9015351351351352, + "grad_norm": 25.1343994140625, + "learning_rate": 2.606897603027536e-07, + "loss": 6.8162078857421875, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.2969999313354492, + "loss_total_v6": 0.426012996584177, + "step": 3909 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9017657657657657, + "grad_norm": 26.089679718017578, + "learning_rate": 2.594808209164823e-07, + "loss": 7.184527397155762, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.3215762749314308, + "loss_total_v6": 0.4490329436957836, + "step": 3910 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9019963963963964, + "grad_norm": 25.686471939086914, + "learning_rate": 2.5827461657581544e-07, + "loss": 7.182915687561035, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.3203158788383007, + "loss_total_v6": 0.4489322155714035, + "step": 3911 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9022270270270271, + "grad_norm": 25.208818435668945, + "learning_rate": 2.5707114797667467e-07, + "loss": 6.908002853393555, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.30082976073026657, + "loss_total_v6": 0.43175020441412926, + "step": 3912 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9024576576576576, + "grad_norm": 28.839330673217773, + "learning_rate": 2.5587041581340235e-07, + "loss": 7.325354099273682, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.3274177871644497, + "loss_total_v6": 0.4578346461057663, + "step": 3913 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9026882882882883, + "grad_norm": 21.37364959716797, + "learning_rate": 2.546724207787593e-07, + "loss": 7.331887245178223, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.32970290444791317, + "loss_total_v6": 0.4582429379224777, + "step": 3914 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.902918918918919, + "grad_norm": 23.761625289916992, + "learning_rate": 2.5347716356393214e-07, + "loss": 7.1664605140686035, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.3189517520368099, + "loss_total_v6": 0.4479037784039974, + "step": 3915 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9031495495495495, + "grad_norm": 24.317373275756836, + "learning_rate": 2.522846448585231e-07, + "loss": 7.000979423522949, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.30978409200906754, + "loss_total_v6": 0.4375612139701843, + "step": 3916 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9033801801801802, + "grad_norm": 22.92003059387207, + "learning_rate": 2.510948653505585e-07, + "loss": 7.024173259735107, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3081056773662567, + "loss_total_v6": 0.4390108399093151, + "step": 3917 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9036108108108108, + "grad_norm": 23.588600158691406, + "learning_rate": 2.4990782572647977e-07, + "loss": 7.034409523010254, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3079367093741894, + "loss_total_v6": 0.4396505728363991, + "step": 3918 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9038414414414414, + "grad_norm": 25.55095863342285, + "learning_rate": 2.4872352667115296e-07, + "loss": 7.150992393493652, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.31859536841511726, + "loss_total_v6": 0.44693703949451447, + "step": 3919 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9040720720720721, + "grad_norm": 25.140878677368164, + "learning_rate": 2.4754196886785984e-07, + "loss": 6.9043378829956055, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.30339307337999344, + "loss_total_v6": 0.43152112886309624, + "step": 3920 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9043027027027027, + "grad_norm": 25.233205795288086, + "learning_rate": 2.4636315299829963e-07, + "loss": 6.770040512084961, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.29443488642573357, + "loss_total_v6": 0.42312750592827797, + "step": 3921 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9045333333333333, + "grad_norm": 24.63266944885254, + "learning_rate": 2.45187079742592e-07, + "loss": 7.38090181350708, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.3319728709757328, + "loss_total_v6": 0.4613063670694828, + "step": 3922 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.904763963963964, + "grad_norm": 24.569419860839844, + "learning_rate": 2.440137497792722e-07, + "loss": 7.4856743812561035, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.33743778243660927, + "loss_total_v6": 0.46785465255379677, + "step": 3923 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9049945945945946, + "grad_norm": 23.42772674560547, + "learning_rate": 2.4284316378529405e-07, + "loss": 6.832369804382324, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.2971707843244076, + "loss_total_v6": 0.42702309787273407, + "step": 3924 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9052252252252252, + "grad_norm": 25.214468002319336, + "learning_rate": 2.4167532243602853e-07, + "loss": 6.361456871032715, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.2684711702167988, + "loss_total_v6": 0.3975910469889641, + "step": 3925 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9054558558558559, + "grad_norm": 24.064592361450195, + "learning_rate": 2.405102264052617e-07, + "loss": 6.537505149841309, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.27923006005585194, + "loss_total_v6": 0.4085940644145012, + "step": 3926 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9056864864864865, + "grad_norm": 25.013456344604492, + "learning_rate": 2.3934787636519683e-07, + "loss": 7.145473003387451, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.3145577684044838, + "loss_total_v6": 0.4465920701622963, + "step": 3927 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9059171171171171, + "grad_norm": 25.818010330200195, + "learning_rate": 2.3818827298645208e-07, + "loss": 7.391632080078125, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.3317432291805744, + "loss_total_v6": 0.4619769938290119, + "step": 3928 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9061477477477478, + "grad_norm": 26.615644454956055, + "learning_rate": 2.3703141693806276e-07, + "loss": 6.909153938293457, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.30364830419421196, + "loss_total_v6": 0.43182212486863136, + "step": 3929 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9063783783783784, + "grad_norm": 23.59856414794922, + "learning_rate": 2.358773088874755e-07, + "loss": 7.072266578674316, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.3133087698370218, + "loss_total_v6": 0.44201667606830597, + "step": 3930 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.906609009009009, + "grad_norm": 22.808231353759766, + "learning_rate": 2.347259495005555e-07, + "loss": 7.104159355163574, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.3156224898993969, + "loss_total_v6": 0.4440099336206913, + "step": 3931 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9068396396396396, + "grad_norm": 25.902446746826172, + "learning_rate": 2.3357733944158023e-07, + "loss": 7.359986305236816, + "loss_alignment": 0.261474609375, + "loss_alignment_w": 0.1307373046875, + "loss_sft": 0.3291855398565531, + "loss_total_v6": 0.4599991664290428, + "step": 3932 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9070702702702703, + "grad_norm": 22.65494728088379, + "learning_rate": 2.3243147937323963e-07, + "loss": 6.9657392501831055, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.3063303641974926, + "loss_total_v6": 0.4353586919605732, + "step": 3933 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9073009009009009, + "grad_norm": 26.05914306640625, + "learning_rate": 2.3128836995663985e-07, + "loss": 7.496921539306641, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.33980395272374153, + "loss_total_v6": 0.46855761110782623, + "step": 3934 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9075315315315315, + "grad_norm": 24.07537269592285, + "learning_rate": 2.301480118512972e-07, + "loss": 7.115033149719238, + "loss_alignment": 0.249755859375, + "loss_alignment_w": 0.1248779296875, + "loss_sft": 0.3197811357676983, + "loss_total_v6": 0.4446895718574524, + "step": 3935 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9077621621621622, + "grad_norm": 29.241289138793945, + "learning_rate": 2.2901040571514322e-07, + "loss": 6.785675048828125, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.2962970696389675, + "loss_total_v6": 0.4241046980023384, + "step": 3936 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9079927927927928, + "grad_norm": 27.13078498840332, + "learning_rate": 2.2787555220452006e-07, + "loss": 6.772331237792969, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.29650069028139114, + "loss_total_v6": 0.42327070608735085, + "step": 3937 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9082234234234234, + "grad_norm": 26.099519729614258, + "learning_rate": 2.2674345197418236e-07, + "loss": 6.7482733726501465, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.2932575698941946, + "loss_total_v6": 0.42176708579063416, + "step": 3938 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9084540540540541, + "grad_norm": 23.188636779785156, + "learning_rate": 2.2561410567729592e-07, + "loss": 6.738569736480713, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.2917813491076231, + "loss_total_v6": 0.42116060853004456, + "step": 3939 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9086846846846847, + "grad_norm": 26.94087791442871, + "learning_rate": 2.2448751396543788e-07, + "loss": 6.9329400062561035, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.30592838674783707, + "loss_total_v6": 0.43330876529216766, + "step": 3940 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9089153153153153, + "grad_norm": 24.998947143554688, + "learning_rate": 2.2336367748859667e-07, + "loss": 7.114413261413574, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3152257800102234, + "loss_total_v6": 0.4446508288383484, + "step": 3941 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.909145945945946, + "grad_norm": 24.155630111694336, + "learning_rate": 2.222425968951686e-07, + "loss": 6.966707229614258, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.3074590228497982, + "loss_total_v6": 0.4354192316532135, + "step": 3942 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9093765765765766, + "grad_norm": 24.212020874023438, + "learning_rate": 2.2112427283196457e-07, + "loss": 6.912324905395508, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.3026410285383463, + "loss_total_v6": 0.43202030286192894, + "step": 3943 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9096072072072072, + "grad_norm": 22.482542037963867, + "learning_rate": 2.200087059441991e-07, + "loss": 6.704623222351074, + "loss_alignment": 0.2509765625, + "loss_alignment_w": 0.12548828125, + "loss_sft": 0.2938711252063513, + "loss_total_v6": 0.41903894767165184, + "step": 3944 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9098378378378379, + "grad_norm": 26.273433685302734, + "learning_rate": 2.1889589687550172e-07, + "loss": 7.337517738342285, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.32999379001557827, + "loss_total_v6": 0.4585948660969734, + "step": 3945 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9100684684684685, + "grad_norm": 24.488750457763672, + "learning_rate": 2.1778584626790777e-07, + "loss": 7.17335319519043, + "loss_alignment": 0.260498046875, + "loss_alignment_w": 0.1302490234375, + "loss_sft": 0.31816181540489197, + "loss_total_v6": 0.44833454862236977, + "step": 3946 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9102990990990991, + "grad_norm": 23.19116973876953, + "learning_rate": 2.1667855476186161e-07, + "loss": 6.998799800872803, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.30745065584778786, + "loss_total_v6": 0.43742499500513077, + "step": 3947 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9105297297297297, + "grad_norm": 23.04344940185547, + "learning_rate": 2.155740229962161e-07, + "loss": 7.096689224243164, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.31655943021178246, + "loss_total_v6": 0.44354307278990746, + "step": 3948 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9107603603603603, + "grad_norm": 26.134275436401367, + "learning_rate": 2.1447225160823038e-07, + "loss": 7.332536697387695, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.33055221661925316, + "loss_total_v6": 0.45828352496027946, + "step": 3949 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.910990990990991, + "grad_norm": 24.40094757080078, + "learning_rate": 2.1337324123357316e-07, + "loss": 7.20259952545166, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.32151563093066216, + "loss_total_v6": 0.45016247779130936, + "step": 3950 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9112216216216216, + "grad_norm": 25.56965446472168, + "learning_rate": 2.122769925063195e-07, + "loss": 7.009466171264648, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.3083156440407038, + "loss_total_v6": 0.4380916357040405, + "step": 3951 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9114522522522522, + "grad_norm": 27.17896270751953, + "learning_rate": 2.1118350605894955e-07, + "loss": 7.177933692932129, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.31777673959732056, + "loss_total_v6": 0.44862085580825806, + "step": 3952 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9116828828828829, + "grad_norm": 23.674720764160156, + "learning_rate": 2.10092782522352e-07, + "loss": 6.968152046203613, + "loss_alignment": 0.2508544921875, + "loss_alignment_w": 0.12542724609375, + "loss_sft": 0.3100670091807842, + "loss_total_v6": 0.43550949543714523, + "step": 3953 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9119135135135135, + "grad_norm": 24.592992782592773, + "learning_rate": 2.090048225258201e-07, + "loss": 6.736294746398926, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.2906473409384489, + "loss_total_v6": 0.42101842164993286, + "step": 3954 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9121441441441441, + "grad_norm": 29.151281356811523, + "learning_rate": 2.07919626697054e-07, + "loss": 6.983212471008301, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.30873468704521656, + "loss_total_v6": 0.4364507608115673, + "step": 3955 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9123747747747748, + "grad_norm": 21.4293155670166, + "learning_rate": 2.068371956621562e-07, + "loss": 7.281703948974609, + "loss_alignment": 0.2520751953125, + "loss_alignment_w": 0.12603759765625, + "loss_sft": 0.3293893076479435, + "loss_total_v6": 0.4551064781844616, + "step": 3956 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9126054054054054, + "grad_norm": 24.13740348815918, + "learning_rate": 2.057575300456377e-07, + "loss": 7.0457563400268555, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3100954666733742, + "loss_total_v6": 0.4403597488999367, + "step": 3957 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.912836036036036, + "grad_norm": 24.48520278930664, + "learning_rate": 2.046806304704124e-07, + "loss": 6.722775459289551, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.29271678626537323, + "loss_total_v6": 0.42017344757914543, + "step": 3958 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9130666666666667, + "grad_norm": 25.64198112487793, + "learning_rate": 2.0360649755779672e-07, + "loss": 6.913064956665039, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30050525069236755, + "loss_total_v6": 0.43206654489040375, + "step": 3959 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9132972972972973, + "grad_norm": 24.440853118896484, + "learning_rate": 2.0253513192751374e-07, + "loss": 6.355625152587891, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.26853392645716667, + "loss_total_v6": 0.39722657203674316, + "step": 3960 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9135279279279279, + "grad_norm": 21.461965560913086, + "learning_rate": 2.0146653419768737e-07, + "loss": 7.02132511138916, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.30878218822181225, + "loss_total_v6": 0.4388328418135643, + "step": 3961 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9137585585585586, + "grad_norm": 21.43674659729004, + "learning_rate": 2.004007049848461e-07, + "loss": 7.308572769165039, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.32731494307518005, + "loss_total_v6": 0.45678577572107315, + "step": 3962 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9139891891891891, + "grad_norm": 28.736013412475586, + "learning_rate": 1.993376449039214e-07, + "loss": 7.074644565582275, + "loss_alignment": 0.2547607421875, + "loss_alignment_w": 0.12738037109375, + "loss_sft": 0.3147849068045616, + "loss_total_v6": 0.442165270447731, + "step": 3963 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9142198198198198, + "grad_norm": 25.971111297607422, + "learning_rate": 1.9827735456824593e-07, + "loss": 7.108339309692383, + "loss_alignment": 0.25048828125, + "loss_alignment_w": 0.125244140625, + "loss_sft": 0.3187524266541004, + "loss_total_v6": 0.4442712217569351, + "step": 3964 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9144504504504505, + "grad_norm": 26.276016235351562, + "learning_rate": 1.9721983458955597e-07, + "loss": 6.175623893737793, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.25774164870381355, + "loss_total_v6": 0.38597650453448296, + "step": 3965 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.914681081081081, + "grad_norm": 25.748580932617188, + "learning_rate": 1.961650855779862e-07, + "loss": 7.149116039276123, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.3197903372347355, + "loss_total_v6": 0.446819756180048, + "step": 3966 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9149117117117117, + "grad_norm": 24.12025260925293, + "learning_rate": 1.9511310814207707e-07, + "loss": 6.352741241455078, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.2663395516574383, + "loss_total_v6": 0.3970463387668133, + "step": 3967 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9151423423423424, + "grad_norm": 23.800399780273438, + "learning_rate": 1.9406390288876586e-07, + "loss": 6.93724250793457, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.30596839264035225, + "loss_total_v6": 0.43357766419649124, + "step": 3968 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9153729729729729, + "grad_norm": 24.209009170532227, + "learning_rate": 1.930174704233928e-07, + "loss": 6.993715763092041, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.3073312249034643, + "loss_total_v6": 0.43710723146796227, + "step": 3969 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9156036036036036, + "grad_norm": 25.9306697845459, + "learning_rate": 1.919738113496983e-07, + "loss": 6.743500709533691, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.29186062701046467, + "loss_total_v6": 0.4214687794446945, + "step": 3970 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9158342342342343, + "grad_norm": 26.26540756225586, + "learning_rate": 1.9093292626982118e-07, + "loss": 6.4803032875061035, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.27786747738718987, + "loss_total_v6": 0.40501896291971207, + "step": 3971 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9160648648648648, + "grad_norm": 25.36853790283203, + "learning_rate": 1.8989481578430225e-07, + "loss": 7.672792434692383, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.3518029525876045, + "loss_total_v6": 0.4795495308935642, + "step": 3972 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9162954954954955, + "grad_norm": 25.434383392333984, + "learning_rate": 1.8885948049207847e-07, + "loss": 7.139243125915527, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.31729645282030106, + "loss_total_v6": 0.44620268791913986, + "step": 3973 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9165261261261262, + "grad_norm": 23.40590476989746, + "learning_rate": 1.878269209904887e-07, + "loss": 6.874756813049316, + "loss_alignment": 0.251220703125, + "loss_alignment_w": 0.1256103515625, + "loss_sft": 0.3038635775446892, + "loss_total_v6": 0.4296723045408726, + "step": 3974 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9167567567567567, + "grad_norm": 26.667451858520508, + "learning_rate": 1.8679713787526699e-07, + "loss": 6.407059669494629, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.2733812853693962, + "loss_total_v6": 0.4004412330687046, + "step": 3975 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9169873873873874, + "grad_norm": 26.112701416015625, + "learning_rate": 1.857701317405486e-07, + "loss": 6.773029327392578, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.2940418627113104, + "loss_total_v6": 0.42331433296203613, + "step": 3976 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.917218018018018, + "grad_norm": 25.746267318725586, + "learning_rate": 1.8474590317886509e-07, + "loss": 7.215156555175781, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.32161378860473633, + "loss_total_v6": 0.4509472884237766, + "step": 3977 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9174486486486486, + "grad_norm": 22.84363555908203, + "learning_rate": 1.8372445278114547e-07, + "loss": 7.512335777282715, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.3359913192689419, + "loss_total_v6": 0.4695209749042988, + "step": 3978 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9176792792792793, + "grad_norm": 25.238662719726562, + "learning_rate": 1.8270578113671667e-07, + "loss": 7.006532669067383, + "loss_alignment": 0.2484130859375, + "loss_alignment_w": 0.12420654296875, + "loss_sft": 0.3136407360434532, + "loss_total_v6": 0.4379083029925823, + "step": 3979 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.91790990990991, + "grad_norm": 26.180082321166992, + "learning_rate": 1.8168988883330185e-07, + "loss": 6.5070576667785645, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.2808213531970978, + "loss_total_v6": 0.4066911041736603, + "step": 3980 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9181405405405405, + "grad_norm": 25.18003273010254, + "learning_rate": 1.8067677645702108e-07, + "loss": 6.869723796844482, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.30170271918177605, + "loss_total_v6": 0.42935773357748985, + "step": 3981 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9183711711711712, + "grad_norm": 24.27387046813965, + "learning_rate": 1.7966644459238958e-07, + "loss": 6.9919233322143555, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.30805844627320766, + "loss_total_v6": 0.4369952157139778, + "step": 3982 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9186018018018018, + "grad_norm": 25.537729263305664, + "learning_rate": 1.7865889382231938e-07, + "loss": 7.087939262390137, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.31444089487195015, + "loss_total_v6": 0.44299618527293205, + "step": 3983 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9188324324324324, + "grad_norm": 23.7597713470459, + "learning_rate": 1.776541247281177e-07, + "loss": 7.368635654449463, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.33174026384949684, + "loss_total_v6": 0.46053971722722054, + "step": 3984 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9190630630630631, + "grad_norm": 22.859580993652344, + "learning_rate": 1.7665213788948532e-07, + "loss": 6.644926071166992, + "loss_alignment": 0.248779296875, + "loss_alignment_w": 0.1243896484375, + "loss_sft": 0.2905062735080719, + "loss_total_v6": 0.4153078980743885, + "step": 3985 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9192936936936937, + "grad_norm": 22.182443618774414, + "learning_rate": 1.7565293388452142e-07, + "loss": 7.118499279022217, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.3154048752039671, + "loss_total_v6": 0.44490621611475945, + "step": 3986 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9195243243243243, + "grad_norm": 22.38319969177246, + "learning_rate": 1.74656513289716e-07, + "loss": 6.740642070770264, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.2931926194578409, + "loss_total_v6": 0.4212901405990124, + "step": 3987 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.919754954954955, + "grad_norm": 24.85375213623047, + "learning_rate": 1.7366287667995417e-07, + "loss": 6.753369331359863, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.29502565041184425, + "loss_total_v6": 0.42208557575941086, + "step": 3988 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9199855855855856, + "grad_norm": 26.786344528198242, + "learning_rate": 1.726720246285163e-07, + "loss": 6.848567008972168, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.3013112172484398, + "loss_total_v6": 0.4280354641377926, + "step": 3989 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9202162162162162, + "grad_norm": 27.433927536010742, + "learning_rate": 1.7168395770707446e-07, + "loss": 7.116334915161133, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.31192789413034916, + "loss_total_v6": 0.4447709247469902, + "step": 3990 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9204468468468469, + "grad_norm": 24.222332000732422, + "learning_rate": 1.706986764856955e-07, + "loss": 7.183966636657715, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.3202137351036072, + "loss_total_v6": 0.4489979147911072, + "step": 3991 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9206774774774775, + "grad_norm": 22.356008529663086, + "learning_rate": 1.6971618153283632e-07, + "loss": 6.82853889465332, + "loss_alignment": 0.251220703125, + "loss_alignment_w": 0.1256103515625, + "loss_sft": 0.3008834011852741, + "loss_total_v6": 0.4267836846411228, + "step": 3992 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9209081081081081, + "grad_norm": 22.808908462524414, + "learning_rate": 1.6873647341535015e-07, + "loss": 6.848935604095459, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.2985571436583996, + "loss_total_v6": 0.4280584827065468, + "step": 3993 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9211387387387387, + "grad_norm": 21.79878044128418, + "learning_rate": 1.6775955269847767e-07, + "loss": 6.93060302734375, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.30543139949440956, + "loss_total_v6": 0.43316271901130676, + "step": 3994 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9213693693693694, + "grad_norm": 23.250234603881836, + "learning_rate": 1.6678541994585629e-07, + "loss": 7.135979652404785, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.3195033874362707, + "loss_total_v6": 0.44599873945116997, + "step": 3995 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9216, + "grad_norm": 24.974361419677734, + "learning_rate": 1.658140757195109e-07, + "loss": 6.7113037109375, + "loss_alignment": 0.2508544921875, + "loss_alignment_w": 0.12542724609375, + "loss_sft": 0.2940597329288721, + "loss_total_v6": 0.41945646330714226, + "step": 3996 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9218306306306306, + "grad_norm": 26.360029220581055, + "learning_rate": 1.648455205798599e-07, + "loss": 7.103912353515625, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3137607518583536, + "loss_total_v6": 0.44399452954530716, + "step": 3997 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9220612612612613, + "grad_norm": 26.02791976928711, + "learning_rate": 1.6387975508571242e-07, + "loss": 7.056683540344238, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.3152492679655552, + "loss_total_v6": 0.4410427175462246, + "step": 3998 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9222918918918919, + "grad_norm": 23.938711166381836, + "learning_rate": 1.6291677979426602e-07, + "loss": 7.066603660583496, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.31197824887931347, + "loss_total_v6": 0.4416627399623394, + "step": 3999 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9225225225225225, + "grad_norm": 25.166362762451172, + "learning_rate": 1.6195659526111184e-07, + "loss": 6.632422924041748, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.28787846490740776, + "loss_total_v6": 0.41452642902731895, + "step": 4000 + }, + { + "epoch": 0.9225225225225225, + "eval_loss": 0.5622133016586304, + "eval_loss_alignment": 0.23042861729452055, + "eval_loss_alignment_w": 0.11521430864726027, + "eval_loss_sft": 0.4469989977913088, + "eval_loss_total_v6": 0.5622133065406318, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 89.7876, + "eval_samples_per_second": 19.435, + "eval_steps_per_second": 2.439, + "step": 4000 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9227531531531532, + "grad_norm": 23.942420959472656, + "learning_rate": 1.6099920204022678e-07, + "loss": 6.997838497161865, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.308321338146925, + "loss_total_v6": 0.4373649023473263, + "step": 4001 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9229837837837838, + "grad_norm": 25.505159378051758, + "learning_rate": 1.600446006839812e-07, + "loss": 6.966422080993652, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3064340893179178, + "loss_total_v6": 0.435401376336813, + "step": 4002 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9232144144144144, + "grad_norm": 25.61471939086914, + "learning_rate": 1.5909279174313175e-07, + "loss": 7.149508953094482, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.31714460998773575, + "loss_total_v6": 0.44684432819485664, + "step": 4003 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9234450450450451, + "grad_norm": 23.24969482421875, + "learning_rate": 1.5814377576682527e-07, + "loss": 6.854141712188721, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.30049992352724075, + "loss_total_v6": 0.42838384583592415, + "step": 4004 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9236756756756757, + "grad_norm": 23.510623931884766, + "learning_rate": 1.5719755330259823e-07, + "loss": 6.92158842086792, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.30364725925028324, + "loss_total_v6": 0.4325992837548256, + "step": 4005 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9239063063063063, + "grad_norm": 24.243852615356445, + "learning_rate": 1.5625412489637337e-07, + "loss": 6.98021125793457, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.30747905746102333, + "loss_total_v6": 0.43626321852207184, + "step": 4006 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.924136936936937, + "grad_norm": 23.941844940185547, + "learning_rate": 1.5531349109246364e-07, + "loss": 7.0403900146484375, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.31018735468387604, + "loss_total_v6": 0.44002438336610794, + "step": 4007 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9243675675675675, + "grad_norm": 23.04007911682129, + "learning_rate": 1.5437565243356656e-07, + "loss": 6.8270955085754395, + "loss_alignment": 0.265869140625, + "loss_alignment_w": 0.1329345703125, + "loss_sft": 0.29392674565315247, + "loss_total_v6": 0.42669346928596497, + "step": 4008 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9245981981981982, + "grad_norm": 23.465192794799805, + "learning_rate": 1.5344060946076934e-07, + "loss": 6.898324489593506, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.3020712025463581, + "loss_total_v6": 0.431145291775465, + "step": 4009 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9248288288288288, + "grad_norm": 24.026607513427734, + "learning_rate": 1.5250836271354653e-07, + "loss": 7.133550643920898, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.31689488142728806, + "loss_total_v6": 0.44584691524505615, + "step": 4010 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9250594594594594, + "grad_norm": 25.545419692993164, + "learning_rate": 1.5157891272975677e-07, + "loss": 7.131875991821289, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.31672918424010277, + "loss_total_v6": 0.44574224576354027, + "step": 4011 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9252900900900901, + "grad_norm": 24.778564453125, + "learning_rate": 1.5065226004564892e-07, + "loss": 6.961319923400879, + "loss_alignment": 0.2642822265625, + "loss_alignment_w": 0.13214111328125, + "loss_sft": 0.3033991251140833, + "loss_total_v6": 0.43508247658610344, + "step": 4012 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9255207207207207, + "grad_norm": 27.197589874267578, + "learning_rate": 1.4972840519585418e-07, + "loss": 6.693493843078613, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.28849107399582863, + "loss_total_v6": 0.41834336146712303, + "step": 4013 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9257513513513513, + "grad_norm": 25.183425903320312, + "learning_rate": 1.488073487133912e-07, + "loss": 7.276252746582031, + "loss_alignment": 0.2655029296875, + "loss_alignment_w": 0.13275146484375, + "loss_sft": 0.32244158163666725, + "loss_total_v6": 0.45476580038666725, + "step": 4014 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.925981981981982, + "grad_norm": 25.2109432220459, + "learning_rate": 1.4788909112966544e-07, + "loss": 7.271897315979004, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.32543473318219185, + "loss_total_v6": 0.45449358969926834, + "step": 4015 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9262126126126126, + "grad_norm": 26.933868408203125, + "learning_rate": 1.4697363297446476e-07, + "loss": 6.9491987228393555, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3056933209300041, + "loss_total_v6": 0.4343249127268791, + "step": 4016 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9264432432432432, + "grad_norm": 24.28647232055664, + "learning_rate": 1.4606097477596504e-07, + "loss": 6.951578617095947, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.3035837411880493, + "loss_total_v6": 0.4344736598432064, + "step": 4017 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9266738738738739, + "grad_norm": 24.356204986572266, + "learning_rate": 1.4515111706072337e-07, + "loss": 7.293231010437012, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.3269512318074703, + "loss_total_v6": 0.4558269567787647, + "step": 4018 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9269045045045045, + "grad_norm": 24.78683853149414, + "learning_rate": 1.4424406035368378e-07, + "loss": 6.564266681671143, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.283969659358263, + "loss_total_v6": 0.4102666601538658, + "step": 4019 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9271351351351351, + "grad_norm": 25.304035186767578, + "learning_rate": 1.4333980517817202e-07, + "loss": 7.27676248550415, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.323709387332201, + "loss_total_v6": 0.4547976404428482, + "step": 4020 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9273657657657658, + "grad_norm": 23.81704330444336, + "learning_rate": 1.4243835205590028e-07, + "loss": 7.356036186218262, + "loss_alignment": 0.266845703125, + "loss_alignment_w": 0.1334228515625, + "loss_sft": 0.32652779668569565, + "loss_total_v6": 0.45975228026509285, + "step": 4021 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9275963963963963, + "grad_norm": 23.44432258605957, + "learning_rate": 1.4153970150696085e-07, + "loss": 7.034375190734863, + "loss_alignment": 0.253173828125, + "loss_alignment_w": 0.1265869140625, + "loss_sft": 0.31306153163313866, + "loss_total_v6": 0.43964846804738045, + "step": 4022 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.927827027027027, + "grad_norm": 20.17484474182129, + "learning_rate": 1.4064385404983227e-07, + "loss": 6.672379493713379, + "loss_alignment": 0.2509765625, + "loss_alignment_w": 0.12548828125, + "loss_sft": 0.2914134059101343, + "loss_total_v6": 0.41702374815940857, + "step": 4023 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9280576576576577, + "grad_norm": 25.342164993286133, + "learning_rate": 1.3975081020137392e-07, + "loss": 6.934229373931885, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.30178227834403515, + "loss_total_v6": 0.4333893246948719, + "step": 4024 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9282882882882882, + "grad_norm": 26.0682373046875, + "learning_rate": 1.3886057047682755e-07, + "loss": 6.697849750518799, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2888243477791548, + "loss_total_v6": 0.4186156168580055, + "step": 4025 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9285189189189189, + "grad_norm": 24.66844940185547, + "learning_rate": 1.3797313538981782e-07, + "loss": 6.739578723907471, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.29324818402528763, + "loss_total_v6": 0.4212236665189266, + "step": 4026 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9287495495495496, + "grad_norm": 25.56226921081543, + "learning_rate": 1.3708850545235074e-07, + "loss": 6.291965484619141, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.2660658098757267, + "loss_total_v6": 0.3932478353381157, + "step": 4027 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9289801801801801, + "grad_norm": 24.96491241455078, + "learning_rate": 1.3620668117481471e-07, + "loss": 6.80881404876709, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.2966446466743946, + "loss_total_v6": 0.4255509003996849, + "step": 4028 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9292108108108108, + "grad_norm": 25.571374893188477, + "learning_rate": 1.3532766306597778e-07, + "loss": 6.655726909637451, + "loss_alignment": 0.2501220703125, + "loss_alignment_w": 0.12506103515625, + "loss_sft": 0.2909371592104435, + "loss_total_v6": 0.4159829318523407, + "step": 4029 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9294414414414415, + "grad_norm": 23.61859893798828, + "learning_rate": 1.3445145163299088e-07, + "loss": 7.050205707550049, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31005312502384186, + "loss_total_v6": 0.44063783809542656, + "step": 4030 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.929672072072072, + "grad_norm": 22.979564666748047, + "learning_rate": 1.335780473813847e-07, + "loss": 7.113160133361816, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.31815342977643013, + "loss_total_v6": 0.4445725083351135, + "step": 4031 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9299027027027027, + "grad_norm": 25.29637336730957, + "learning_rate": 1.3270745081506996e-07, + "loss": 7.3640899658203125, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3288927301764488, + "loss_total_v6": 0.4602556340396404, + "step": 4032 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9301333333333334, + "grad_norm": 23.323745727539062, + "learning_rate": 1.3183966243633884e-07, + "loss": 6.805477619171143, + "loss_alignment": 0.252197265625, + "loss_alignment_w": 0.1260986328125, + "loss_sft": 0.29890803806483746, + "loss_total_v6": 0.4253423623740673, + "step": 4033 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9303639639639639, + "grad_norm": 23.9410400390625, + "learning_rate": 1.3097468274586132e-07, + "loss": 6.701398849487305, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2890767063945532, + "loss_total_v6": 0.41883744671940804, + "step": 4034 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9305945945945946, + "grad_norm": 27.228439331054688, + "learning_rate": 1.3011251224268828e-07, + "loss": 7.233419418334961, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.32183967903256416, + "loss_total_v6": 0.45208870992064476, + "step": 4035 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9308252252252253, + "grad_norm": 23.06603240966797, + "learning_rate": 1.292531514242501e-07, + "loss": 6.765267372131348, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.2954946178942919, + "loss_total_v6": 0.4228292144834995, + "step": 4036 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9310558558558558, + "grad_norm": 24.753686904907227, + "learning_rate": 1.2839660078635464e-07, + "loss": 7.042062759399414, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.31309949234128, + "loss_total_v6": 0.4401289150118828, + "step": 4037 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9312864864864865, + "grad_norm": 29.539064407348633, + "learning_rate": 1.2754286082319044e-07, + "loss": 7.046374797821045, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.31135484017431736, + "loss_total_v6": 0.4403984285891056, + "step": 4038 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9315171171171172, + "grad_norm": 24.767513275146484, + "learning_rate": 1.266919320273219e-07, + "loss": 6.895228385925293, + "loss_alignment": 0.2501220703125, + "loss_alignment_w": 0.12506103515625, + "loss_sft": 0.30601280741393566, + "loss_total_v6": 0.4309517592191696, + "step": 4039 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9317477477477477, + "grad_norm": 25.0363826751709, + "learning_rate": 1.2584381488969454e-07, + "loss": 7.051865577697754, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.31427672877907753, + "loss_total_v6": 0.44074157252907753, + "step": 4040 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9319783783783784, + "grad_norm": 28.29424285888672, + "learning_rate": 1.249985098996287e-07, + "loss": 7.262588024139404, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.32474612072110176, + "loss_total_v6": 0.45391176640987396, + "step": 4041 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.932209009009009, + "grad_norm": 24.9166259765625, + "learning_rate": 1.241560175448242e-07, + "loss": 7.311028957366943, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.32675134018063545, + "loss_total_v6": 0.45693931356072426, + "step": 4042 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9324396396396396, + "grad_norm": 26.67698097229004, + "learning_rate": 1.2331633831135835e-07, + "loss": 7.449667930603027, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3364080861210823, + "loss_total_v6": 0.4656042642891407, + "step": 4043 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9326702702702703, + "grad_norm": 22.843582153320312, + "learning_rate": 1.2247947268368366e-07, + "loss": 7.137957572937012, + "loss_alignment": 0.2498779296875, + "loss_alignment_w": 0.12493896484375, + "loss_sft": 0.3212291821837425, + "loss_total_v6": 0.4461223669350147, + "step": 4044 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9329009009009009, + "grad_norm": 24.470781326293945, + "learning_rate": 1.2164542114463053e-07, + "loss": 7.037172317504883, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3120919596403837, + "loss_total_v6": 0.43982328474521637, + "step": 4045 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9331315315315315, + "grad_norm": 23.271778106689453, + "learning_rate": 1.208141841754057e-07, + "loss": 7.101736545562744, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.31432665325701237, + "loss_total_v6": 0.4438585303723812, + "step": 4046 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9333621621621622, + "grad_norm": 26.479135513305664, + "learning_rate": 1.1998576225559112e-07, + "loss": 7.2756547927856445, + "loss_alignment": 0.2650146484375, + "loss_alignment_w": 0.13250732421875, + "loss_sft": 0.3221753239631653, + "loss_total_v6": 0.4547284245491028, + "step": 4047 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9335927927927928, + "grad_norm": 23.824440002441406, + "learning_rate": 1.191601558631461e-07, + "loss": 6.852876663208008, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.2981931045651436, + "loss_total_v6": 0.4283048063516617, + "step": 4048 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9338234234234234, + "grad_norm": 23.533123016357422, + "learning_rate": 1.1833736547440466e-07, + "loss": 6.667085647583008, + "loss_alignment": 0.2642822265625, + "loss_alignment_w": 0.13214111328125, + "loss_sft": 0.2845059596002102, + "loss_total_v6": 0.4166928566992283, + "step": 4049 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9340540540540541, + "grad_norm": 27.02227210998535, + "learning_rate": 1.1751739156407649e-07, + "loss": 6.3243327140808105, + "loss_alignment": 0.25048828125, + "loss_alignment_w": 0.125244140625, + "loss_sft": 0.2698130141943693, + "loss_total_v6": 0.39527078345417976, + "step": 4050 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9342846846846847, + "grad_norm": 23.52703285217285, + "learning_rate": 1.1670023460524538e-07, + "loss": 7.455169200897217, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.33858293294906616, + "loss_total_v6": 0.46594807133078575, + "step": 4051 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9345153153153153, + "grad_norm": 23.019155502319336, + "learning_rate": 1.1588589506937198e-07, + "loss": 7.207062244415283, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.32390024699270725, + "loss_total_v6": 0.4504414014518261, + "step": 4052 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.934745945945946, + "grad_norm": 25.701475143432617, + "learning_rate": 1.1507437342628769e-07, + "loss": 7.108736038208008, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.3153897635638714, + "loss_total_v6": 0.4442960098385811, + "step": 4053 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9349765765765766, + "grad_norm": 25.352949142456055, + "learning_rate": 1.1426567014420297e-07, + "loss": 7.172935962677002, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3209586553275585, + "loss_total_v6": 0.4483084976673126, + "step": 4054 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9352072072072072, + "grad_norm": 23.071083068847656, + "learning_rate": 1.134597856896974e-07, + "loss": 6.499045372009277, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.277024706825614, + "loss_total_v6": 0.40619034320116043, + "step": 4055 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9354378378378378, + "grad_norm": 22.983638763427734, + "learning_rate": 1.1265672052772791e-07, + "loss": 6.380789756774902, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.27308221347630024, + "loss_total_v6": 0.3987993746995926, + "step": 4056 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9356684684684685, + "grad_norm": 21.34698486328125, + "learning_rate": 1.1185647512162334e-07, + "loss": 7.029624938964844, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.31192542612552643, + "loss_total_v6": 0.4393515810370445, + "step": 4057 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9358990990990991, + "grad_norm": 26.820505142211914, + "learning_rate": 1.1105904993308547e-07, + "loss": 7.196843147277832, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3202708326280117, + "loss_total_v6": 0.4498026929795742, + "step": 4058 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9361297297297297, + "grad_norm": 25.207460403442383, + "learning_rate": 1.1026444542219017e-07, + "loss": 6.791705131530762, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.29424781538546085, + "loss_total_v6": 0.4244815669953823, + "step": 4059 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9363603603603604, + "grad_norm": 27.523168563842773, + "learning_rate": 1.094726620473835e-07, + "loss": 7.005043029785156, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.30796289071440697, + "loss_total_v6": 0.43781518191099167, + "step": 4060 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.936590990990991, + "grad_norm": 23.88092613220215, + "learning_rate": 1.086837002654867e-07, + "loss": 7.188695907592773, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.31941068544983864, + "loss_total_v6": 0.44929349422454834, + "step": 4061 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9368216216216216, + "grad_norm": 21.12816619873047, + "learning_rate": 1.0789756053169065e-07, + "loss": 6.3778076171875, + "loss_alignment": 0.24755859375, + "loss_alignment_w": 0.123779296875, + "loss_sft": 0.27481841295957565, + "loss_total_v6": 0.39861297979950905, + "step": 4062 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9370522522522523, + "grad_norm": 25.596982955932617, + "learning_rate": 1.0711424329956033e-07, + "loss": 6.949207782745361, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.30438164062798023, + "loss_total_v6": 0.434325497597456, + "step": 4063 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9372828828828829, + "grad_norm": 24.06736183166504, + "learning_rate": 1.0633374902103088e-07, + "loss": 6.940884590148926, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.304685415700078, + "loss_total_v6": 0.43380527570843697, + "step": 4064 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9375135135135135, + "grad_norm": 26.319103240966797, + "learning_rate": 1.0555607814640767e-07, + "loss": 7.01133918762207, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.3129645138978958, + "loss_total_v6": 0.4382086805999279, + "step": 4065 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9377441441441442, + "grad_norm": 24.046199798583984, + "learning_rate": 1.0478123112437011e-07, + "loss": 7.080432415008545, + "loss_alignment": 0.2518310546875, + "loss_alignment_w": 0.12591552734375, + "loss_sft": 0.31653522700071335, + "loss_total_v6": 0.44252703711390495, + "step": 4066 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9379747747747748, + "grad_norm": 26.517173767089844, + "learning_rate": 1.040092084019656e-07, + "loss": 6.898670196533203, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.3025200553238392, + "loss_total_v6": 0.4311668984591961, + "step": 4067 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9382054054054054, + "grad_norm": 24.145591735839844, + "learning_rate": 1.0324001042461395e-07, + "loss": 7.007108688354492, + "loss_alignment": 0.2557373046875, + "loss_alignment_w": 0.12786865234375, + "loss_sft": 0.3099687956273556, + "loss_total_v6": 0.43794427812099457, + "step": 4068 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.938436036036036, + "grad_norm": 24.735258102416992, + "learning_rate": 1.0247363763610408e-07, + "loss": 7.258588790893555, + "loss_alignment": 0.2608642578125, + "loss_alignment_w": 0.13043212890625, + "loss_sft": 0.32312287017703056, + "loss_total_v6": 0.45366179943084717, + "step": 4069 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9386666666666666, + "grad_norm": 28.35607147216797, + "learning_rate": 1.0171009047859504e-07, + "loss": 7.09446907043457, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.31628334894776344, + "loss_total_v6": 0.44340432062745094, + "step": 4070 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9388972972972973, + "grad_norm": 22.790924072265625, + "learning_rate": 1.0094936939261669e-07, + "loss": 7.000735282897949, + "loss_alignment": 0.252197265625, + "loss_alignment_w": 0.1260986328125, + "loss_sft": 0.3110811375081539, + "loss_total_v6": 0.4375459738075733, + "step": 4071 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9391279279279279, + "grad_norm": 24.05650520324707, + "learning_rate": 1.0019147481706626e-07, + "loss": 7.298792839050293, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3267647847533226, + "loss_total_v6": 0.4561745822429657, + "step": 4072 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9393585585585585, + "grad_norm": 26.684274673461914, + "learning_rate": 9.943640718921177e-08, + "loss": 7.001452922821045, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.3103782497346401, + "loss_total_v6": 0.4375907890498638, + "step": 4073 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9395891891891892, + "grad_norm": 28.05307960510254, + "learning_rate": 9.868416694469085e-08, + "loss": 6.9016571044921875, + "loss_alignment": 0.251708984375, + "loss_alignment_w": 0.1258544921875, + "loss_sft": 0.3057431988418102, + "loss_total_v6": 0.4313535578548908, + "step": 4074 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9398198198198198, + "grad_norm": 26.653413772583008, + "learning_rate": 9.793475451750856e-08, + "loss": 6.501247406005859, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.2796037048101425, + "loss_total_v6": 0.4063279516994953, + "step": 4075 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9400504504504504, + "grad_norm": 24.44666862487793, + "learning_rate": 9.718817034003903e-08, + "loss": 7.288024425506592, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.3229026719927788, + "loss_total_v6": 0.4555015414953232, + "step": 4076 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9402810810810811, + "grad_norm": 24.267486572265625, + "learning_rate": 9.64444148430238e-08, + "loss": 7.152674674987793, + "loss_alignment": 0.249755859375, + "loss_alignment_w": 0.1248779296875, + "loss_sft": 0.32169121876358986, + "loss_total_v6": 0.44704215973615646, + "step": 4077 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9405117117117117, + "grad_norm": 26.468868255615234, + "learning_rate": 9.570348845557409e-08, + "loss": 6.705940246582031, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.29114577919244766, + "loss_total_v6": 0.41912124305963516, + "step": 4078 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9407423423423423, + "grad_norm": 26.517032623291016, + "learning_rate": 9.496539160516738e-08, + "loss": 6.587437629699707, + "loss_alignment": 0.2481689453125, + "loss_alignment_w": 0.12408447265625, + "loss_sft": 0.287325207144022, + "loss_total_v6": 0.411714855581522, + "step": 4079 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.940972972972973, + "grad_norm": 26.538101196289062, + "learning_rate": 9.423012471764914e-08, + "loss": 7.474991798400879, + "loss_alignment": 0.256103515625, + "loss_alignment_w": 0.1280517578125, + "loss_sft": 0.3388911075890064, + "loss_total_v6": 0.46718698740005493, + "step": 4080 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9412036036036036, + "grad_norm": 28.536487579345703, + "learning_rate": 9.349768821723116e-08, + "loss": 7.122713565826416, + "loss_alignment": 0.2489013671875, + "loss_alignment_w": 0.12445068359375, + "loss_sft": 0.3204747885465622, + "loss_total_v6": 0.445169597864151, + "step": 4081 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9414342342342342, + "grad_norm": 25.053449630737305, + "learning_rate": 9.276808252649483e-08, + "loss": 7.169943809509277, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.31977980211377144, + "loss_total_v6": 0.44812148064374924, + "step": 4082 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9416648648648649, + "grad_norm": 21.997936248779297, + "learning_rate": 9.204130806638511e-08, + "loss": 6.896265506744385, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.30336155742406845, + "loss_total_v6": 0.43101659044623375, + "step": 4083 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9418954954954954, + "grad_norm": 22.042152404785156, + "learning_rate": 9.131736525621604e-08, + "loss": 6.916928291320801, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.30375270172953606, + "loss_total_v6": 0.43230801075696945, + "step": 4084 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9421261261261261, + "grad_norm": 22.25571060180664, + "learning_rate": 9.059625451366738e-08, + "loss": 7.001593112945557, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.3084339387714863, + "loss_total_v6": 0.4375995732843876, + "step": 4085 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9423567567567568, + "grad_norm": 25.007516860961914, + "learning_rate": 8.987797625478245e-08, + "loss": 7.1239213943481445, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31502658873796463, + "loss_total_v6": 0.44524508714675903, + "step": 4086 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9425873873873873, + "grad_norm": 24.534854888916016, + "learning_rate": 8.91625308939742e-08, + "loss": 6.7976298332214355, + "loss_alignment": 0.2515869140625, + "loss_alignment_w": 0.12579345703125, + "loss_sft": 0.2991499602794647, + "loss_total_v6": 0.4248518608510494, + "step": 4087 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.942818018018018, + "grad_norm": 25.792985916137695, + "learning_rate": 8.844991884401855e-08, + "loss": 6.696080207824707, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2901633121073246, + "loss_total_v6": 0.4185049906373024, + "step": 4088 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9430486486486487, + "grad_norm": 24.737279891967773, + "learning_rate": 8.774014051605717e-08, + "loss": 6.687280178070068, + "loss_alignment": 0.251953125, + "loss_alignment_w": 0.1259765625, + "loss_sft": 0.2921615391969681, + "loss_total_v6": 0.417955007404089, + "step": 4089 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9432792792792792, + "grad_norm": 25.08712387084961, + "learning_rate": 8.703319631959806e-08, + "loss": 7.286433696746826, + "loss_alignment": 0.261962890625, + "loss_alignment_w": 0.1309814453125, + "loss_sft": 0.32443591207265854, + "loss_total_v6": 0.45540209114551544, + "step": 4090 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9435099099099099, + "grad_norm": 23.262737274169922, + "learning_rate": 8.632908666251161e-08, + "loss": 7.12309455871582, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.31772148236632347, + "loss_total_v6": 0.44519341364502907, + "step": 4091 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9437405405405406, + "grad_norm": 26.038246154785156, + "learning_rate": 8.56278119510362e-08, + "loss": 6.845348834991455, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.29833294451236725, + "loss_total_v6": 0.42783429101109505, + "step": 4092 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9439711711711711, + "grad_norm": 28.813627243041992, + "learning_rate": 8.492937258977152e-08, + "loss": 6.871241569519043, + "loss_alignment": 0.2620849609375, + "loss_alignment_w": 0.13104248046875, + "loss_sft": 0.2984253764152527, + "loss_total_v6": 0.4294525980949402, + "step": 4093 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9442018018018018, + "grad_norm": 25.179672241210938, + "learning_rate": 8.423376898168246e-08, + "loss": 7.148777008056641, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.31751085817813873, + "loss_total_v6": 0.44679858162999153, + "step": 4094 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9444324324324325, + "grad_norm": 24.160951614379883, + "learning_rate": 8.354100152809907e-08, + "loss": 6.909255027770996, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.30505839735269547, + "loss_total_v6": 0.43182842433452606, + "step": 4095 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.944663063063063, + "grad_norm": 24.437700271606445, + "learning_rate": 8.285107062871334e-08, + "loss": 7.40933895111084, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.33278888277709484, + "loss_total_v6": 0.463083665817976, + "step": 4096 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9448936936936937, + "grad_norm": 25.21337127685547, + "learning_rate": 8.216397668158127e-08, + "loss": 6.902833938598633, + "loss_alignment": 0.2518310546875, + "loss_alignment_w": 0.12591552734375, + "loss_sft": 0.3053590003401041, + "loss_total_v6": 0.43142712861299515, + "step": 4097 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9451243243243244, + "grad_norm": 26.526777267456055, + "learning_rate": 8.147972008312244e-08, + "loss": 6.9694976806640625, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.30655003897845745, + "loss_total_v6": 0.4355936124920845, + "step": 4098 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9453549549549549, + "grad_norm": 24.90041732788086, + "learning_rate": 8.079830122811827e-08, + "loss": 6.8175458908081055, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.2973429672420025, + "loss_total_v6": 0.4260966181755066, + "step": 4099 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9455855855855856, + "grad_norm": 25.654367446899414, + "learning_rate": 8.011972050971484e-08, + "loss": 7.034201622009277, + "loss_alignment": 0.2515869140625, + "loss_alignment_w": 0.12579345703125, + "loss_sft": 0.31395097076892853, + "loss_total_v6": 0.4396376200020313, + "step": 4100 + }, + { + "epoch": 0.9455855855855856, + "eval_loss": 0.562717080116272, + "eval_loss_alignment": 0.23135612871004566, + "eval_loss_alignment_w": 0.11567806435502283, + "eval_loss_sft": 0.4470390196226231, + "eval_loss_total_v6": 0.5627170830760917, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 89.7204, + "eval_samples_per_second": 19.449, + "eval_steps_per_second": 2.441, + "step": 4100 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9458162162162163, + "grad_norm": 27.57133674621582, + "learning_rate": 7.944397831941952e-08, + "loss": 7.136780261993408, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.3178291693329811, + "loss_total_v6": 0.4460487887263298, + "step": 4101 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9460468468468468, + "grad_norm": 23.639835357666016, + "learning_rate": 7.877107504710213e-08, + "loss": 6.99226188659668, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.30640115588903427, + "loss_total_v6": 0.4370163753628731, + "step": 4102 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9462774774774775, + "grad_norm": 24.321659088134766, + "learning_rate": 7.810101108099433e-08, + "loss": 7.292376518249512, + "loss_alignment": 0.26220703125, + "loss_alignment_w": 0.131103515625, + "loss_sft": 0.3245174027979374, + "loss_total_v6": 0.4557735174894333, + "step": 4103 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9465081081081081, + "grad_norm": 23.701509475708008, + "learning_rate": 7.743378680769021e-08, + "loss": 6.880425453186035, + "loss_alignment": 0.2520751953125, + "loss_alignment_w": 0.12603759765625, + "loss_sft": 0.3035465218126774, + "loss_total_v6": 0.4300266057252884, + "step": 4104 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9467387387387387, + "grad_norm": 25.43344497680664, + "learning_rate": 7.676940261214516e-08, + "loss": 6.699748516082764, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.2903620693832636, + "loss_total_v6": 0.41873428598046303, + "step": 4105 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9469693693693694, + "grad_norm": 25.394784927368164, + "learning_rate": 7.610785887767647e-08, + "loss": 6.977513313293457, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.30862266570329666, + "loss_total_v6": 0.43609459325671196, + "step": 4106 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9472, + "grad_norm": 22.80183982849121, + "learning_rate": 7.544915598596158e-08, + "loss": 7.537437915802002, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3422141298651695, + "loss_total_v6": 0.4710898697376251, + "step": 4107 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9474306306306306, + "grad_norm": 23.999555587768555, + "learning_rate": 7.479329431703986e-08, + "loss": 6.978068828582764, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.3090388644486666, + "loss_total_v6": 0.4361293166875839, + "step": 4108 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9476612612612613, + "grad_norm": 26.2685489654541, + "learning_rate": 7.414027424931137e-08, + "loss": 7.000340461730957, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.30902700312435627, + "loss_total_v6": 0.4375212602317333, + "step": 4109 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9478918918918919, + "grad_norm": 24.730257034301758, + "learning_rate": 7.349009615953585e-08, + "loss": 7.134748458862305, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.31683245673775673, + "loss_total_v6": 0.44592180475592613, + "step": 4110 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9481225225225225, + "grad_norm": 23.649127960205078, + "learning_rate": 7.284276042283489e-08, + "loss": 6.713534832000732, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.29035397060215473, + "loss_total_v6": 0.4195959232747555, + "step": 4111 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9483531531531532, + "grad_norm": 25.974674224853516, + "learning_rate": 7.21982674126881e-08, + "loss": 6.908254623413086, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.29950274899601936, + "loss_total_v6": 0.43176591396331787, + "step": 4112 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9485837837837838, + "grad_norm": 25.936004638671875, + "learning_rate": 7.155661750093745e-08, + "loss": 6.706514835357666, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.289457481354475, + "loss_total_v6": 0.4191571772098541, + "step": 4113 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9488144144144144, + "grad_norm": 22.806604385375977, + "learning_rate": 7.09178110577824e-08, + "loss": 7.2195940017700195, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.32123501412570477, + "loss_total_v6": 0.4512246362864971, + "step": 4114 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.949045045045045, + "grad_norm": 24.00418472290039, + "learning_rate": 7.028184845178254e-08, + "loss": 6.843061447143555, + "loss_alignment": 0.2509765625, + "loss_alignment_w": 0.12548828125, + "loss_sft": 0.3018368389457464, + "loss_total_v6": 0.427691325545311, + "step": 4115 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9492756756756757, + "grad_norm": 28.013307571411133, + "learning_rate": 6.964873004985717e-08, + "loss": 6.844566822052002, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.30089334212243557, + "loss_total_v6": 0.4277854263782501, + "step": 4116 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9495063063063063, + "grad_norm": 22.12483787536621, + "learning_rate": 6.901845621728464e-08, + "loss": 6.92840576171875, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.306087464094162, + "loss_total_v6": 0.4330253340303898, + "step": 4117 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.949736936936937, + "grad_norm": 23.38802719116211, + "learning_rate": 6.839102731770242e-08, + "loss": 7.3356523513793945, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.3302129078656435, + "loss_total_v6": 0.45847827941179276, + "step": 4118 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9499675675675676, + "grad_norm": 23.77912712097168, + "learning_rate": 6.776644371310426e-08, + "loss": 7.093493938446045, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3126518353819847, + "loss_total_v6": 0.4433433711528778, + "step": 4119 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9501981981981982, + "grad_norm": 26.978187561035156, + "learning_rate": 6.71447057638458e-08, + "loss": 7.03158712387085, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3118954487144947, + "loss_total_v6": 0.4394741952419281, + "step": 4120 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9504288288288288, + "grad_norm": 24.772130966186523, + "learning_rate": 6.652581382863787e-08, + "loss": 6.923425674438477, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3038841634988785, + "loss_total_v6": 0.4327141158282757, + "step": 4121 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9506594594594595, + "grad_norm": 27.41098403930664, + "learning_rate": 6.590976826455153e-08, + "loss": 7.117558002471924, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.3163226004689932, + "loss_total_v6": 0.44484738633036613, + "step": 4122 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9508900900900901, + "grad_norm": 24.80782127380371, + "learning_rate": 6.529656942701468e-08, + "loss": 6.840693473815918, + "loss_alignment": 0.2540283203125, + "loss_alignment_w": 0.12701416015625, + "loss_sft": 0.30086488649249077, + "loss_total_v6": 0.42754336446523666, + "step": 4123 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9511207207207207, + "grad_norm": 24.06102180480957, + "learning_rate": 6.468621766981153e-08, + "loss": 7.117256164550781, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.3172192871570587, + "loss_total_v6": 0.4448285289108753, + "step": 4124 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9513513513513514, + "grad_norm": 23.114337921142578, + "learning_rate": 6.407871334508598e-08, + "loss": 7.1683807373046875, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.31894972920417786, + "loss_total_v6": 0.44802381098270416, + "step": 4125 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.951581981981982, + "grad_norm": 27.619613647460938, + "learning_rate": 6.347405680333818e-08, + "loss": 6.947521209716797, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.30444410257041454, + "loss_total_v6": 0.4342200793325901, + "step": 4126 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9518126126126126, + "grad_norm": 26.363977432250977, + "learning_rate": 6.28722483934241e-08, + "loss": 7.472223281860352, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.33557475358247757, + "loss_total_v6": 0.4670139439404011, + "step": 4127 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9520432432432433, + "grad_norm": 23.590553283691406, + "learning_rate": 6.227328846255931e-08, + "loss": 6.915925979614258, + "loss_alignment": 0.262451171875, + "loss_alignment_w": 0.1312255859375, + "loss_sft": 0.30141652189195156, + "loss_total_v6": 0.4322453700006008, + "step": 4128 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9522738738738739, + "grad_norm": 26.062219619750977, + "learning_rate": 6.167717735631184e-08, + "loss": 7.093234062194824, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.3137799836695194, + "loss_total_v6": 0.4433271028101444, + "step": 4129 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9525045045045045, + "grad_norm": 29.298295974731445, + "learning_rate": 6.108391541861047e-08, + "loss": 6.996996879577637, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.308283980935812, + "loss_total_v6": 0.4373123161494732, + "step": 4130 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9527351351351351, + "grad_norm": 25.906892776489258, + "learning_rate": 6.04935029917364e-08, + "loss": 6.940686225891113, + "loss_alignment": 0.2589111328125, + "loss_alignment_w": 0.12945556640625, + "loss_sft": 0.3042457513511181, + "loss_total_v6": 0.433792881667614, + "step": 4131 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9529657657657657, + "grad_norm": 26.976638793945312, + "learning_rate": 5.990594041632992e-08, + "loss": 6.865137577056885, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.30170599929988384, + "loss_total_v6": 0.4290711097419262, + "step": 4132 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9531963963963964, + "grad_norm": 24.02523422241211, + "learning_rate": 5.932122803138429e-08, + "loss": 7.46375036239624, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.3369830548763275, + "loss_total_v6": 0.4664843939244747, + "step": 4133 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.953427027027027, + "grad_norm": 22.52488899230957, + "learning_rate": 5.8739366174250246e-08, + "loss": 6.945723533630371, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.306040707975626, + "loss_total_v6": 0.4341077208518982, + "step": 4134 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9536576576576576, + "grad_norm": 24.899654388427734, + "learning_rate": 5.816035518063368e-08, + "loss": 6.862731456756592, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.2992667891085148, + "loss_total_v6": 0.4289207234978676, + "step": 4135 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9538882882882883, + "grad_norm": 25.584070205688477, + "learning_rate": 5.758419538459459e-08, + "loss": 6.747668266296387, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.2959205340594053, + "loss_total_v6": 0.42172926664352417, + "step": 4136 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9541189189189189, + "grad_norm": 27.014095306396484, + "learning_rate": 5.701088711855041e-08, + "loss": 6.607873916625977, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.2836281359195709, + "loss_total_v6": 0.4129921421408653, + "step": 4137 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9543495495495495, + "grad_norm": 23.36618423461914, + "learning_rate": 5.6440430713269325e-08, + "loss": 7.019376277923584, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3074396476149559, + "loss_total_v6": 0.4387110099196434, + "step": 4138 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9545801801801802, + "grad_norm": 28.02335548400879, + "learning_rate": 5.587282649787862e-08, + "loss": 7.284377098083496, + "loss_alignment": 0.2486572265625, + "loss_alignment_w": 0.12432861328125, + "loss_sft": 0.3307160846889019, + "loss_total_v6": 0.4552735798060894, + "step": 4139 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9548108108108108, + "grad_norm": 23.378555297851562, + "learning_rate": 5.5308074799856334e-08, + "loss": 6.865266799926758, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.30061545968055725, + "loss_total_v6": 0.42907917872071266, + "step": 4140 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9550414414414414, + "grad_norm": 25.511062622070312, + "learning_rate": 5.474617594503739e-08, + "loss": 6.774442195892334, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.29498464055359364, + "loss_total_v6": 0.4234026223421097, + "step": 4141 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9552720720720721, + "grad_norm": 27.55914878845215, + "learning_rate": 5.418713025760969e-08, + "loss": 7.159890174865723, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3192430306226015, + "loss_total_v6": 0.44749314337968826, + "step": 4142 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9555027027027027, + "grad_norm": 27.62607192993164, + "learning_rate": 5.36309380601141e-08, + "loss": 6.823802947998047, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.298512214794755, + "loss_total_v6": 0.42648766189813614, + "step": 4143 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9557333333333333, + "grad_norm": 23.486434936523438, + "learning_rate": 5.307759967344672e-08, + "loss": 6.843925476074219, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.29819823801517487, + "loss_total_v6": 0.42774536088109016, + "step": 4144 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.955963963963964, + "grad_norm": 23.121482849121094, + "learning_rate": 5.252711541685662e-08, + "loss": 7.220487594604492, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.32184017822146416, + "loss_total_v6": 0.45128047838807106, + "step": 4145 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9561945945945945, + "grad_norm": 25.155275344848633, + "learning_rate": 5.197948560794586e-08, + "loss": 7.415623664855957, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3306182101368904, + "loss_total_v6": 0.4634764865040779, + "step": 4146 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9564252252252252, + "grad_norm": 24.69803237915039, + "learning_rate": 5.1434710562669465e-08, + "loss": 6.619551181793213, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.2856549359858036, + "loss_total_v6": 0.4137219525873661, + "step": 4147 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9566558558558559, + "grad_norm": 24.94186019897461, + "learning_rate": 5.089279059533658e-08, + "loss": 7.176910400390625, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.31997108459472656, + "loss_total_v6": 0.44855690374970436, + "step": 4148 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9568864864864864, + "grad_norm": 26.190458297729492, + "learning_rate": 5.035372601860766e-08, + "loss": 6.944288730621338, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3054322339594364, + "loss_total_v6": 0.4340180493891239, + "step": 4149 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9571171171171171, + "grad_norm": 25.55521583557129, + "learning_rate": 4.981751714349614e-08, + "loss": 7.041781425476074, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.3103506062179804, + "loss_total_v6": 0.44011134281754494, + "step": 4150 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9573477477477478, + "grad_norm": 27.052997589111328, + "learning_rate": 4.928416427936844e-08, + "loss": 7.22972297668457, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.32194435596466064, + "loss_total_v6": 0.45185769721865654, + "step": 4151 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9575783783783783, + "grad_norm": 25.112892150878906, + "learning_rate": 4.87536677339423e-08, + "loss": 6.379884719848633, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.2686463613063097, + "loss_total_v6": 0.39874278381466866, + "step": 4152 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.957809009009009, + "grad_norm": 25.27898597717285, + "learning_rate": 4.8226027813289e-08, + "loss": 7.132994651794434, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.3174704909324646, + "loss_total_v6": 0.4458121582865715, + "step": 4153 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9580396396396397, + "grad_norm": 22.95831871032715, + "learning_rate": 4.7701244821830584e-08, + "loss": 7.695679664611816, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3533859960734844, + "loss_total_v6": 0.4809799939393997, + "step": 4154 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9582702702702702, + "grad_norm": 26.420602798461914, + "learning_rate": 4.7179319062340415e-08, + "loss": 6.716416358947754, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.2928381487727165, + "loss_total_v6": 0.4197760038077831, + "step": 4155 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9585009009009009, + "grad_norm": 24.945941925048828, + "learning_rate": 4.666025083594483e-08, + "loss": 6.423697471618652, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.2746652830392122, + "loss_total_v6": 0.40148109570145607, + "step": 4156 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9587315315315316, + "grad_norm": 22.481231689453125, + "learning_rate": 4.614404044211873e-08, + "loss": 7.360815525054932, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.33058011904358864, + "loss_total_v6": 0.46005095914006233, + "step": 4157 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9589621621621621, + "grad_norm": 26.947675704956055, + "learning_rate": 4.563068817869165e-08, + "loss": 6.7150349617004395, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.29122595116496086, + "loss_total_v6": 0.41968969255685806, + "step": 4158 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9591927927927928, + "grad_norm": 27.223657608032227, + "learning_rate": 4.512019434184223e-08, + "loss": 7.157187461853027, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.3197301998734474, + "loss_total_v6": 0.4473241940140724, + "step": 4159 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9594234234234235, + "grad_norm": 24.818309783935547, + "learning_rate": 4.461255922609986e-08, + "loss": 7.102978229522705, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.31571653857827187, + "loss_total_v6": 0.44393614679574966, + "step": 4160 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.959654054054054, + "grad_norm": 24.017255783081055, + "learning_rate": 4.4107783124344715e-08, + "loss": 6.708203315734863, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.29139406234025955, + "loss_total_v6": 0.41926272585988045, + "step": 4161 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9598846846846847, + "grad_norm": 33.71091842651367, + "learning_rate": 4.360586632780828e-08, + "loss": 7.377427101135254, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.33209140226244926, + "loss_total_v6": 0.46108920499682426, + "step": 4162 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9601153153153154, + "grad_norm": 24.829212188720703, + "learning_rate": 4.310680912607168e-08, + "loss": 6.546133041381836, + "loss_alignment": 0.249755859375, + "loss_alignment_w": 0.1248779296875, + "loss_sft": 0.28437742963433266, + "loss_total_v6": 0.40913328900933266, + "step": 4163 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9603459459459459, + "grad_norm": 25.29764175415039, + "learning_rate": 4.261061180706627e-08, + "loss": 7.191005706787109, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.32024169713258743, + "loss_total_v6": 0.44943786785006523, + "step": 4164 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9605765765765766, + "grad_norm": 24.57176971435547, + "learning_rate": 4.211727465707361e-08, + "loss": 7.302857398986816, + "loss_alignment": 0.2633056640625, + "loss_alignment_w": 0.13165283203125, + "loss_sft": 0.3246079199016094, + "loss_total_v6": 0.4564285986125469, + "step": 4165 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9608072072072072, + "grad_norm": 23.745145797729492, + "learning_rate": 4.162679796072433e-08, + "loss": 7.069943428039551, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.31459788978099823, + "loss_total_v6": 0.4418714642524719, + "step": 4166 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9610378378378378, + "grad_norm": 24.197410583496094, + "learning_rate": 4.1139182000999314e-08, + "loss": 6.6746039390563965, + "loss_alignment": 0.2647705078125, + "loss_alignment_w": 0.13238525390625, + "loss_sft": 0.28462490253150463, + "loss_total_v6": 0.4171627424657345, + "step": 4167 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9612684684684685, + "grad_norm": 30.42010498046875, + "learning_rate": 4.0654427059229065e-08, + "loss": 6.736574649810791, + "loss_alignment": 0.2467041015625, + "loss_alignment_w": 0.12335205078125, + "loss_sft": 0.29742446914315224, + "loss_total_v6": 0.42103591933846474, + "step": 4168 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9614990990990991, + "grad_norm": 23.128442764282227, + "learning_rate": 4.017253341509375e-08, + "loss": 7.244480133056641, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.322424178943038, + "loss_total_v6": 0.45278000459074974, + "step": 4169 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9617297297297297, + "grad_norm": 24.51725959777832, + "learning_rate": 3.969350134662209e-08, + "loss": 7.21097993850708, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3221156857907772, + "loss_total_v6": 0.4506862573325634, + "step": 4170 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9619603603603604, + "grad_norm": 26.495691299438477, + "learning_rate": 3.921733113019077e-08, + "loss": 7.227716445922852, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.3229633495211601, + "loss_total_v6": 0.4517322666943073, + "step": 4171 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.962190990990991, + "grad_norm": 25.472631454467773, + "learning_rate": 3.874402304052838e-08, + "loss": 7.72910737991333, + "loss_alignment": 0.251708984375, + "loss_alignment_w": 0.1258544921875, + "loss_sft": 0.35713842138648033, + "loss_total_v6": 0.48306921124458313, + "step": 4172 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9624216216216216, + "grad_norm": 24.296403884887695, + "learning_rate": 3.827357735070869e-08, + "loss": 6.918018817901611, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.3019440323114395, + "loss_total_v6": 0.4323761835694313, + "step": 4173 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9626522522522523, + "grad_norm": 23.6592960357666, + "learning_rate": 3.780599433215626e-08, + "loss": 6.979062557220459, + "loss_alignment": 0.2510986328125, + "loss_alignment_w": 0.12554931640625, + "loss_sft": 0.3104895167052746, + "loss_total_v6": 0.4361914210021496, + "step": 4174 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9628828828828829, + "grad_norm": 27.719680786132812, + "learning_rate": 3.7341274254644175e-08, + "loss": 7.050926208496094, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.31073901802301407, + "loss_total_v6": 0.44068286940455437, + "step": 4175 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9631135135135135, + "grad_norm": 24.2208251953125, + "learning_rate": 3.687941738629186e-08, + "loss": 6.735686779022217, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.29134174808859825, + "loss_total_v6": 0.42098042741417885, + "step": 4176 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9633441441441442, + "grad_norm": 22.71321678161621, + "learning_rate": 3.6420423993567824e-08, + "loss": 7.329331398010254, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3297720439732075, + "loss_total_v6": 0.45808320492506027, + "step": 4177 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9635747747747748, + "grad_norm": 25.905729293823242, + "learning_rate": 3.5964294341289676e-08, + "loss": 6.599207401275635, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.2852837238460779, + "loss_total_v6": 0.4124504588544369, + "step": 4178 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9638054054054054, + "grad_norm": 23.84506607055664, + "learning_rate": 3.5511028692620795e-08, + "loss": 6.831300258636475, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.2991181220859289, + "loss_total_v6": 0.42695626243948936, + "step": 4179 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.964036036036036, + "grad_norm": 24.439931869506836, + "learning_rate": 3.5060627309074225e-08, + "loss": 7.326108455657959, + "loss_alignment": 0.2607421875, + "loss_alignment_w": 0.13037109375, + "loss_sft": 0.32783110067248344, + "loss_total_v6": 0.45788176730275154, + "step": 4180 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9642666666666667, + "grad_norm": 26.98463249206543, + "learning_rate": 3.4613090450507644e-08, + "loss": 7.260766983032227, + "loss_alignment": 0.263427734375, + "loss_alignment_w": 0.1317138671875, + "loss_sft": 0.32209933549165726, + "loss_total_v6": 0.45379794761538506, + "step": 4181 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9644972972972973, + "grad_norm": 24.920724868774414, + "learning_rate": 3.416841837512952e-08, + "loss": 6.804501533508301, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.2964361347258091, + "loss_total_v6": 0.4252813421189785, + "step": 4182 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9647279279279279, + "grad_norm": 24.344106674194336, + "learning_rate": 3.37266113394924e-08, + "loss": 6.988824844360352, + "loss_alignment": 0.24755859375, + "loss_alignment_w": 0.123779296875, + "loss_sft": 0.3131138142198324, + "loss_total_v6": 0.43680156022310257, + "step": 4183 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9649585585585586, + "grad_norm": 26.135089874267578, + "learning_rate": 3.3287669598497384e-08, + "loss": 6.685599327087402, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.28918785229325294, + "loss_total_v6": 0.41784996166825294, + "step": 4184 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9651891891891892, + "grad_norm": 23.549598693847656, + "learning_rate": 3.2851593405393014e-08, + "loss": 6.967766761779785, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.3092800099402666, + "loss_total_v6": 0.43548543751239777, + "step": 4185 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9654198198198198, + "grad_norm": 26.379056930541992, + "learning_rate": 3.241838301177247e-08, + "loss": 7.243228435516357, + "loss_alignment": 0.259765625, + "loss_alignment_w": 0.1298828125, + "loss_sft": 0.32300204783678055, + "loss_total_v6": 0.45270177721977234, + "step": 4186 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9656504504504505, + "grad_norm": 24.42043113708496, + "learning_rate": 3.198803866757749e-08, + "loss": 7.10371208190918, + "loss_alignment": 0.251220703125, + "loss_alignment_w": 0.1256103515625, + "loss_sft": 0.31858527660369873, + "loss_total_v6": 0.4439820200204849, + "step": 4187 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9658810810810811, + "grad_norm": 24.379261016845703, + "learning_rate": 3.156056062109503e-08, + "loss": 6.943970680236816, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3062821067869663, + "loss_total_v6": 0.4339981637895107, + "step": 4188 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9661117117117117, + "grad_norm": 26.97103500366211, + "learning_rate": 3.113594911896001e-08, + "loss": 7.081948280334473, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.31177764385938644, + "loss_total_v6": 0.44262176379561424, + "step": 4189 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9663423423423424, + "grad_norm": 24.08152961730957, + "learning_rate": 3.071420440615147e-08, + "loss": 7.046802997589111, + "loss_alignment": 0.25732421875, + "loss_alignment_w": 0.128662109375, + "loss_sft": 0.31174782663583755, + "loss_total_v6": 0.44042519479990005, + "step": 4190 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.966572972972973, + "grad_norm": 25.959949493408203, + "learning_rate": 3.029532672599478e-08, + "loss": 6.567042827606201, + "loss_alignment": 0.257080078125, + "loss_alignment_w": 0.1285400390625, + "loss_sft": 0.2819306757301092, + "loss_total_v6": 0.41044018790125847, + "step": 4191 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9668036036036036, + "grad_norm": 24.37021255493164, + "learning_rate": 2.987931632016272e-08, + "loss": 6.772321701049805, + "loss_alignment": 0.2491455078125, + "loss_alignment_w": 0.12457275390625, + "loss_sft": 0.29856003634631634, + "loss_total_v6": 0.4232701249420643, + "step": 4192 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9670342342342342, + "grad_norm": 29.107816696166992, + "learning_rate": 2.9466173428672197e-08, + "loss": 7.2709197998046875, + "loss_alignment": 0.263671875, + "loss_alignment_w": 0.1318359375, + "loss_sft": 0.32247447595000267, + "loss_total_v6": 0.45443249121308327, + "step": 4193 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9672648648648648, + "grad_norm": 22.211294174194336, + "learning_rate": 2.9055898289885865e-08, + "loss": 7.370214462280273, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3323272541165352, + "loss_total_v6": 0.4606384113430977, + "step": 4194 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9674954954954955, + "grad_norm": 27.20415687561035, + "learning_rate": 2.8648491140513267e-08, + "loss": 6.977643966674805, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3073795773088932, + "loss_total_v6": 0.4361027292907238, + "step": 4195 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9677261261261261, + "grad_norm": 24.6934871673584, + "learning_rate": 2.824395221560805e-08, + "loss": 7.628379821777344, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.348874531686306, + "loss_total_v6": 0.4767737165093422, + "step": 4196 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9679567567567567, + "grad_norm": 27.370161056518555, + "learning_rate": 2.7842281748567957e-08, + "loss": 7.003366470336914, + "loss_alignment": 0.258056640625, + "loss_alignment_w": 0.1290283203125, + "loss_sft": 0.3089262191206217, + "loss_total_v6": 0.4377104192972183, + "step": 4197 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9681873873873874, + "grad_norm": 24.726171493530273, + "learning_rate": 2.744347997113872e-08, + "loss": 7.128683567047119, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3166212309151888, + "loss_total_v6": 0.44554273039102554, + "step": 4198 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.968418018018018, + "grad_norm": 27.49131965637207, + "learning_rate": 2.7047547113407957e-08, + "loss": 6.946994304656982, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.30618116445839405, + "loss_total_v6": 0.4341871403157711, + "step": 4199 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9686486486486486, + "grad_norm": 24.731868743896484, + "learning_rate": 2.665448340381016e-08, + "loss": 6.819608688354492, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.29722775891423225, + "loss_total_v6": 0.42622555419802666, + "step": 4200 + }, + { + "epoch": 0.9686486486486486, + "eval_loss": 0.5620110630989075, + "eval_loss_alignment": 0.2303773366152968, + "eval_loss_alignment_w": 0.1151886683076484, + "eval_loss_sft": 0.44682237651352197, + "eval_loss_total_v6": 0.5620110452379266, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.4489, + "eval_samples_per_second": 19.293, + "eval_steps_per_second": 2.421, + "step": 4200 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9688792792792793, + "grad_norm": 22.617555618286133, + "learning_rate": 2.6264289069123373e-08, + "loss": 7.015050888061523, + "loss_alignment": 0.2545166015625, + "loss_alignment_w": 0.12725830078125, + "loss_sft": 0.3111213371157646, + "loss_total_v6": 0.4384406879544258, + "step": 4201 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9691099099099099, + "grad_norm": 23.881725311279297, + "learning_rate": 2.5876964334470288e-08, + "loss": 6.804864406585693, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.29720648750662804, + "loss_total_v6": 0.42530402168631554, + "step": 4202 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9693405405405405, + "grad_norm": 26.73775863647461, + "learning_rate": 2.549250942331882e-08, + "loss": 6.791476249694824, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.29577460139989853, + "loss_total_v6": 0.4244672395288944, + "step": 4203 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9695711711711712, + "grad_norm": 22.045082092285156, + "learning_rate": 2.511092455747932e-08, + "loss": 6.7076735496521, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.29113204404711723, + "loss_total_v6": 0.4192295968532562, + "step": 4204 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9698018018018018, + "grad_norm": 24.03424072265625, + "learning_rate": 2.473220995710901e-08, + "loss": 6.6471266746521, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.2848759666085243, + "loss_total_v6": 0.4154454283416271, + "step": 4205 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9700324324324324, + "grad_norm": 23.84770393371582, + "learning_rate": 2.4356365840706442e-08, + "loss": 6.6476006507873535, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.2877437248826027, + "loss_total_v6": 0.4154750369489193, + "step": 4206 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9702630630630631, + "grad_norm": 24.73983383178711, + "learning_rate": 2.398339242511538e-08, + "loss": 6.8709635734558105, + "loss_alignment": 0.2606201171875, + "loss_alignment_w": 0.13031005859375, + "loss_sft": 0.29917094483971596, + "loss_total_v6": 0.42943523079156876, + "step": 4207 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9704936936936936, + "grad_norm": 23.17449378967285, + "learning_rate": 2.3613289925523143e-08, + "loss": 6.564330577850342, + "loss_alignment": 0.2576904296875, + "loss_alignment_w": 0.12884521484375, + "loss_sft": 0.28137965872883797, + "loss_total_v6": 0.41027066111564636, + "step": 4208 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9707243243243243, + "grad_norm": 26.110116958618164, + "learning_rate": 2.3246058555460027e-08, + "loss": 6.642007827758789, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.2878061681985855, + "loss_total_v6": 0.4151255041360855, + "step": 4209 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.970954954954955, + "grad_norm": 23.36362075805664, + "learning_rate": 2.2881698526801556e-08, + "loss": 6.998106002807617, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.30958928912878036, + "loss_total_v6": 0.43738164007663727, + "step": 4210 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9711855855855855, + "grad_norm": 22.56301498413086, + "learning_rate": 2.252021004976457e-08, + "loss": 6.766903877258301, + "loss_alignment": 0.2481689453125, + "loss_alignment_w": 0.12408447265625, + "loss_sft": 0.2989080511033535, + "loss_total_v6": 0.4229314848780632, + "step": 4211 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9714162162162162, + "grad_norm": 22.560049057006836, + "learning_rate": 2.2161593332910014e-08, + "loss": 6.616125106811523, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.28689037077128887, + "loss_total_v6": 0.4135078079998493, + "step": 4212 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9716468468468469, + "grad_norm": 22.800352096557617, + "learning_rate": 2.1805848583142385e-08, + "loss": 6.843725204467773, + "loss_alignment": 0.2529296875, + "loss_alignment_w": 0.12646484375, + "loss_sft": 0.3013290073722601, + "loss_total_v6": 0.42773282527923584, + "step": 4213 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9718774774774774, + "grad_norm": 24.939313888549805, + "learning_rate": 2.145297600570917e-08, + "loss": 6.572965621948242, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.2817362677305937, + "loss_total_v6": 0.41081035137176514, + "step": 4214 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9721081081081081, + "grad_norm": 24.187742233276367, + "learning_rate": 2.1102975804200287e-08, + "loss": 6.831198215484619, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.2986692376434803, + "loss_total_v6": 0.426949892193079, + "step": 4215 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9723387387387388, + "grad_norm": 22.10063362121582, + "learning_rate": 2.0755848180547543e-08, + "loss": 7.33284854888916, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.32881693355739117, + "loss_total_v6": 0.4583030417561531, + "step": 4216 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9725693693693693, + "grad_norm": 24.02370834350586, + "learning_rate": 2.0411593335027958e-08, + "loss": 6.808805465698242, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.2975290883332491, + "loss_total_v6": 0.42555033415555954, + "step": 4217 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9728, + "grad_norm": 26.200599670410156, + "learning_rate": 2.007021146625876e-08, + "loss": 7.097574710845947, + "loss_alignment": 0.2611083984375, + "loss_alignment_w": 0.13055419921875, + "loss_sft": 0.3130442164838314, + "loss_total_v6": 0.443598423153162, + "step": 4218 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9730306306306307, + "grad_norm": 27.38396453857422, + "learning_rate": 1.9731702771200733e-08, + "loss": 6.696093559265137, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.29188843071460724, + "loss_total_v6": 0.41850585862994194, + "step": 4219 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9732612612612612, + "grad_norm": 25.88935661315918, + "learning_rate": 1.9396067445155987e-08, + "loss": 7.010588645935059, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.31052203848958015, + "loss_total_v6": 0.43816180527210236, + "step": 4220 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9734918918918919, + "grad_norm": 24.189680099487305, + "learning_rate": 1.9063305681770726e-08, + "loss": 7.806121826171875, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.3579387404024601, + "loss_total_v6": 0.4878826029598713, + "step": 4221 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9737225225225226, + "grad_norm": 26.5618953704834, + "learning_rate": 1.8733417673031383e-08, + "loss": 7.255074501037598, + "loss_alignment": 0.2537841796875, + "loss_alignment_w": 0.12689208984375, + "loss_sft": 0.32641275972127914, + "loss_total_v6": 0.45344216749072075, + "step": 4222 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9739531531531531, + "grad_norm": 28.84576416015625, + "learning_rate": 1.840640360926682e-08, + "loss": 6.76536750793457, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.29469214379787445, + "loss_total_v6": 0.42283545807003975, + "step": 4223 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9741837837837838, + "grad_norm": 22.692325592041016, + "learning_rate": 1.8082263679148338e-08, + "loss": 7.068594932556152, + "loss_alignment": 0.2530517578125, + "loss_alignment_w": 0.12652587890625, + "loss_sft": 0.31481882371008396, + "loss_total_v6": 0.4417871981859207, + "step": 4224 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9744144144144145, + "grad_norm": 23.4298095703125, + "learning_rate": 1.7760998069688562e-08, + "loss": 7.043455123901367, + "loss_alignment": 0.2550048828125, + "loss_alignment_w": 0.12750244140625, + "loss_sft": 0.31286607682704926, + "loss_total_v6": 0.44021591916680336, + "step": 4225 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.974645045045045, + "grad_norm": 24.669017791748047, + "learning_rate": 1.7442606966242005e-08, + "loss": 6.77078914642334, + "loss_alignment": 0.2476806640625, + "loss_alignment_w": 0.12384033203125, + "loss_sft": 0.299196669831872, + "loss_total_v6": 0.42317432165145874, + "step": 4226 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9748756756756757, + "grad_norm": 24.41450309753418, + "learning_rate": 1.7127090552505054e-08, + "loss": 7.295831680297852, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3273274041712284, + "loss_total_v6": 0.4559894986450672, + "step": 4227 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9751063063063063, + "grad_norm": 26.945585250854492, + "learning_rate": 1.681444901051432e-08, + "loss": 6.628599166870117, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.2864493113011122, + "loss_total_v6": 0.414287444204092, + "step": 4228 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9753369369369369, + "grad_norm": 23.49100685119629, + "learning_rate": 1.650468252064885e-08, + "loss": 6.87598991394043, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.30166711285710335, + "loss_total_v6": 0.42974938452243805, + "step": 4229 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9755675675675676, + "grad_norm": 25.216344833374023, + "learning_rate": 1.619779126162846e-08, + "loss": 7.102308750152588, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.3178872223943472, + "loss_total_v6": 0.44389430433511734, + "step": 4230 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9757981981981982, + "grad_norm": 25.454795837402344, + "learning_rate": 1.5893775410514843e-08, + "loss": 6.641031742095947, + "loss_alignment": 0.2596435546875, + "loss_alignment_w": 0.12982177734375, + "loss_sft": 0.2851511426270008, + "loss_total_v6": 0.4150644764304161, + "step": 4231 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9760288288288288, + "grad_norm": 28.00189208984375, + "learning_rate": 1.5592635142709367e-08, + "loss": 7.139985084533691, + "loss_alignment": 0.2603759765625, + "loss_alignment_w": 0.13018798828125, + "loss_sft": 0.31610687635838985, + "loss_total_v6": 0.4462490864098072, + "step": 4232 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9762594594594595, + "grad_norm": 21.868417739868164, + "learning_rate": 1.5294370631955268e-08, + "loss": 6.976943016052246, + "loss_alignment": 0.2523193359375, + "loss_alignment_w": 0.12615966796875, + "loss_sft": 0.30991455540060997, + "loss_total_v6": 0.4360589496791363, + "step": 4233 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9764900900900901, + "grad_norm": 25.777259826660156, + "learning_rate": 1.499898205033712e-08, + "loss": 7.424284934997559, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3332499638199806, + "loss_total_v6": 0.4640177823603153, + "step": 4234 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9767207207207207, + "grad_norm": 28.73773765563965, + "learning_rate": 1.4706469568279147e-08, + "loss": 6.901495933532715, + "loss_alignment": 0.2581787109375, + "loss_alignment_w": 0.12908935546875, + "loss_sft": 0.30245251953601837, + "loss_total_v6": 0.4313435070216656, + "step": 4235 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9769513513513514, + "grad_norm": 25.4860782623291, + "learning_rate": 1.4416833354546355e-08, + "loss": 6.95567512512207, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.3075171932578087, + "loss_total_v6": 0.4347297139465809, + "step": 4236 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.977181981981982, + "grad_norm": 26.0363826751709, + "learning_rate": 1.4130073576244518e-08, + "loss": 7.198760986328125, + "loss_alignment": 0.262939453125, + "loss_alignment_w": 0.1314697265625, + "loss_sft": 0.3183155208826065, + "loss_total_v6": 0.4499225653707981, + "step": 4237 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9774126126126126, + "grad_norm": 25.195005416870117, + "learning_rate": 1.384619039882018e-08, + "loss": 6.838173866271973, + "loss_alignment": 0.2548828125, + "loss_alignment_w": 0.12744140625, + "loss_sft": 0.2999139316380024, + "loss_total_v6": 0.4273858480155468, + "step": 4238 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9776432432432433, + "grad_norm": 26.323749542236328, + "learning_rate": 1.3565183986060104e-08, + "loss": 7.0581464767456055, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.31086985394358635, + "loss_total_v6": 0.44113414362072945, + "step": 4239 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9778738738738739, + "grad_norm": 26.14790153503418, + "learning_rate": 1.3287054500090713e-08, + "loss": 6.896941184997559, + "loss_alignment": 0.2584228515625, + "loss_alignment_w": 0.12921142578125, + "loss_sft": 0.3021068200469017, + "loss_total_v6": 0.431058831512928, + "step": 4240 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9781045045045045, + "grad_norm": 23.654817581176758, + "learning_rate": 1.30118021013792e-08, + "loss": 7.492063522338867, + "loss_alignment": 0.264892578125, + "loss_alignment_w": 0.1324462890625, + "loss_sft": 0.3357466608285904, + "loss_total_v6": 0.4682539813220501, + "step": 4241 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9783351351351351, + "grad_norm": 22.73267364501953, + "learning_rate": 1.2739426948732426e-08, + "loss": 7.0845842361450195, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3151009790599346, + "loss_total_v6": 0.4427865147590637, + "step": 4242 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9785657657657658, + "grad_norm": 20.77918243408203, + "learning_rate": 1.246992919929746e-08, + "loss": 7.065215110778809, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.31134216859936714, + "loss_total_v6": 0.44157594814896584, + "step": 4243 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9787963963963964, + "grad_norm": 24.671236038208008, + "learning_rate": 1.2203309008561592e-08, + "loss": 7.042999267578125, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.3113117255270481, + "loss_total_v6": 0.4401874653995037, + "step": 4244 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.979027027027027, + "grad_norm": 25.38129234313965, + "learning_rate": 1.1939566530351221e-08, + "loss": 6.864953994750977, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.3025490175932646, + "loss_total_v6": 0.42905960977077484, + "step": 4245 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9792576576576577, + "grad_norm": 26.086610794067383, + "learning_rate": 1.1678701916832402e-08, + "loss": 6.702481269836426, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.2915857657790184, + "loss_total_v6": 0.4189051017165184, + "step": 4246 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9794882882882883, + "grad_norm": 25.106178283691406, + "learning_rate": 1.142071531851141e-08, + "loss": 7.229852676391602, + "loss_alignment": 0.2525634765625, + "loss_alignment_w": 0.12628173828125, + "loss_sft": 0.3253398835659027, + "loss_total_v6": 0.4518657736480236, + "step": 4247 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9797189189189189, + "grad_norm": 30.341102600097656, + "learning_rate": 1.1165606884234182e-08, + "loss": 6.622803211212158, + "loss_alignment": 0.26123046875, + "loss_alignment_w": 0.130615234375, + "loss_sft": 0.2836304046213627, + "loss_total_v6": 0.4139252007007599, + "step": 4248 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9799495495495496, + "grad_norm": 21.42714500427246, + "learning_rate": 1.09133767611852e-08, + "loss": 6.507528305053711, + "loss_alignment": 0.25341796875, + "loss_alignment_w": 0.126708984375, + "loss_sft": 0.27990472689270973, + "loss_total_v6": 0.40672053024172783, + "step": 4249 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9801801801801802, + "grad_norm": 25.152103424072266, + "learning_rate": 1.0664025094889174e-08, + "loss": 7.07253360748291, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.3158126249909401, + "loss_total_v6": 0.4420333318412304, + "step": 4250 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9804108108108108, + "grad_norm": 25.719970703125, + "learning_rate": 1.0417552029209354e-08, + "loss": 6.662463188171387, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.28993913158774376, + "loss_total_v6": 0.41640396788716316, + "step": 4251 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9806414414414414, + "grad_norm": 22.33028793334961, + "learning_rate": 1.0173957706348658e-08, + "loss": 6.909627914428711, + "loss_alignment": 0.2568359375, + "loss_alignment_w": 0.12841796875, + "loss_sft": 0.3036016523838043, + "loss_total_v6": 0.4318517707288265, + "step": 4252 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.980872072072072, + "grad_norm": 24.734088897705078, + "learning_rate": 9.933242266848553e-09, + "loss": 7.099040985107422, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.31316637992858887, + "loss_total_v6": 0.44369006529450417, + "step": 4253 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9811027027027027, + "grad_norm": 25.069528579711914, + "learning_rate": 9.695405849590722e-09, + "loss": 7.156245231628418, + "loss_alignment": 0.2559814453125, + "loss_alignment_w": 0.12799072265625, + "loss_sft": 0.31916777789592743, + "loss_total_v6": 0.4472653158009052, + "step": 4254 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9813333333333333, + "grad_norm": 24.600997924804688, + "learning_rate": 9.460448591794846e-09, + "loss": 7.400568962097168, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3323933631181717, + "loss_total_v6": 0.4625355638563633, + "step": 4255 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.981563963963964, + "grad_norm": 27.994827270507812, + "learning_rate": 9.228370629019712e-09, + "loss": 7.284411907196045, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.3272697776556015, + "loss_total_v6": 0.4552757665514946, + "step": 4256 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9817945945945946, + "grad_norm": 24.798419952392578, + "learning_rate": 8.9991720951621e-09, + "loss": 6.830172538757324, + "loss_alignment": 0.252685546875, + "loss_alignment_w": 0.1263427734375, + "loss_sft": 0.30055825784802437, + "loss_total_v6": 0.42688576877117157, + "step": 4257 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9820252252252252, + "grad_norm": 22.826669692993164, + "learning_rate": 8.772853122459568e-09, + "loss": 6.935929775238037, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.30509291775524616, + "loss_total_v6": 0.4334956146776676, + "step": 4258 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9822558558558558, + "grad_norm": 26.240055084228516, + "learning_rate": 8.549413841485443e-09, + "loss": 6.762167930603027, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.2930425964295864, + "loss_total_v6": 0.4226354993879795, + "step": 4259 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9824864864864865, + "grad_norm": 22.906648635864258, + "learning_rate": 8.328854381154938e-09, + "loss": 6.8406572341918945, + "loss_alignment": 0.2523193359375, + "loss_alignment_w": 0.12615966796875, + "loss_sft": 0.3013661354780197, + "loss_total_v6": 0.4275410547852516, + "step": 4260 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9827171171171171, + "grad_norm": 25.113386154174805, + "learning_rate": 8.11117486871793e-09, + "loss": 6.700223922729492, + "loss_alignment": 0.2498779296875, + "loss_alignment_w": 0.12493896484375, + "loss_sft": 0.29387080669403076, + "loss_total_v6": 0.41876400262117386, + "step": 4261 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9829477477477477, + "grad_norm": 24.60114288330078, + "learning_rate": 7.896375429766734e-09, + "loss": 7.090097904205322, + "loss_alignment": 0.2618408203125, + "loss_alignment_w": 0.13092041015625, + "loss_sft": 0.31185976788401604, + "loss_total_v6": 0.44313111901283264, + "step": 4262 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9831783783783784, + "grad_norm": 21.475671768188477, + "learning_rate": 7.684456188228883e-09, + "loss": 7.090466499328613, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.31481246277689934, + "loss_total_v6": 0.44315414130687714, + "step": 4263 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.983409009009009, + "grad_norm": 24.738798141479492, + "learning_rate": 7.475417266371576e-09, + "loss": 6.839784145355225, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.3005486447364092, + "loss_total_v6": 0.42748650908470154, + "step": 4264 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9836396396396396, + "grad_norm": 26.855859756469727, + "learning_rate": 7.269258784799449e-09, + "loss": 6.536825180053711, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.2794927414506674, + "loss_total_v6": 0.40855158492922783, + "step": 4265 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9838702702702703, + "grad_norm": 23.074478149414062, + "learning_rate": 7.065980862456801e-09, + "loss": 6.79789924621582, + "loss_alignment": 0.2528076171875, + "loss_alignment_w": 0.12640380859375, + "loss_sft": 0.2986327577382326, + "loss_total_v6": 0.42486870661377907, + "step": 4266 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9841009009009009, + "grad_norm": 26.12693214416504, + "learning_rate": 6.86558361662426e-09, + "loss": 7.264452934265137, + "loss_alignment": 0.255615234375, + "loss_alignment_w": 0.1278076171875, + "loss_sft": 0.32644957304000854, + "loss_total_v6": 0.45402831211686134, + "step": 4267 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9843315315315315, + "grad_norm": 22.30116081237793, + "learning_rate": 6.668067162921566e-09, + "loss": 7.151524543762207, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.31865914538502693, + "loss_total_v6": 0.44697029888629913, + "step": 4268 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9845621621621622, + "grad_norm": 24.966148376464844, + "learning_rate": 6.473431615305337e-09, + "loss": 7.02075719833374, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.309936860576272, + "loss_total_v6": 0.43879732117056847, + "step": 4269 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9847927927927927, + "grad_norm": 25.018957138061523, + "learning_rate": 6.281677086071303e-09, + "loss": 6.713330268859863, + "loss_alignment": 0.25439453125, + "loss_alignment_w": 0.127197265625, + "loss_sft": 0.29247741028666496, + "loss_total_v6": 0.41958314552903175, + "step": 4270 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9850234234234234, + "grad_norm": 23.978824615478516, + "learning_rate": 6.092803685852078e-09, + "loss": 6.992071151733398, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.3071216307580471, + "loss_total_v6": 0.4370044395327568, + "step": 4271 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9852540540540541, + "grad_norm": 23.715662002563477, + "learning_rate": 5.906811523618272e-09, + "loss": 7.313023567199707, + "loss_alignment": 0.2508544921875, + "loss_alignment_w": 0.12542724609375, + "loss_sft": 0.3314841352403164, + "loss_total_v6": 0.4570639617741108, + "step": 4272 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9854846846846846, + "grad_norm": 26.707231521606445, + "learning_rate": 5.723700706678492e-09, + "loss": 7.167505264282227, + "loss_alignment": 0.26318359375, + "loss_alignment_w": 0.131591796875, + "loss_sft": 0.31633149087429047, + "loss_total_v6": 0.44796906039118767, + "step": 4273 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9857153153153153, + "grad_norm": 23.4776668548584, + "learning_rate": 5.5434713406782305e-09, + "loss": 6.781920433044434, + "loss_alignment": 0.250244140625, + "loss_alignment_w": 0.1251220703125, + "loss_sft": 0.29885475896298885, + "loss_total_v6": 0.4238700307905674, + "step": 4274 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.985945945945946, + "grad_norm": 22.90653419494629, + "learning_rate": 5.366123529600975e-09, + "loss": 6.982731819152832, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3080485463142395, + "loss_total_v6": 0.4364207424223423, + "step": 4275 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9861765765765765, + "grad_norm": 26.902515411376953, + "learning_rate": 5.191657375767656e-09, + "loss": 7.536625862121582, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.3405306786298752, + "loss_total_v6": 0.4710390977561474, + "step": 4276 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9864072072072072, + "grad_norm": 27.248916625976562, + "learning_rate": 5.0200729798371985e-09, + "loss": 6.8183274269104, + "loss_alignment": 0.2587890625, + "loss_alignment_w": 0.12939453125, + "loss_sft": 0.2969950679689646, + "loss_total_v6": 0.4261454641819, + "step": 4277 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9866378378378379, + "grad_norm": 24.456912994384766, + "learning_rate": 4.851370440804304e-09, + "loss": 7.511081218719482, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.33960552513599396, + "loss_total_v6": 0.46944256499409676, + "step": 4278 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9868684684684684, + "grad_norm": 22.344572067260742, + "learning_rate": 4.6855498560033354e-09, + "loss": 6.9390788078308105, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.30635782331228256, + "loss_total_v6": 0.43369242548942566, + "step": 4279 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9870990990990991, + "grad_norm": 26.180925369262695, + "learning_rate": 4.522611321103321e-09, + "loss": 7.230800151824951, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.32349177449941635, + "loss_total_v6": 0.45192500203847885, + "step": 4280 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9873297297297298, + "grad_norm": 24.443166732788086, + "learning_rate": 4.362554930112395e-09, + "loss": 6.9706549644470215, + "loss_alignment": 0.25244140625, + "loss_alignment_w": 0.126220703125, + "loss_sft": 0.30936895310878754, + "loss_total_v6": 0.43566593900322914, + "step": 4281 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9875603603603603, + "grad_norm": 25.96173095703125, + "learning_rate": 4.2053807753750235e-09, + "loss": 7.249316215515137, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.32445067167282104, + "loss_total_v6": 0.45308225229382515, + "step": 4282 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.987790990990991, + "grad_norm": 23.053787231445312, + "learning_rate": 4.051088947573112e-09, + "loss": 7.282243728637695, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.32562362030148506, + "loss_total_v6": 0.45514022558927536, + "step": 4283 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9880216216216217, + "grad_norm": 22.958703994750977, + "learning_rate": 3.899679535725454e-09, + "loss": 6.978462219238281, + "loss_alignment": 0.2625732421875, + "loss_alignment_w": 0.13128662109375, + "loss_sft": 0.3048520013689995, + "loss_total_v6": 0.4361538738012314, + "step": 4284 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9882522522522522, + "grad_norm": 23.772994995117188, + "learning_rate": 3.751152627187727e-09, + "loss": 6.487811088562012, + "loss_alignment": 0.256591796875, + "loss_alignment_w": 0.1282958984375, + "loss_sft": 0.27717703953385353, + "loss_total_v6": 0.40548818930983543, + "step": 4285 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9884828828828829, + "grad_norm": 25.85938835144043, + "learning_rate": 3.605508307651939e-09, + "loss": 7.152587890625, + "loss_alignment": 0.2601318359375, + "loss_alignment_w": 0.13006591796875, + "loss_sft": 0.3167114444077015, + "loss_total_v6": 0.4470367543399334, + "step": 4286 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9887135135135136, + "grad_norm": 23.269319534301758, + "learning_rate": 3.4627466611492077e-09, + "loss": 7.191344261169434, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.3223685622215271, + "loss_total_v6": 0.4494590237736702, + "step": 4287 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9889441441441441, + "grad_norm": 24.94928741455078, + "learning_rate": 3.3228677700442024e-09, + "loss": 6.834315299987793, + "loss_alignment": 0.2586669921875, + "loss_alignment_w": 0.12933349609375, + "loss_sft": 0.2981011550873518, + "loss_total_v6": 0.42714472487568855, + "step": 4288 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9891747747747748, + "grad_norm": 23.752628326416016, + "learning_rate": 3.1858717150412554e-09, + "loss": 6.714667320251465, + "loss_alignment": 0.2513427734375, + "loss_alignment_w": 0.12567138671875, + "loss_sft": 0.2938274461776018, + "loss_total_v6": 0.41966668888926506, + "step": 4289 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9894054054054054, + "grad_norm": 24.334447860717773, + "learning_rate": 3.0517585751804744e-09, + "loss": 7.429020881652832, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.3345072604715824, + "loss_total_v6": 0.4643137827515602, + "step": 4290 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.989636036036036, + "grad_norm": 25.042654037475586, + "learning_rate": 2.920528427837743e-09, + "loss": 7.193742752075195, + "loss_alignment": 0.2615966796875, + "loss_alignment_w": 0.13079833984375, + "loss_sft": 0.3189631812274456, + "loss_total_v6": 0.4496089369058609, + "step": 4291 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9898666666666667, + "grad_norm": 25.8033504486084, + "learning_rate": 2.792181348726941e-09, + "loss": 7.312526702880859, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.32824873365461826, + "loss_total_v6": 0.4570329114794731, + "step": 4292 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9900972972972973, + "grad_norm": 23.3116512298584, + "learning_rate": 2.6667174118977235e-09, + "loss": 7.186532497406006, + "loss_alignment": 0.253662109375, + "loss_alignment_w": 0.1268310546875, + "loss_sft": 0.32222040370106697, + "loss_total_v6": 0.44915827736258507, + "step": 4293 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9903279279279279, + "grad_norm": 24.38680648803711, + "learning_rate": 2.5441366897366317e-09, + "loss": 6.886909484863281, + "loss_alignment": 0.2506103515625, + "loss_alignment_w": 0.12530517578125, + "loss_sft": 0.30529452860355377, + "loss_total_v6": 0.4304318465292454, + "step": 4294 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9905585585585586, + "grad_norm": 23.71177864074707, + "learning_rate": 2.4244392529670924e-09, + "loss": 6.894552230834961, + "loss_alignment": 0.2532958984375, + "loss_alignment_w": 0.12664794921875, + "loss_sft": 0.3041700068861246, + "loss_total_v6": 0.43090950325131416, + "step": 4295 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9907891891891892, + "grad_norm": 27.466724395751953, + "learning_rate": 2.3076251706477536e-09, + "loss": 7.371181011199951, + "loss_alignment": 0.260009765625, + "loss_alignment_w": 0.1300048828125, + "loss_sft": 0.33055660128593445, + "loss_total_v6": 0.46069881692528725, + "step": 4296 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9910198198198198, + "grad_norm": 24.312179565429688, + "learning_rate": 2.1936945101752594e-09, + "loss": 6.854947090148926, + "loss_alignment": 0.255126953125, + "loss_alignment_w": 0.1275634765625, + "loss_sft": 0.3007943909615278, + "loss_total_v6": 0.42843418195843697, + "step": 4297 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9912504504504505, + "grad_norm": 22.558420181274414, + "learning_rate": 2.082647337282029e-09, + "loss": 7.254402160644531, + "loss_alignment": 0.259521484375, + "loss_alignment_w": 0.1297607421875, + "loss_sft": 0.3237919956445694, + "loss_total_v6": 0.4534001350402832, + "step": 4298 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9914810810810811, + "grad_norm": 26.19986915588379, + "learning_rate": 1.9744837160362572e-09, + "loss": 7.0023674964904785, + "loss_alignment": 0.2623291015625, + "loss_alignment_w": 0.13116455078125, + "loss_sft": 0.3064376227557659, + "loss_total_v6": 0.4376479648053646, + "step": 4299 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9917117117117117, + "grad_norm": 31.388328552246094, + "learning_rate": 1.869203708843581e-09, + "loss": 7.290214538574219, + "loss_alignment": 0.25830078125, + "loss_alignment_w": 0.129150390625, + "loss_sft": 0.32671689987182617, + "loss_total_v6": 0.4556383974850178, + "step": 4300 + }, + { + "epoch": 0.9917117117117117, + "eval_loss": 0.5619313716888428, + "eval_loss_alignment": 0.2303840253995434, + "eval_loss_alignment_w": 0.1151920126997717, + "eval_loss_sft": 0.44673937028401517, + "eval_loss_total_v6": 0.5619313827626509, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.3934, + "eval_samples_per_second": 19.305, + "eval_steps_per_second": 2.423, + "step": 4300 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9919423423423424, + "grad_norm": 27.614648818969727, + "learning_rate": 1.7668073764448569e-09, + "loss": 7.128278732299805, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.3159703090786934, + "loss_total_v6": 0.445517435669899, + "step": 4301 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.992172972972973, + "grad_norm": 27.106773376464844, + "learning_rate": 1.6672947779178273e-09, + "loss": 7.179193496704102, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.3198848944157362, + "loss_total_v6": 0.44869958236813545, + "step": 4302 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9924036036036036, + "grad_norm": 26.912668228149414, + "learning_rate": 1.5706659706771211e-09, + "loss": 6.863919258117676, + "loss_alignment": 0.263916015625, + "loss_alignment_w": 0.1319580078125, + "loss_sft": 0.2971132528036833, + "loss_total_v6": 0.42899495363235474, + "step": 4303 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9926342342342342, + "grad_norm": 23.496976852416992, + "learning_rate": 1.476921010471477e-09, + "loss": 6.843374252319336, + "loss_alignment": 0.259033203125, + "loss_alignment_w": 0.1295166015625, + "loss_sft": 0.2980417124927044, + "loss_total_v6": 0.4277109131217003, + "step": 4304 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9928648648648649, + "grad_norm": 24.943523406982422, + "learning_rate": 1.3860599513881855e-09, + "loss": 7.130498886108398, + "loss_alignment": 0.251708984375, + "loss_alignment_w": 0.1258544921875, + "loss_sft": 0.31951176561415195, + "loss_total_v6": 0.4456561729311943, + "step": 4305 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9930954954954955, + "grad_norm": 24.621299743652344, + "learning_rate": 1.2980828458492023e-09, + "loss": 7.087548732757568, + "loss_alignment": 0.2554931640625, + "loss_alignment_w": 0.12774658203125, + "loss_sft": 0.3152099475264549, + "loss_total_v6": 0.4429717995226383, + "step": 4306 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9933261261261261, + "grad_norm": 24.575368881225586, + "learning_rate": 1.2129897446133687e-09, + "loss": 6.707370758056641, + "loss_alignment": 0.2515869140625, + "loss_alignment_w": 0.12579345703125, + "loss_sft": 0.29355454444885254, + "loss_total_v6": 0.41921066492795944, + "step": 4307 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9935567567567568, + "grad_norm": 25.05780601501465, + "learning_rate": 1.130780696774192e-09, + "loss": 6.999386787414551, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3070753179490566, + "loss_total_v6": 0.43746168166399, + "step": 4308 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9937873873873874, + "grad_norm": 25.366310119628906, + "learning_rate": 1.051455749763175e-09, + "loss": 7.198148727416992, + "loss_alignment": 0.260986328125, + "loss_alignment_w": 0.1304931640625, + "loss_sft": 0.3194216303527355, + "loss_total_v6": 0.449884295463562, + "step": 4309 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.994018018018018, + "grad_norm": 24.903478622436523, + "learning_rate": 9.750149493464866e-10, + "loss": 7.2584991455078125, + "loss_alignment": 0.2542724609375, + "loss_alignment_w": 0.12713623046875, + "loss_sft": 0.3267641067504883, + "loss_total_v6": 0.4536561965942383, + "step": 4310 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9942486486486487, + "grad_norm": 24.553123474121094, + "learning_rate": 9.014583396266263e-10, + "loss": 6.940491199493408, + "loss_alignment": 0.2572021484375, + "loss_alignment_w": 0.12860107421875, + "loss_sft": 0.3053474836051464, + "loss_total_v6": 0.4337806962430477, + "step": 4311 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9944792792792793, + "grad_norm": 24.01221466064453, + "learning_rate": 8.307859630429793e-10, + "loss": 7.1369524002075195, + "loss_alignment": 0.2593994140625, + "loss_alignment_w": 0.12969970703125, + "loss_sft": 0.3162377364933491, + "loss_total_v6": 0.44605952501296997, + "step": 4312 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9947099099099099, + "grad_norm": 26.09644889831543, + "learning_rate": 7.629978603684862e-10, + "loss": 6.279335021972656, + "loss_alignment": 0.249755859375, + "loss_alignment_w": 0.1248779296875, + "loss_sft": 0.26754998974502087, + "loss_total_v6": 0.3924584351480007, + "step": 4313 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9949405405405405, + "grad_norm": 23.76227378845215, + "learning_rate": 6.980940707146388e-10, + "loss": 6.958683013916016, + "loss_alignment": 0.25634765625, + "loss_alignment_w": 0.128173828125, + "loss_sft": 0.30669810250401497, + "loss_total_v6": 0.43491770699620247, + "step": 4314 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9951711711711712, + "grad_norm": 25.23593521118164, + "learning_rate": 6.360746315275946e-10, + "loss": 7.180349349975586, + "loss_alignment": 0.2567138671875, + "loss_alignment_w": 0.12835693359375, + "loss_sft": 0.3202775754034519, + "loss_total_v6": 0.4487718418240547, + "step": 4315 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9954018018018018, + "grad_norm": 24.55379867553711, + "learning_rate": 5.769395785898413e-10, + "loss": 7.196072578430176, + "loss_alignment": 0.2579345703125, + "loss_alignment_w": 0.12896728515625, + "loss_sft": 0.3209856189787388, + "loss_total_v6": 0.4497545324265957, + "step": 4316 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9956324324324324, + "grad_norm": 25.805665969848633, + "learning_rate": 5.206889460185327e-10, + "loss": 7.332320213317871, + "loss_alignment": 0.254638671875, + "loss_alignment_w": 0.1273193359375, + "loss_sft": 0.33107272535562515, + "loss_total_v6": 0.45826998353004456, + "step": 4317 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.995863063063063, + "grad_norm": 24.870128631591797, + "learning_rate": 4.673227662677083e-10, + "loss": 6.964435577392578, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.30516551062464714, + "loss_total_v6": 0.43527721613645554, + "step": 4318 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9960936936936937, + "grad_norm": 26.30704689025879, + "learning_rate": 4.16841070127183e-10, + "loss": 6.978847026824951, + "loss_alignment": 0.2598876953125, + "loss_alignment_w": 0.12994384765625, + "loss_sft": 0.3065087553113699, + "loss_total_v6": 0.43617793545126915, + "step": 4319 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9963243243243243, + "grad_norm": 26.29311752319336, + "learning_rate": 3.6924388672254784e-10, + "loss": 6.953705310821533, + "loss_alignment": 0.2591552734375, + "loss_alignment_w": 0.12957763671875, + "loss_sft": 0.30505945533514023, + "loss_total_v6": 0.43460657075047493, + "step": 4320 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9965549549549549, + "grad_norm": 23.309558868408203, + "learning_rate": 3.245312435151693e-10, + "loss": 7.170981407165527, + "loss_alignment": 0.25537109375, + "loss_alignment_w": 0.127685546875, + "loss_sft": 0.32079072296619415, + "loss_total_v6": 0.44818635284900665, + "step": 4321 + }, + { + "action_cond_aligned": 0.9765625, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9967855855855856, + "grad_norm": 23.427677154541016, + "learning_rate": 2.827031663016344e-10, + "loss": 7.084221839904785, + "loss_alignment": 0.25244140625, + "loss_alignment_w": 0.126220703125, + "loss_sft": 0.3166804797947407, + "loss_total_v6": 0.44276386871933937, + "step": 4322 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9970162162162162, + "grad_norm": 24.457151412963867, + "learning_rate": 2.437596792148611e-10, + "loss": 7.123974800109863, + "loss_alignment": 0.2613525390625, + "loss_alignment_w": 0.13067626953125, + "loss_sft": 0.31438906490802765, + "loss_total_v6": 0.44524843245744705, + "step": 4323 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9972468468468468, + "grad_norm": 23.494674682617188, + "learning_rate": 2.0770080472298782e-10, + "loss": 6.941632270812988, + "loss_alignment": 0.2574462890625, + "loss_alignment_w": 0.12872314453125, + "loss_sft": 0.30496097542345524, + "loss_total_v6": 0.4338519908487797, + "step": 4324 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9974774774774775, + "grad_norm": 24.854103088378906, + "learning_rate": 1.7452656363103893e-10, + "loss": 6.932000637054443, + "loss_alignment": 0.26025390625, + "loss_alignment_w": 0.130126953125, + "loss_sft": 0.3033977337181568, + "loss_total_v6": 0.4332500286400318, + "step": 4325 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9977081081081081, + "grad_norm": 26.437068939208984, + "learning_rate": 1.4423697507814916e-10, + "loss": 7.0733137130737305, + "loss_alignment": 0.2564697265625, + "loss_alignment_w": 0.12823486328125, + "loss_sft": 0.31386251747608185, + "loss_total_v6": 0.44208211451768875, + "step": 4326 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9979387387387387, + "grad_norm": 28.58077049255371, + "learning_rate": 1.1683205653978402e-10, + "loss": 6.927838325500488, + "loss_alignment": 0.258544921875, + "loss_alignment_w": 0.1292724609375, + "loss_sft": 0.3035038225352764, + "loss_total_v6": 0.4329899065196514, + "step": 4327 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9981693693693694, + "grad_norm": 25.678606033325195, + "learning_rate": 9.231182382773984e-11, + "loss": 6.545172214508057, + "loss_alignment": 0.2481689453125, + "loss_alignment_w": 0.12408447265625, + "loss_sft": 0.28459206223487854, + "loss_total_v6": 0.40907325968146324, + "step": 4328 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9984, + "grad_norm": 25.857147216796875, + "learning_rate": 7.067629108903351e-11, + "loss": 6.931822299957275, + "loss_alignment": 0.2569580078125, + "loss_alignment_w": 0.12847900390625, + "loss_sft": 0.3046225607395172, + "loss_total_v6": 0.43323889747262, + "step": 4329 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9986306306306306, + "grad_norm": 23.509653091430664, + "learning_rate": 5.1925470806457603e-11, + "loss": 6.947049140930176, + "loss_alignment": 0.2552490234375, + "loss_alignment_w": 0.12762451171875, + "loss_sft": 0.3064439743757248, + "loss_total_v6": 0.4341905601322651, + "step": 4330 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9988612612612613, + "grad_norm": 27.188053131103516, + "learning_rate": 3.6059373797470155e-11, + "loss": 7.120849132537842, + "loss_alignment": 0.2562255859375, + "loss_alignment_w": 0.12811279296875, + "loss_sft": 0.3170013055205345, + "loss_total_v6": 0.4450530633330345, + "step": 4331 + }, + { + "action_cond_aligned": 0.9921875, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9990918918918918, + "grad_norm": 21.385330200195312, + "learning_rate": 2.3078009216415122e-11, + "loss": 7.081700325012207, + "loss_alignment": 0.249755859375, + "loss_alignment_w": 0.1248779296875, + "loss_sft": 0.3178656790405512, + "loss_total_v6": 0.44260628148913383, + "step": 4332 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9993225225225225, + "grad_norm": 24.256654739379883, + "learning_rate": 1.2981384553412136e-11, + "loss": 7.0345916748046875, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.3127393610775471, + "loss_total_v6": 0.4396619684994221, + "step": 4333 + }, + { + "action_cond_aligned": 0.984375, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9995531531531532, + "grad_norm": 23.765607833862305, + "learning_rate": 5.7695056326911855e-12, + "loss": 6.702056407928467, + "loss_alignment": 0.257568359375, + "loss_alignment_w": 0.1287841796875, + "loss_sft": 0.290277449414134, + "loss_total_v6": 0.4188785217702389, + "step": 4334 + }, + { + "action_cond_aligned": 1.0, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 0.9997837837837837, + "grad_norm": 26.764110565185547, + "learning_rate": 1.4423766164783914e-12, + "loss": 6.828737735748291, + "loss_alignment": 0.25927734375, + "loss_alignment_w": 0.129638671875, + "loss_sft": 0.2973862998187542, + "loss_total_v6": 0.4267960973083973, + "step": 4335 + }, + { + "action_cond_aligned": 0.9916666666666667, + "action_cond_missing": 0.0, + "action_cond_skipped": 0.0, + "epoch": 1.0, + "grad_norm": 24.512462615966797, + "learning_rate": 0.0, + "loss": 6.1031999588012695, + "loss_alignment": 0.255859375, + "loss_alignment_w": 0.1279296875, + "loss_sft": 0.27890146176020303, + "loss_total_v6": 0.4068799694379171, + "step": 4336 + }, + { + "epoch": 1.0, + "eval_loss": 0.561885416507721, + "eval_loss_alignment": 0.23037510702054795, + "eval_loss_alignment_w": 0.11518755351027397, + "eval_loss_sft": 0.44669782340560604, + "eval_loss_total_v6": 0.5618853764651028, + "eval_loss_wm_recon": 0.0, + "eval_loss_wm_recon_w": 0.0, + "eval_runtime": 90.6142, + "eval_samples_per_second": 19.257, + "eval_steps_per_second": 2.417, + "step": 4336 + } + ], + "logging_steps": 1, + "max_steps": 4336, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 5.448439683098031e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/training_args.bin b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..a4c0420f0e5ae03c851d8a109f47d09dcafc9b96 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf7c80f0ad8044e3ac70b36c262bfd71094837fa2ed1853cb519e1703dbfcb8b +size 9169 diff --git a/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/zero_to_fp32.py b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694 --- /dev/null +++ b/qwen-cua__runs__opencua_exact_wm_joint_v7_siglip2_delta_pad_pool32x32_phase2_baseline_wmlr__v3-20260618-000502__checkpoint-4336/zero_to_fp32.py @@ -0,0 +1,760 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: +# python zero_to_fp32.py . output_dir/ +# or +# python zero_to_fp32.py . output_dir/ --safe_serialization + +import argparse +import torch +import glob +import math +import os +import re +import gc +import json +import numpy as np +from tqdm import tqdm +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device, weights_only=False) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + total_files = len(files) + state_dicts = [] + for f in tqdm(files, desc='Loading checkpoint shards'): + state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +class GatheredTensor: + """ + A pseudo tensor that collects partitioned weights. + It is more memory efficient when there are multiple groups. + """ + + def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape): + self.flat_groups = flat_groups + self.flat_groups_offset = flat_groups_offset + self.offset = offset + self.partitioned_numel = partitioned_numel + self.shape = shape + self.dtype = self.flat_groups[0][0].dtype + + def contiguous(self): + """ + Merge partitioned weights from flat_groups into a single tensor. + """ + end_idx = self.offset + self.partitioned_numel + world_size = len(self.flat_groups) + pad_flat_param_chunks = [] + + for rank_i in range(world_size): + # for each rank, we need to collect weights from related group/groups + flat_groups_at_rank_i = self.flat_groups[rank_i] + start_group_id = None + end_group_id = None + for group_id in range(len(self.flat_groups_offset)): + if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]: + start_group_id = group_id + if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]: + end_group_id = group_id + break + # collect weights from related group/groups + for group_id in range(start_group_id, end_group_id + 1): + flat_tensor = flat_groups_at_rank_i[group_id] + start_offset = self.offset - self.flat_groups_offset[group_id] + end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id] + pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset]) + + # collect weights from all ranks + pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0) + param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous() + return param + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size + + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]])) + for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'): + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # memory efficient tensor + tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape) + state_dict[name] = tensor + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def to_torch_tensor(state_dict, return_empty_tensor=False): + """ + Convert state_dict of GatheredTensor to torch tensor + """ + torch_state_dict = {} + converted_tensors = {} + for name, tensor in state_dict.items(): + tensor_id = id(tensor) + if tensor_id in converted_tensors: # shared tensors + shared_tensor = torch_state_dict[converted_tensors[tensor_id]] + torch_state_dict[name] = shared_tensor + else: + converted_tensors[tensor_id] = name + if return_empty_tensor: + torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype) + else: + torch_state_dict[name] = tensor.contiguous() + return torch_state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag=None, + exclude_frozen_parameters=False, + lazy_mode=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient. + Convert the pesduo tensor to torch tensor by ``.contiguous()`` + + Returns: + - pytorch ``state_dict`` + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + Note: the above usage may not work if your application doesn't have sufficient free CPU memory. + You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. Or you can load state_dict in lazy mode :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu + for name, lazy_tensor in state_dict.item(): + tensor = lazy_tensor.contiguous() # to cpu + print(name, tensor) + # del tensor to release memory if it no longer in use + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + if lazy_mode: + return state_dict + else: + return to_torch_tensor(state_dict) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, + output_dir, + max_shard_size="5GB", + safe_serialization=False, + tag=None, + exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_dir``: directory to the pytorch fp32 state_dict output files + - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB + - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`). + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + # Dependency pre-check + if safe_serialization: + try: + from safetensors.torch import save_file + except ImportError: + print('If you want to use `safe_serialization`, please `pip install safetensors`') + raise + if max_shard_size is not None: + try: + from huggingface_hub import split_torch_state_dict_into_shards + except ImportError: + print('If you want to use `max_shard_size`, please `pip install huggingface_hub`') + raise + + # Convert zero checkpoint to state_dict + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag, + exclude_frozen_parameters, + lazy_mode=True) + + # Shard the model if it is too big. + weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin" + if max_shard_size is not None: + filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors") + # an memory-efficient approach for sharding + empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True) + state_dict_split = split_torch_state_dict_into_shards(empty_state_dict, + filename_pattern=filename_pattern, + max_shard_size=max_shard_size) + else: + from collections import namedtuple + StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"]) + state_dict_split = StateDictSplit(is_sharded=False, + filename_to_tensors={weights_name: list(state_dict.keys())}) + + # Save the model by shard + os.makedirs(output_dir, exist_ok=True) + filename_to_tensors = state_dict_split.filename_to_tensors.items() + for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"): + shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors} + shard_state_dict = to_torch_tensor(shard_state_dict) + output_path = os.path.join(output_dir, shard_file) + if safe_serialization: + save_file(shard_state_dict, output_path, metadata={"format": "pt"}) + else: + torch.save(shard_state_dict, output_path) + # release the memory of current shard + for tensor_name in list(shard_state_dict.keys()): + del state_dict[tensor_name] + del shard_state_dict[tensor_name] + del shard_state_dict + gc.collect() + + # Save index if sharded + if state_dict_split.is_sharded: + index = { + "metadata": state_dict_split.metadata, + "weight_map": state_dict_split.tensor_to_filename, + } + save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json" + save_index_file = os.path.join(output_dir, save_index_file) + with open(save_index_file, "w", encoding="utf-8") as f: + content = json.dumps(index, indent=2, sort_keys=True) + "\n" + f.write(content) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info("Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info("Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument("output_dir", + type=str, + help="directory to the pytorch fp32 state_dict output files" + "(e.g. path/checkpoint-12-output/)") + parser.add_argument( + "--max_shard_size", + type=str, + default="5GB", + help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size" + "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`" + "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances" + "without CPU OOM issues.") + parser.add_argument( + "--safe_serialization", + default=False, + action='store_true', + help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_dir, + max_shard_size=args.max_shard_size, + safe_serialization=args.safe_serialization, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/args.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/args.json new file mode 100644 index 0000000000000000000000000000000000000000..361deec09ff99a6714ee8d82edf8edf103b5c271 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/args.json @@ -0,0 +1,382 @@ +{ + "output_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_coord/v5-20260622-101413", + "per_device_train_batch_size": 1, + "num_train_epochs": 3.0, + "max_steps": 2000, + "learning_rate": 1e-05, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_steps": 100.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "optim_target_modules": null, + "gradient_accumulation_steps": 32, + "average_tokens_across_devices": true, + "max_grad_norm": 1.0, + "label_smoothing_factor": 0.0, + "bf16": true, + "fp16": false, + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": true, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "use_liger_kernel": false, + "liger_kernel_config": null, + "use_cache": false, + "neftune_noise_alpha": null, + "torch_empty_cache_steps": null, + "auto_find_batch_size": false, + "logging_strategy": "steps", + "logging_steps": 10, + "logging_first_step": true, + "log_on_each_node": true, + "logging_nan_inf_filter": true, + "include_num_input_tokens_seen": false, + "log_level": "passive", + "log_level_replica": "warning", + "disable_tqdm": null, + "report_to": [ + "wandb" + ], + "run_name": "probe2000_idm_code_coord", + "project": "huggingface", + "trackio_space_id": null, + "trackio_bucket_id": null, + "trackio_static_space_id": null, + "eval_strategy": "steps", + "eval_steps": 500.0, + "eval_delay": 0, + "per_device_eval_batch_size": 1, + "prediction_loss_only": false, + "eval_on_start": false, + "eval_do_concat_batches": true, + "eval_use_gather_object": false, + "eval_accumulation_steps": null, + "include_for_metrics": [], + "batch_eval_metrics": false, + "save_only_model": false, + "save_strategy": "steps", + "save_steps": 500.0, + "save_on_each_node": false, + "save_total_limit": 6, + "enable_jit_checkpoint": false, + "push_to_hub": false, + "hub_token": null, + "hub_private_repo": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_always_push": false, + "hub_revision": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "restore_callback_states_from_checkpoint": false, + "full_determinism": false, + "seed": 42, + "data_seed": 42, + "use_cpu": false, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "dataloader_drop_last": false, + "dataloader_num_workers": 8, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 4, + "remove_unused_columns": true, + "label_names": null, + "train_sampling_strategy": "random", + "length_column_name": "length", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "ddp_static_graph": null, + "ddp_backend": null, + "ddp_timeout": 18000000, + "fsdp": [], + "fsdp_config": null, + "deepspeed": { + "bf16": { + "enabled": true + }, + "fp16": { + "enabled": false + }, + "zero_optimization": { + "stage": 2, + "overlap_comm": true, + "contiguous_gradients": true, + "reduce_bucket_size": 500000000.0, + "allgather_bucket_size": 500000000.0, + "reduce_scatter": true, + "round_robin_gradients": true + }, + "gradient_clipping": "auto", + "gradient_accumulation_steps": "auto", + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "steps_per_print": 100, + "wall_clock_breakdown": false, + "_comment_bf16_safety": "DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives." + }, + "debug": null, + "skip_memory_metrics": true, + "do_train": false, + "do_eval": false, + "do_predict": false, + "resume_from_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_coord/v3-20260621-233930/checkpoint-1500", + "warmup_ratio": null, + "logging_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_coord/v5-20260622-101413/runs", + "local_rank": 0, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "group_by_length": false, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "model_type": "qwen3_5", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attention_2", + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_5", + "system": null, + "max_length": 12288, + "truncation_strategy": "delete", + "max_pixels": 16777216, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": false, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l1/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l2/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l3/train_wm.jsonl#50000" + ], + "val_dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/val_wm.jsonl" + ], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 4, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "use_hf": false, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual", + "model.visual.merger" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "rank": 0, + "global_world_size": 4, + "local_world_size": 4, + "model_suffix": "checkpoint-4336", + "model_info": "ModelInfo(model_type='qwen3_5', model_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling=None, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_5', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.5-0.8B', hf_model_id='Qwen/Qwen3.5-0.8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B', hf_model_id='Qwen/Qwen3.5-2B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B', hf_model_id='Qwen/Qwen3.5-4B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B', hf_model_id='Qwen/Qwen3.5-9B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B', hf_model_id='Qwen/Qwen3.5-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B-FP8', hf_model_id='Qwen/Qwen3.5-27B-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-0.8B-Base', hf_model_id='Qwen/Qwen3.5-0.8B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B-Base', hf_model_id='Qwen/Qwen3.5-2B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B-Base', hf_model_id='Qwen/Qwen3.5-4B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B-Base', hf_model_id='Qwen/Qwen3.5-9B-Base', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.6-27B', hf_model_id='Qwen/Qwen3.6-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.6-27B-FP8', hf_model_id='Qwen/Qwen3.6-27B-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[])], loader=, template=None, model_arch=MultiModelKeys(arch_name='qwen2_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3_5ForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=5.0.0.dev', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "template_meta": "QwenTemplateMeta(template_type='qwen3_5', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='qwen3_5', is_thinking=True, thinking_prefix='\\n', non_thinking_prefix='\\n\\n\\n\\n', history_thinking_prefix='')", + "_val_dataset_exists": true, + "hub": "", + "evaluation_strategy": "steps", + "training_args": "Seq2SeqTrainingArguments(output_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_coord/v5-20260622-101413', per_device_train_batch_size=1, num_train_epochs=3.0, max_steps=2000, learning_rate=1e-05, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_steps=100.0, optim=, optim_args=None, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=32, average_tokens_across_devices=None, max_grad_norm=1.0, label_smoothing_factor=0.0, bf16=True, fp16=False, bf16_full_eval=False, fp16_full_eval=False, tf32=True, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, use_liger_kernel=False, liger_kernel_config=None, use_cache=False, neftune_noise_alpha=None, torch_empty_cache_steps=None, auto_find_batch_size=False, logging_strategy=, logging_steps=10, logging_first_step=True, log_on_each_node=True, logging_nan_inf_filter=True, include_num_input_tokens_seen=None, log_level='passive', log_level_replica='warning', disable_tqdm=False, report_to=['wandb'], run_name='probe2000_idm_code_coord', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=, eval_steps=500, eval_delay=0, per_device_eval_batch_size=1, prediction_loss_only=False, eval_on_start=False, eval_do_concat_batches=True, eval_use_gather_object=False, eval_accumulation_steps=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=, save_steps=500, save_on_each_node=False, save_total_limit=6, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=, hub_always_push=False, hub_revision=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, restore_callback_states_from_checkpoint=False, full_determinism=False, seed=42, data_seed=42, use_cpu=False, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, dataloader_drop_last=False, dataloader_num_workers=8, dataloader_pin_memory=True, dataloader_persistent_workers=True, dataloader_prefetch_factor=4, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, deepspeed={'bf16': {'enabled': True}, 'fp16': {'enabled': False}, 'zero_optimization': {'stage': 2, 'overlap_comm': True, 'contiguous_gradients': True, 'reduce_bucket_size': 500000000.0, 'allgather_bucket_size': 500000000.0, 'reduce_scatter': True, 'round_robin_gradients': True}, 'gradient_clipping': 'auto', 'gradient_accumulation_steps': 'auto', 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'steps_per_print': 100, 'wall_clock_breakdown': False, '_comment_bf16_safety': 'DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives.'}, debug=[], skip_memory_metrics=True, do_train=False, do_eval=True, do_predict=False, resume_from_checkpoint='/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_coord/v3-20260621-233930/checkpoint-1500', warmup_ratio=None, logging_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_coord/v5-20260622-101413/runs', local_rank=0, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, group_by_length=False, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/chat_template.jinja b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..a585dec894e63da457d9440ec6aa7caa16d20860 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/chat_template.jinja @@ -0,0 +1,154 @@ +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- macro render_content(content, do_vision_count, is_system_content=false) %} + {%- if content is string %} + {{- content }} + {%- elif content is iterable and content is not mapping %} + {%- for item in content %} + {%- if 'image' in item or 'image_url' in item or item.type == 'image' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain images.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set image_count.value = image_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Picture ' ~ image_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|image_pad|><|vision_end|>' }} + {%- elif 'video' in item or item.type == 'video' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain videos.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set video_count.value = video_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Video ' ~ video_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|video_pad|><|vision_end|>' }} + {%- elif 'text' in item %} + {{- item.text }} + {%- else %} + {{- raise_exception('Unexpected item type in content.') }} + {%- endif %} + {%- endfor %} + {%- elif content is none or content is undefined %} + {{- '' }} + {%- else %} + {{- raise_exception('Unexpected content type.') }} + {%- endif %} +{%- endmacro %} +{%- if not messages %} + {{- raise_exception('No messages provided.') }} +{%- endif %} +{%- if tools and tools is iterable and tools is not mapping %} + {{- '<|im_start|>system\n' }} + {{- "# Tools\n\nYou have access to the following functions:\n\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n" }} + {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {%- if content %} + {{- '\n\n' + content }} + {%- endif %} + {%- endif %} + {{- '<|im_end|>\n' }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" %} + {%- set content = render_content(message.content, false)|trim %} + {%- if not(content.startswith('') and content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if ns.multi_step_tool %} + {{- raise_exception('No user query found in messages.') }} +{%- endif %} +{%- for message in messages %} + {%- set content = render_content(message.content, true)|trim %} + {%- if message.role == "system" %} + {%- if not loop.first %} + {{- raise_exception('System message must be at the beginning.') }} + {%- endif %} + {%- elif message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = '' %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- set reasoning_content = reasoning_content|trim %} + {%- if loop.index0 > ns.last_query_index %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {%- if loop.first %} + {%- if content|trim %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n\n' }} + {%- endif %} + {%- else %} + {{- '\n\n\n' }} + {%- endif %} + {%- if tool_call.arguments is defined %} + {%- for args_name, args_value in tool_call.arguments|items %} + {{- '\n' }} + {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %} + {{- args_value }} + {{- '\n\n' }} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.previtem and loop.previtem.role != "tool" %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if not loop.last and loop.nextitem.role != "tool" %} + {{- '<|im_end|>\n' }} + {%- elif loop.last %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- else %} + {{- raise_exception('Unexpected message role.') }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n' }} + {%- endif %} +{%- endif %} \ No newline at end of file diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/config.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b1ade33543aaa3d299835283e0e20968262a8a0e --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/config.json @@ -0,0 +1,113 @@ +{ + "architectures": [ + "Qwen3_5ForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 248046, + "hidden_size": 2560, + "image_token_id": 248056, + "model_type": "qwen3_5", + "pad_token_id": 248044, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "attn_output_gate": true, + "bos_token_id": null, + "dtype": "bfloat16", + "eos_token_id": 248044, + "full_attention_interval": 4, + "head_dim": 256, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9216, + "layer_types": [ + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention" + ], + "linear_conv_kernel_dim": 4, + "linear_key_head_dim": 128, + "linear_num_key_heads": 16, + "linear_num_value_heads": 32, + "linear_value_head_dim": 128, + "mamba_ssm_dtype": "float32", + "max_position_embeddings": 262144, + "mlp_only_layers": [], + "model_type": "qwen3_5_text", + "mtp_num_hidden_layers": 1, + "mtp_use_dedicated_embeddings": false, + "num_attention_heads": 16, + "num_hidden_layers": 32, + "num_key_value_heads": 4, + "pad_token_id": 248044, + "partial_rotary_factor": 0.25, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "mrope_interleaved": true, + "mrope_section": [ + 11, + 11, + 10 + ], + "partial_rotary_factor": 0.25, + "rope_theta": 10000000, + "rope_type": "default" + }, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 248320 + }, + "tie_word_embeddings": true, + "transformers_version": "5.8.1", + "use_cache": false, + "video_token_id": 248057, + "vision_config": { + "deepstack_visual_indexes": [], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_5_vision", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 248054, + "vision_start_token_id": 248053 +} diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/generation_config.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1f1ab970fed81cb788c2945bed33833d247b9091 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "eos_token_id": [ + 248044, + 248046 + ], + "transformers_version": "5.8.1", + "use_cache": true +} diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..b1e1fc189f46c1e0703d68ab1d613aa964c91228 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7f3e95b6d0f2c14a5f7f3ce1ccbecea7cd6b6de7ef18521f30afeb0bd7614cc6 +size 12675600417 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..683c7c3ad395d3809aa5552ceb4a251d6b4c6f85 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a93a0ee8b508bc117b1ecef4a7dea2b75ffa2da79d7120905dda5eb3010273f9 +size 12675605601 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..3f4366e33a14bbe93a6078f868b95489be0d2401 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e2d1d8a2af5a0e0f98da783afb0a20716127666dce5b313f9562202274055a3 +size 12675605729 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..fa259ac271389b2a99ce3e50498be4756e2047e9 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f08a56ad7018578ccfffd05f72263ca03dae94da31b93f45b8f48b3d2feb2251 +size 12675604577 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/mp_rank_00_model_states.pt b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/mp_rank_00_model_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..ed49e1290aec2cd3dfc1e17b648bc19ec838bf1a --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/global_step2000/mp_rank_00_model_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b6ab3946db8d04e50573656eca59c24f43004ab98fad0379267dd6ff6906a1f +size 9784824803 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/latest b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/latest new file mode 100644 index 0000000000000000000000000000000000000000..2a79fdc19587e6bc9de060e90633f3a151b04516 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/latest @@ -0,0 +1 @@ +global_step2000 \ No newline at end of file diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00001-of-00003.safetensors b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e592e62c820bd9f0f4cec8eeb1eae0c5117dc332 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:236f0c46ec33a82c926dd8631c0f92e14f7f40204fc8892ff2fe327d829bab12 +size 4997083614 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00002-of-00003.safetensors b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..49710bf0044a844b37260dd4a86cff3d4e75fa57 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:776d7d67e03e06e1ebeeb7f323016ad6603620fe444200b08c71f7556db810d5 +size 4993498424 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00003-of-00003.safetensors b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ab8023a740649342f1bc7e224320898de73054c --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9132529e66cac3079a828b70ff720888e735b62dcc8a839f7c42730ef896d42c +size 398327248 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model.safetensors.index.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..a138254f39ab95cffdb02f5f8564913a9c1cf4cb --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/model.safetensors.index.json @@ -0,0 +1,748 @@ +{ + "metadata": { + "total_parameters": 4558710159, + "total_size": 10388818718 + }, + "weight_map": { + "_idm_head.anl_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.anl_head.weight": "model-00001-of-00003.safetensors", + "_idm_head.coord_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.coord_head.weight": "model-00001-of-00003.safetensors", + "_idm_head.ln_next.bias": "model-00001-of-00003.safetensors", + "_idm_head.ln_next.weight": "model-00001-of-00003.safetensors", + "_idm_head.ln_state.bias": "model-00001-of-00003.safetensors", + "_idm_head.ln_state.weight": "model-00001-of-00003.safetensors", + "_idm_head.text_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.text_head.weight": "model-00001-of-00003.safetensors", + "_idm_head.trunk.0.bias": "model-00001-of-00003.safetensors", + "_idm_head.trunk.0.weight": "model-00001-of-00003.safetensors", + "_idm_head.trunk.2.bias": "model-00001-of-00003.safetensors", + "_idm_head.trunk.2.weight": "model-00001-of-00003.safetensors", + "_idm_head.type_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.type_head.weight": "model-00001-of-00003.safetensors", + "lm_head.weight": "model-00001-of-00003.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00003.safetensors" + } +} diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/preprocessor_config.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/processor_config.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..33818c7f9e991ad735fd240209f4fa73e6c28c50 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/processor_config.json @@ -0,0 +1,60 @@ +{ + "image_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_0.pth b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..b396fc0c423b86fe0303871876cafff4c6e2fe8e --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e1151ec5bf6bee4180571c0e8613fb542c4045c69203c75d518064bbc8a5215e +size 15429 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_1.pth b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..0686ebca7e90e83ef59441d8b7c9b15525d11a31 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:87f09a6bd97da9e86bd526c0ff6ffed0ccae6d70fd4c0891e6188dcb63b6038b +size 15429 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_2.pth b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..91180de4ebf0ddfac167c6c33ef2c66feb1691df --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:43a28053c6bfe54748f017e8c55ab99bfe8a7d7878c87e7a73a755c11741bb5b +size 15429 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_3.pth b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..9b678ecc192324dc9a54734a0a55be23b4b7ee4d --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ac61fe0f290ea42f93b0fa110f05d8d497644ec52d3387dc749d96ce5f882b4 +size 15429 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/scheduler.pt b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4d8bfa2f2055831824c38d31d554572c47542880 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:36f4b99926451a01a926e5b6c040343926d2653b170ae2a34a2bb9514da0cf5f +size 1465 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/tokenizer.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523 +size 19989325 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/tokenizer_config.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1d134cd298be1e3be25db393d93a1cefe80e3214 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/tokenizer_config.json @@ -0,0 +1,33 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "image_token": "<|image_pad|>", + "is_local": true, + "local_files_only": false, + "model_max_length": 262144, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +} diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/trainer_state.json b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..57b7241fc6c74a2504f5ae47268623e98f8243be --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/trainer_state.json @@ -0,0 +1,2884 @@ +{ + "best_global_step": 2000, + "best_metric": 0.78171819, + "best_model_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_coord/v5-20260622-101413/checkpoint-2000", + "epoch": 0.46126126126126127, + "eval_steps": 500, + "global_step": 2000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00023063063063063062, + "grad_norm": 1590.688720703125, + "learning_rate": 1.0000000000000001e-07, + "loss": 49.45393371582031, + "loss_idm": 3.06640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0621337890625, + "loss_idm_text": 0.0, + "loss_idm_type": 3.003211200237274, + "loss_sft": 0.013178251218050718, + "step": 1, + "token_acc": 0.855064946292334 + }, + { + "epoch": 0.0023063063063063064, + "grad_norm": 1611.734375, + "learning_rate": 1.0000000000000002e-06, + "loss": 47.76775105794271, + "loss_idm": 3.451388888888889, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.05978732638888889, + "loss_idm_text": 0.0, + "loss_idm_type": 2.8977474106682672, + "loss_sft": 0.013165522780683305, + "step": 10, + "token_acc": 0.8557971074689107 + }, + { + "epoch": 0.004612612612612613, + "grad_norm": 339.4031982421875, + "learning_rate": 2.0000000000000003e-06, + "loss": 24.062301635742188, + "loss_idm": 1.3859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0595703125, + "loss_idm_text": 0.0, + "loss_idm_type": 1.4169265031814575, + "loss_sft": 0.013038619607686996, + "step": 20, + "token_acc": 0.8575614504505172 + }, + { + "epoch": 0.006918918918918919, + "grad_norm": 117.9345474243164, + "learning_rate": 3e-06, + "loss": 17.185517883300783, + "loss_idm": 0.934375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0529296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.9935595035552979, + "loss_sft": 0.013021596148610115, + "step": 30, + "token_acc": 0.8576696365011294 + }, + { + "epoch": 0.009225225225225226, + "grad_norm": 174.67822265625, + "learning_rate": 4.000000000000001e-06, + "loss": 15.97247314453125, + "loss_idm": 0.9, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.049169921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.921882438659668, + "loss_sft": 0.01299305036664009, + "step": 40, + "token_acc": 0.8573861458475359 + }, + { + "epoch": 0.011531531531531532, + "grad_norm": 383.7381591796875, + "learning_rate": 5e-06, + "loss": 12.401361846923828, + "loss_idm": 0.69296875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.044677734375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.7016407489776612, + "loss_sft": 0.013680103048682214, + "step": 50, + "token_acc": 0.8517720027290206 + }, + { + "epoch": 0.013837837837837838, + "grad_norm": 276.0512390136719, + "learning_rate": 6e-06, + "loss": 10.131932067871094, + "loss_idm": 0.559375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.04560546875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5592228412628174, + "loss_sft": 0.01410299502313137, + "step": 60, + "token_acc": 0.8488453901987737 + }, + { + "epoch": 0.016144144144144144, + "grad_norm": 236.4015655517578, + "learning_rate": 7e-06, + "loss": 10.888622283935547, + "loss_idm": 0.61328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.03974609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.6118354797363281, + "loss_sft": 0.013782905787229538, + "step": 70, + "token_acc": 0.8515422419612366 + }, + { + "epoch": 0.01845045045045045, + "grad_norm": 173.59402465820312, + "learning_rate": 8.000000000000001e-06, + "loss": 10.349563598632812, + "loss_idm": 0.5734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.041650390625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5762068748474121, + "loss_sft": 0.013817435875535012, + "step": 80, + "token_acc": 0.8507013906275447 + }, + { + "epoch": 0.020756756756756756, + "grad_norm": 231.35693359375, + "learning_rate": 9e-06, + "loss": 10.654405212402343, + "loss_idm": 0.59375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0390625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5972403526306153, + "loss_sft": 0.01390336975455284, + "step": 90, + "token_acc": 0.8490709364931573 + }, + { + "epoch": 0.023063063063063063, + "grad_norm": 179.43270874023438, + "learning_rate": 1e-05, + "loss": 10.548146057128907, + "loss_idm": 0.584375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.03662109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5928622126579285, + "loss_sft": 0.014141988009214401, + "step": 100, + "token_acc": 0.847036073395476 + }, + { + "epoch": 0.025369369369369368, + "grad_norm": 221.4450225830078, + "learning_rate": 9.999316524962347e-06, + "loss": 10.619392395019531, + "loss_idm": 0.5953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.033154296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.6010093808174133, + "loss_sft": 0.014010535180568695, + "step": 110, + "token_acc": 0.8487004546091859 + }, + { + "epoch": 0.027675675675675675, + "grad_norm": 186.53411865234375, + "learning_rate": 9.99726628670463e-06, + "loss": 10.216886138916015, + "loss_idm": 0.57109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.03154296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5778079628944397, + "loss_sft": 0.013904009386897087, + "step": 120, + "token_acc": 0.8502426663442629 + }, + { + "epoch": 0.029981981981981983, + "grad_norm": 157.05239868164062, + "learning_rate": 9.993849845741525e-06, + "loss": 10.855589294433594, + "loss_idm": 0.6046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.03525390625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.6137941002845764, + "loss_sft": 0.014121369644999503, + "step": 130, + "token_acc": 0.8481210495695818 + }, + { + "epoch": 0.03228828828828829, + "grad_norm": 166.25057983398438, + "learning_rate": 9.989068136093873e-06, + "loss": 10.426358032226563, + "loss_idm": 0.58203125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.030859375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5915833473205566, + "loss_sft": 0.01406811736524105, + "step": 140, + "token_acc": 0.8483285794652639 + }, + { + "epoch": 0.034594594594594595, + "grad_norm": 140.43893432617188, + "learning_rate": 9.98292246503335e-06, + "loss": 10.939217376708985, + "loss_idm": 0.60859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.03046875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.6234304666519165, + "loss_sft": 0.014257545396685601, + "step": 150, + "token_acc": 0.8462056666356372 + }, + { + "epoch": 0.0369009009009009, + "grad_norm": 101.10719299316406, + "learning_rate": 9.975414512725058e-06, + "loss": 9.8798828125, + "loss_idm": 0.54765625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.029736328125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5589980244636535, + "loss_sft": 0.013876666501164436, + "step": 160, + "token_acc": 0.8505486664698356 + }, + { + "epoch": 0.03920720720720721, + "grad_norm": 144.91943359375, + "learning_rate": 9.966546331768192e-06, + "loss": 10.420352172851562, + "loss_idm": 0.5953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.029296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5927331566810607, + "loss_sft": 0.014148441702127456, + "step": 170, + "token_acc": 0.8475072962442912 + }, + { + "epoch": 0.04151351351351351, + "grad_norm": 168.03384399414062, + "learning_rate": 9.956320346634877e-06, + "loss": 10.239119720458984, + "loss_idm": 0.56328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.028076171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5830882906913757, + "loss_sft": 0.013758457824587822, + "step": 180, + "token_acc": 0.851148950239709 + }, + { + "epoch": 0.04381981981981982, + "grad_norm": 145.99766540527344, + "learning_rate": 9.944739353007344e-06, + "loss": 10.057091522216798, + "loss_idm": 0.5609375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.026904296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5724347591400146, + "loss_sft": 0.014053185656666755, + "step": 190, + "token_acc": 0.8491978583015695 + }, + { + "epoch": 0.04612612612612613, + "grad_norm": 121.27855682373047, + "learning_rate": 9.931806517013612e-06, + "loss": 10.908540344238281, + "loss_idm": 0.61328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.026611328125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.6256417632102966, + "loss_sft": 0.014259455353021621, + "step": 200, + "token_acc": 0.8467008716078382 + }, + { + "epoch": 0.048432432432432435, + "grad_norm": 147.652099609375, + "learning_rate": 9.917525374361913e-06, + "loss": 10.310504150390624, + "loss_idm": 0.5828125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02548828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5899111151695251, + "loss_sft": 0.014259019494056701, + "step": 210, + "token_acc": 0.8468536389223965 + }, + { + "epoch": 0.050738738738738735, + "grad_norm": 133.29347229003906, + "learning_rate": 9.901899829374048e-06, + "loss": 9.91939468383789, + "loss_idm": 0.55078125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.028125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5629432678222657, + "loss_sft": 0.013959363475441933, + "step": 220, + "token_acc": 0.8485922361662006 + }, + { + "epoch": 0.05304504504504504, + "grad_norm": 105.41419219970703, + "learning_rate": 9.884934153917998e-06, + "loss": 10.20447235107422, + "loss_idm": 0.57265625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0267578125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.582277512550354, + "loss_sft": 0.014081236347556115, + "step": 230, + "token_acc": 0.8486716679030506 + }, + { + "epoch": 0.05535135135135135, + "grad_norm": 108.32715606689453, + "learning_rate": 9.86663298624003e-06, + "loss": 10.037399291992188, + "loss_idm": 0.56328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.025537109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5717163324356079, + "loss_sft": 0.014507130160927773, + "step": 240, + "token_acc": 0.8445775773754048 + }, + { + "epoch": 0.05765765765765766, + "grad_norm": 101.92463684082031, + "learning_rate": 9.847001329696653e-06, + "loss": 10.207322692871093, + "loss_idm": 0.5765625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0251953125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5826281547546387, + "loss_sft": 0.014590685814619064, + "step": 250, + "token_acc": 0.84426266631776 + }, + { + "epoch": 0.059963963963963966, + "grad_norm": 120.43878173828125, + "learning_rate": 9.826044551386743e-06, + "loss": 10.321310424804688, + "loss_idm": 0.5796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02509765625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5899416208267212, + "loss_sft": 0.014607197791337966, + "step": 260, + "token_acc": 0.8437843269576525 + }, + { + "epoch": 0.06227027027027027, + "grad_norm": 112.54505920410156, + "learning_rate": 9.803768380684242e-06, + "loss": 9.775898742675782, + "loss_idm": 0.55390625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02724609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5538631200790405, + "loss_sft": 0.014404882118105889, + "step": 270, + "token_acc": 0.8459256892908046 + }, + { + "epoch": 0.06457657657657657, + "grad_norm": 94.44606018066406, + "learning_rate": 9.780178907671788e-06, + "loss": 10.339688873291015, + "loss_idm": 0.5859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.026123046875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5898990511894227, + "loss_sft": 0.014457888901233673, + "step": 280, + "token_acc": 0.8451656208816066 + }, + { + "epoch": 0.06688288288288288, + "grad_norm": 93.52350616455078, + "learning_rate": 9.755282581475769e-06, + "loss": 9.293934631347657, + "loss_idm": 0.51640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02529296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5259280681610108, + "loss_sft": 0.014395680651068688, + "step": 290, + "token_acc": 0.8458327979361675 + }, + { + "epoch": 0.06918918918918919, + "grad_norm": 113.6584701538086, + "learning_rate": 9.729086208503174e-06, + "loss": 9.748943328857422, + "loss_idm": 0.5515625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0262939453125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5521375179290772, + "loss_sft": 0.014811449870467185, + "step": 300, + "token_acc": 0.8420287020511943 + }, + { + "epoch": 0.0714954954954955, + "grad_norm": 81.37059783935547, + "learning_rate": 9.701596950580807e-06, + "loss": 9.784170532226563, + "loss_idm": 0.54453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0253662109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5561857938766479, + "loss_sft": 0.014250487089157104, + "step": 310, + "token_acc": 0.847669727830882 + }, + { + "epoch": 0.0738018018018018, + "grad_norm": 108.87104797363281, + "learning_rate": 9.672822322997305e-06, + "loss": 9.601348876953125, + "loss_idm": 0.5375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.023486328125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5469521760940552, + "loss_sft": 0.014529699832201004, + "step": 320, + "token_acc": 0.8448489136666951 + }, + { + "epoch": 0.07610810810810811, + "grad_norm": 130.01808166503906, + "learning_rate": 9.642770192448537e-06, + "loss": 9.793380737304688, + "loss_idm": 0.54375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.023828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5585392713546753, + "loss_sft": 0.014446238800883294, + "step": 330, + "token_acc": 0.8454624950201325 + }, + { + "epoch": 0.07841441441441442, + "grad_norm": 102.37490844726562, + "learning_rate": 9.611448774886925e-06, + "loss": 10.019007873535156, + "loss_idm": 0.5578125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0240966796875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5720710635185242, + "loss_sft": 0.01458285003900528, + "step": 340, + "token_acc": 0.8438792841749271 + }, + { + "epoch": 0.08072072072072072, + "grad_norm": 102.40097045898438, + "learning_rate": 9.578866633275289e-06, + "loss": 9.723292541503906, + "loss_idm": 0.54921875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02216796875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.555128300189972, + "loss_sft": 0.01469133198261261, + "step": 350, + "token_acc": 0.8427996218283734 + }, + { + "epoch": 0.08302702702702702, + "grad_norm": 114.85263061523438, + "learning_rate": 9.545032675245814e-06, + "loss": 9.351869201660156, + "loss_idm": 0.52578125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02275390625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5321333289146424, + "loss_sft": 0.014326457306742668, + "step": 360, + "token_acc": 0.8459759580797804 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 111.28559112548828, + "learning_rate": 9.509956150664796e-06, + "loss": 9.204309844970703, + "loss_idm": 0.51796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0224609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5225722551345825, + "loss_sft": 0.014612309634685516, + "step": 370, + "token_acc": 0.8429501243235337 + }, + { + "epoch": 0.08763963963963964, + "grad_norm": 159.0714111328125, + "learning_rate": 9.473646649103819e-06, + "loss": 10.069181823730469, + "loss_idm": 0.55625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0218017578125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.577532160282135, + "loss_sft": 0.014597192034125328, + "step": 380, + "token_acc": 0.8435810512784968 + }, + { + "epoch": 0.08994594594594595, + "grad_norm": 109.76734161376953, + "learning_rate": 9.43611409721806e-06, + "loss": 9.046263885498046, + "loss_idm": 0.51171875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.024365234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5109918117523193, + "loss_sft": 0.01441824659705162, + "step": 390, + "token_acc": 0.8451461865571329 + }, + { + "epoch": 0.09225225225225225, + "grad_norm": 114.03449249267578, + "learning_rate": 9.397368756032445e-06, + "loss": 8.885872650146485, + "loss_idm": 0.49765625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0222412109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5030822515487671, + "loss_sft": 0.014403243735432624, + "step": 400, + "token_acc": 0.8454346932611032 + }, + { + "epoch": 0.09455855855855856, + "grad_norm": 95.63912200927734, + "learning_rate": 9.357421218136387e-06, + "loss": 9.656077575683593, + "loss_idm": 0.53828125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02294921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5517061471939086, + "loss_sft": 0.013985250145196915, + "step": 410, + "token_acc": 0.8492683217596042 + }, + { + "epoch": 0.09686486486486487, + "grad_norm": 104.48960876464844, + "learning_rate": 9.31628240478787e-06, + "loss": 9.242665100097657, + "loss_idm": 0.51328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0219482421875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.525335168838501, + "loss_sft": 0.014777269214391708, + "step": 420, + "token_acc": 0.8418984633339817 + }, + { + "epoch": 0.09917117117117118, + "grad_norm": 147.1559600830078, + "learning_rate": 9.273963562927695e-06, + "loss": 9.189979553222656, + "loss_idm": 0.5125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02255859375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5218879580497742, + "loss_sft": 0.014679400622844696, + "step": 430, + "token_acc": 0.8437356123388582 + }, + { + "epoch": 0.10147747747747747, + "grad_norm": 93.56449127197266, + "learning_rate": 9.230476262104678e-06, + "loss": 9.361734008789062, + "loss_idm": 0.52421875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0218505859375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5325615644454956, + "loss_sft": 0.014835886657238007, + "step": 440, + "token_acc": 0.8411223489145692 + }, + { + "epoch": 0.10378378378378378, + "grad_norm": 82.03609466552734, + "learning_rate": 9.185832391312644e-06, + "loss": 8.240397644042968, + "loss_idm": 0.45546875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0206298828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.46435696482658384, + "loss_sft": 0.014625262469053268, + "step": 450, + "token_acc": 0.8430563318063227 + }, + { + "epoch": 0.10609009009009009, + "grad_norm": 101.49946594238281, + "learning_rate": 9.140044155740102e-06, + "loss": 8.768144226074218, + "loss_idm": 0.48203125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0253173828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.49250184297561644, + "loss_sft": 0.014648660272359847, + "step": 460, + "token_acc": 0.8423310494803166 + }, + { + "epoch": 0.1083963963963964, + "grad_norm": 119.373291015625, + "learning_rate": 9.093124073433464e-06, + "loss": 9.084942626953126, + "loss_idm": 0.509375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02119140625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5161856293678284, + "loss_sft": 0.01479191929101944, + "step": 470, + "token_acc": 0.842016730205665 + }, + { + "epoch": 0.1107027027027027, + "grad_norm": 110.53947448730469, + "learning_rate": 9.045084971874738e-06, + "loss": 8.959126281738282, + "loss_idm": 0.5015625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0206787109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5086698532104492, + "loss_sft": 0.014888135716319084, + "step": 480, + "token_acc": 0.8410448617813955 + }, + { + "epoch": 0.11300900900900901, + "grad_norm": 88.150146484375, + "learning_rate": 8.995939984474624e-06, + "loss": 9.11611557006836, + "loss_idm": 0.51484375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.021240234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.517701530456543, + "loss_sft": 0.014868151023983955, + "step": 490, + "token_acc": 0.8409646758677575 + }, + { + "epoch": 0.11531531531531532, + "grad_norm": 105.73524475097656, + "learning_rate": 8.94570254698197e-06, + "loss": 8.90081558227539, + "loss_idm": 0.49921875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0226318359375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5032265663146973, + "loss_sft": 0.01476171426475048, + "step": 500, + "token_acc": 0.8424521747539707 + }, + { + "epoch": 0.11531531531531532, + "eval_loss": 0.873286783695221, + "eval_runtime": 169.8257, + "eval_samples_per_second": 10.275, + "eval_steps_per_second": 2.573, + "eval_token_acc": 0.8295710593091115, + "step": 500 + }, + { + "epoch": 0.11762162162162162, + "grad_norm": 79.52201843261719, + "learning_rate": 8.894386393810563e-06, + "loss": 8.980358123779297, + "loss_idm": 0.501171875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0220947265625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5085306227207184, + "loss_sft": 0.014769895002245903, + "step": 510, + "token_acc": 0.8416994573590164 + }, + { + "epoch": 0.11992792792792793, + "grad_norm": 87.73248291015625, + "learning_rate": 8.842005554284296e-06, + "loss": 8.839938354492187, + "loss_idm": 0.4921875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0215576171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5001400589942933, + "loss_sft": 0.014775922894477845, + "step": 520, + "token_acc": 0.8430164098119748 + }, + { + "epoch": 0.12223423423423424, + "grad_norm": 108.73457336425781, + "learning_rate": 8.788574348801676e-06, + "loss": 9.055592346191407, + "loss_idm": 0.50859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.021142578125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5146129488945007, + "loss_sft": 0.014646383747458459, + "step": 530, + "token_acc": 0.843842382599978 + }, + { + "epoch": 0.12454054054054053, + "grad_norm": 98.13040924072266, + "learning_rate": 8.734107384920771e-06, + "loss": 8.952853393554687, + "loss_idm": 0.50625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0204833984375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5087056159973145, + "loss_sft": 0.014654205366969108, + "step": 540, + "token_acc": 0.8428865032158228 + }, + { + "epoch": 0.12684684684684686, + "grad_norm": 120.66363525390625, + "learning_rate": 8.67861955336566e-06, + "loss": 8.789979553222656, + "loss_idm": 0.48984375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02001953125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.49941076040267945, + "loss_sft": 0.014644744619727135, + "step": 550, + "token_acc": 0.8429765894640101 + }, + { + "epoch": 0.12915315315315315, + "grad_norm": 105.6294174194336, + "learning_rate": 8.622126023955446e-06, + "loss": 8.306752014160157, + "loss_idm": 0.47109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0207763671875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4683968424797058, + "loss_sft": 0.014503800868988037, + "step": 560, + "token_acc": 0.844274669579772 + }, + { + "epoch": 0.13145945945945947, + "grad_norm": 87.99810028076172, + "learning_rate": 8.564642241456986e-06, + "loss": 9.134395599365234, + "loss_idm": 0.5078125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0187744140625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5218061327934265, + "loss_sft": 0.014652644842863083, + "step": 570, + "token_acc": 0.8432568832791741 + }, + { + "epoch": 0.13376576576576577, + "grad_norm": 94.22286224365234, + "learning_rate": 8.506183921362443e-06, + "loss": 9.002159118652344, + "loss_idm": 0.50390625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0199462890625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5122206807136536, + "loss_sft": 0.01476593129336834, + "step": 580, + "token_acc": 0.8415455047829566 + }, + { + "epoch": 0.13607207207207206, + "grad_norm": 96.09319305419922, + "learning_rate": 8.446767045592829e-06, + "loss": 8.899197387695313, + "loss_idm": 0.48984375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.020263671875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5060975551605225, + "loss_sft": 0.014485136792063712, + "step": 590, + "token_acc": 0.8453194613833785 + }, + { + "epoch": 0.13837837837837838, + "grad_norm": 91.08513641357422, + "learning_rate": 8.386407858128707e-06, + "loss": 8.522463989257812, + "loss_idm": 0.47265625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0227294921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4799580335617065, + "loss_sft": 0.01448415480554104, + "step": 600, + "token_acc": 0.8451588424854661 + }, + { + "epoch": 0.14068468468468467, + "grad_norm": 91.31858825683594, + "learning_rate": 8.325122860569241e-06, + "loss": 8.168463897705077, + "loss_idm": 0.4453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0220458984375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.45817803144454955, + "loss_sft": 0.01455080546438694, + "step": 610, + "token_acc": 0.8434149712322799 + }, + { + "epoch": 0.142990990990991, + "grad_norm": 92.19358825683594, + "learning_rate": 8.262928807620843e-06, + "loss": 9.086221313476562, + "loss_idm": 0.5046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0214111328125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5163055658340454, + "loss_sft": 0.014690371602773667, + "step": 620, + "token_acc": 0.8422049082316321 + }, + { + "epoch": 0.1452972972972973, + "grad_norm": 85.43914794921875, + "learning_rate": 8.199842702516584e-06, + "loss": 8.663780212402344, + "loss_idm": 0.47265625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0191650390625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.49150881767272947, + "loss_sft": 0.01495927982032299, + "step": 630, + "token_acc": 0.8403844066897375 + }, + { + "epoch": 0.1476036036036036, + "grad_norm": 100.33062744140625, + "learning_rate": 8.135881792367686e-06, + "loss": 8.27200927734375, + "loss_idm": 0.4625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0180419921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.46880000829696655, + "loss_sft": 0.014622250199317932, + "step": 640, + "token_acc": 0.8430498513734264 + }, + { + "epoch": 0.1499099099099099, + "grad_norm": 102.80355834960938, + "learning_rate": 8.071063563448341e-06, + "loss": 8.619700622558593, + "loss_idm": 0.4796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0202880859375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4885416507720947, + "loss_sft": 0.014433516189455986, + "step": 650, + "token_acc": 0.845255299241181 + }, + { + "epoch": 0.15221621621621623, + "grad_norm": 88.60161590576172, + "learning_rate": 8.005405736415127e-06, + "loss": 8.179424285888672, + "loss_idm": 0.460546875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0212158203125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4592796742916107, + "loss_sft": 0.014803672954440118, + "step": 660, + "token_acc": 0.8411518466838719 + }, + { + "epoch": 0.15452252252252252, + "grad_norm": 96.72508239746094, + "learning_rate": 7.938926261462366e-06, + "loss": 8.372673797607423, + "loss_idm": 0.45859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0193115234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4736079216003418, + "loss_sft": 0.014764882624149323, + "step": 670, + "token_acc": 0.8415420786843315 + }, + { + "epoch": 0.15682882882882884, + "grad_norm": 109.25978088378906, + "learning_rate": 7.871643313414718e-06, + "loss": 8.376298522949218, + "loss_idm": 0.46875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0222412109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.47051931619644166, + "loss_sft": 0.014839059859514236, + "step": 680, + "token_acc": 0.8409984786981344 + }, + { + "epoch": 0.15913513513513514, + "grad_norm": 103.52886962890625, + "learning_rate": 7.803575286758365e-06, + "loss": 7.899095916748047, + "loss_idm": 0.44375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0193359375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.44420482516288756, + "loss_sft": 0.014591414108872413, + "step": 690, + "token_acc": 0.8434285377428007 + }, + { + "epoch": 0.16144144144144143, + "grad_norm": 116.31279754638672, + "learning_rate": 7.734740790612137e-06, + "loss": 7.9990592956542965, + "loss_idm": 0.44296875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01865234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4515872120857239, + "loss_sft": 0.014406795054674149, + "step": 700, + "token_acc": 0.8455261067379216 + }, + { + "epoch": 0.16374774774774775, + "grad_norm": 104.14566802978516, + "learning_rate": 7.66515864363997e-06, + "loss": 8.241651916503907, + "loss_idm": 0.4578125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0193603515625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.46591639518737793, + "loss_sft": 0.014483961462974548, + "step": 710, + "token_acc": 0.8448154530557234 + }, + { + "epoch": 0.16605405405405405, + "grad_norm": 104.71328735351562, + "learning_rate": 7.594847868906076e-06, + "loss": 8.089508819580079, + "loss_idm": 0.45234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0209228515625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4547982931137085, + "loss_sft": 0.01439845934510231, + "step": 720, + "token_acc": 0.8451516312280465 + }, + { + "epoch": 0.16836036036036037, + "grad_norm": 107.36788177490234, + "learning_rate": 7.52382768867422e-06, + "loss": 8.494679260253907, + "loss_idm": 0.471875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0205322265625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.48007341623306277, + "loss_sft": 0.014671024680137635, + "step": 730, + "token_acc": 0.8430621123298506 + }, + { + "epoch": 0.17066666666666666, + "grad_norm": 84.9700927734375, + "learning_rate": 7.452117519152542e-06, + "loss": 7.954332733154297, + "loss_idm": 0.4421875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.020751953125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4460784077644348, + "loss_sft": 0.014631084725260734, + "step": 740, + "token_acc": 0.84308325249699 + }, + { + "epoch": 0.17297297297297298, + "grad_norm": 126.57601928710938, + "learning_rate": 7.379736965185369e-06, + "loss": 8.158969116210937, + "loss_idm": 0.45859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0198974609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.459906268119812, + "loss_sft": 0.014672993496060372, + "step": 750, + "token_acc": 0.8428532139788794 + }, + { + "epoch": 0.17527927927927928, + "grad_norm": 92.83258819580078, + "learning_rate": 7.30670581489344e-06, + "loss": 8.280567932128907, + "loss_idm": 0.46171875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.019189453125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.46855597496032714, + "loss_sft": 0.014394653588533401, + "step": 760, + "token_acc": 0.8446953594377756 + }, + { + "epoch": 0.1775855855855856, + "grad_norm": 111.44788360595703, + "learning_rate": 7.233044034264034e-06, + "loss": 7.862223815917969, + "loss_idm": 0.4359375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0191162109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4427155554294586, + "loss_sft": 0.014474740624427796, + "step": 770, + "token_acc": 0.8450180938562969 + }, + { + "epoch": 0.1798918918918919, + "grad_norm": 115.48260498046875, + "learning_rate": 7.158771761692464e-06, + "loss": 7.9805961608886715, + "loss_idm": 0.44453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0202392578125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.44799396991729734, + "loss_sft": 0.01469801552593708, + "step": 780, + "token_acc": 0.8424233709705329 + }, + { + "epoch": 0.18219819819819819, + "grad_norm": 94.31640625, + "learning_rate": 7.083909302476453e-06, + "loss": 9.633565521240234, + "loss_idm": 0.5421875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0183837890625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5536909341812134, + "loss_sft": 0.014749213308095931, + "step": 790, + "token_acc": 0.8422723821357079 + }, + { + "epoch": 0.1845045045045045, + "grad_norm": 96.44784545898438, + "learning_rate": 7.008477123264849e-06, + "loss": 7.915254974365235, + "loss_idm": 0.4421875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0171142578125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4477837860584259, + "loss_sft": 0.014466813579201699, + "step": 800, + "token_acc": 0.8447225350327021 + }, + { + "epoch": 0.1868108108108108, + "grad_norm": 84.42618560791016, + "learning_rate": 6.932495846462262e-06, + "loss": 8.682425689697265, + "loss_idm": 0.4859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017529296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.49512912034988404, + "loss_sft": 0.014556843042373657, + "step": 810, + "token_acc": 0.8437006309304983 + }, + { + "epoch": 0.18911711711711712, + "grad_norm": 114.39678192138672, + "learning_rate": 6.855986244591104e-06, + "loss": 7.2060691833496096, + "loss_idm": 0.394921875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.019873046875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.40005741119384763, + "loss_sft": 0.014659629762172699, + "step": 820, + "token_acc": 0.8422255628738242 + }, + { + "epoch": 0.19142342342342342, + "grad_norm": 125.64624786376953, + "learning_rate": 6.778969234612583e-06, + "loss": 8.314984130859376, + "loss_idm": 0.465234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.018212890625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4713524401187897, + "loss_sft": 0.014716658741235733, + "step": 830, + "token_acc": 0.8419139432862319 + }, + { + "epoch": 0.19372972972972974, + "grad_norm": 82.18567657470703, + "learning_rate": 6.701465872208216e-06, + "loss": 8.177500152587891, + "loss_idm": 0.46015625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.018994140625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.46165603399276733, + "loss_sft": 0.014731941372156143, + "step": 840, + "token_acc": 0.8417456898652139 + }, + { + "epoch": 0.19603603603603603, + "grad_norm": 76.3083724975586, + "learning_rate": 6.6234973460234184e-06, + "loss": 6.866006469726562, + "loss_idm": 0.375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0199951171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.37900240421295167, + "loss_sft": 0.014587616920471192, + "step": 850, + "token_acc": 0.8433828349109465 + }, + { + "epoch": 0.19834234234234235, + "grad_norm": 87.86666107177734, + "learning_rate": 6.545084971874738e-06, + "loss": 9.54729766845703, + "loss_idm": 0.52734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.020166015625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.5463018894195557, + "loss_sft": 0.014688147231936455, + "step": 860, + "token_acc": 0.843223021860908 + }, + { + "epoch": 0.20064864864864865, + "grad_norm": 79.54348754882812, + "learning_rate": 6.466250186922325e-06, + "loss": 7.474161529541016, + "loss_idm": 0.41015625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0190673828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4184339165687561, + "loss_sft": 0.014442858472466468, + "step": 870, + "token_acc": 0.8446895228961879 + }, + { + "epoch": 0.20295495495495494, + "grad_norm": 101.24223327636719, + "learning_rate": 6.387014543809224e-06, + "loss": 7.865777587890625, + "loss_idm": 0.435546875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02099609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4404805421829224, + "loss_sft": 0.014567815512418748, + "step": 880, + "token_acc": 0.8441662849959559 + }, + { + "epoch": 0.20526126126126126, + "grad_norm": 113.0296401977539, + "learning_rate": 6.3073997047691e-06, + "loss": 7.874870300292969, + "loss_idm": 0.440625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0199462890625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4417296290397644, + "loss_sft": 0.014670121297240258, + "step": 890, + "token_acc": 0.8433289124014535 + }, + { + "epoch": 0.20756756756756756, + "grad_norm": 103.46904754638672, + "learning_rate": 6.227427435703997e-06, + "loss": 8.081219482421876, + "loss_idm": 0.444140625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0183837890625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.45728793144226076, + "loss_sft": 0.014389193430542947, + "step": 900, + "token_acc": 0.8454201846502517 + }, + { + "epoch": 0.20987387387387388, + "grad_norm": 88.96587371826172, + "learning_rate": 6.147119600233758e-06, + "loss": 8.543597412109374, + "loss_idm": 0.4796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.019921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4841804504394531, + "loss_sft": 0.014478649944067001, + "step": 910, + "token_acc": 0.8443082839472513 + }, + { + "epoch": 0.21218018018018017, + "grad_norm": 117.9122085571289, + "learning_rate": 6.066498153718735e-06, + "loss": 7.415792083740234, + "loss_idm": 0.4140625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0222412109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4112999439239502, + "loss_sft": 0.014585891366004944, + "step": 920, + "token_acc": 0.8436708029516633 + }, + { + "epoch": 0.2144864864864865, + "grad_norm": 86.82301330566406, + "learning_rate": 5.985585137257401e-06, + "loss": 7.67747802734375, + "loss_idm": 0.427734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0196044921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4302915811538696, + "loss_sft": 0.014485430717468262, + "step": 930, + "token_acc": 0.8436629761319727 + }, + { + "epoch": 0.2167927927927928, + "grad_norm": 95.94313049316406, + "learning_rate": 5.904402671660551e-06, + "loss": 7.310446166992188, + "loss_idm": 0.4046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0189697265625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.40787339210510254, + "loss_sft": 0.014458712935447694, + "step": 940, + "token_acc": 0.845064712919843 + }, + { + "epoch": 0.2190990990990991, + "grad_norm": 90.85774230957031, + "learning_rate": 5.82297295140367e-06, + "loss": 7.37610092163086, + "loss_idm": 0.41015625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.019921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.41144019961357114, + "loss_sft": 0.014352937042713166, + "step": 950, + "token_acc": 0.8450359871010459 + }, + { + "epoch": 0.2214054054054054, + "grad_norm": 91.5264892578125, + "learning_rate": 5.74131823855921e-06, + "loss": 7.979737854003906, + "loss_idm": 0.439453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0195068359375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.450022554397583, + "loss_sft": 0.014237866923213004, + "step": 960, + "token_acc": 0.8463489009812982 + }, + { + "epoch": 0.22371171171171172, + "grad_norm": 89.29737854003906, + "learning_rate": 5.659460856710346e-06, + "loss": 7.60855712890625, + "loss_idm": 0.41953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177490234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4289973974227905, + "loss_sft": 0.014107778668403625, + "step": 970, + "token_acc": 0.8469193499209564 + }, + { + "epoch": 0.22601801801801802, + "grad_norm": 106.5138168334961, + "learning_rate": 5.577423184847932e-06, + "loss": 8.291693115234375, + "loss_idm": 0.46171875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0203857421875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4682368040084839, + "loss_sft": 0.014279923215508462, + "step": 980, + "token_acc": 0.8464010922588995 + }, + { + "epoch": 0.2283243243243243, + "grad_norm": 117.88599395751953, + "learning_rate": 5.495227651252315e-06, + "loss": 7.344970703125, + "loss_idm": 0.402734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0205810546875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.40857953429222105, + "loss_sft": 0.014561992883682252, + "step": 990, + "token_acc": 0.8436562465051372 + }, + { + "epoch": 0.23063063063063063, + "grad_norm": 107.79817199707031, + "learning_rate": 5.412896727361663e-06, + "loss": 7.102320861816406, + "loss_idm": 0.397265625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0206787109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3930091679096222, + "loss_sft": 0.014505467936396598, + "step": 1000, + "token_acc": 0.8436447490694896 + }, + { + "epoch": 0.23063063063063063, + "eval_loss": 0.826145350933075, + "eval_runtime": 168.0761, + "eval_samples_per_second": 10.382, + "eval_steps_per_second": 2.6, + "eval_token_acc": 0.8308116622599108, + "step": 1000 + }, + { + "epoch": 0.23293693693693693, + "grad_norm": 88.33614349365234, + "learning_rate": 5.3304529216284974e-06, + "loss": 7.70971450805664, + "loss_idm": 0.42890625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0201171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4321062684059143, + "loss_sft": 0.014323608577251434, + "step": 1010, + "token_acc": 0.8463429671668039 + }, + { + "epoch": 0.23524324324324325, + "grad_norm": 103.5328369140625, + "learning_rate": 5.247918773366112e-06, + "loss": 7.400896453857422, + "loss_idm": 0.411328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.018603515625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.41432325839996337, + "loss_sft": 0.014487444236874581, + "step": 1020, + "token_acc": 0.8436925959265521 + }, + { + "epoch": 0.23754954954954954, + "grad_norm": 95.37596893310547, + "learning_rate": 5.165316846586541e-06, + "loss": 7.066916656494141, + "loss_idm": 0.398828125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01865234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3934369027614594, + "loss_sft": 0.01427248790860176, + "step": 1030, + "token_acc": 0.8449177570364499 + }, + { + "epoch": 0.23985585585585587, + "grad_norm": 106.63520050048828, + "learning_rate": 5.082669723831793e-06, + "loss": 7.509798431396485, + "loss_idm": 0.41484375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177490234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.42152470350265503, + "loss_sft": 0.014691157266497612, + "step": 1040, + "token_acc": 0.8420689476522522 + }, + { + "epoch": 0.24216216216216216, + "grad_norm": 110.2610855102539, + "learning_rate": 5e-06, + "loss": 6.679129028320313, + "loss_idm": 0.367578125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177490234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3698115944862366, + "loss_sft": 0.01454925611615181, + "step": 1050, + "token_acc": 0.8429293549000219 + }, + { + "epoch": 0.24446846846846848, + "grad_norm": 116.46825408935547, + "learning_rate": 4.917330276168208e-06, + "loss": 7.56883544921875, + "loss_idm": 0.426171875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0185791015625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4245879828929901, + "loss_sft": 0.014570457488298416, + "step": 1060, + "token_acc": 0.843766214243804 + }, + { + "epoch": 0.24677477477477477, + "grad_norm": 90.62398529052734, + "learning_rate": 4.8346831534134595e-06, + "loss": 8.461653137207032, + "loss_idm": 0.48203125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.019677734375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4794220805168152, + "loss_sft": 0.014480674639344215, + "step": 1070, + "token_acc": 0.8445398105746493 + }, + { + "epoch": 0.24908108108108107, + "grad_norm": 97.12873840332031, + "learning_rate": 4.752081226633888e-06, + "loss": 7.096697998046875, + "loss_idm": 0.391015625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017236328125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.39673874378204343, + "loss_sft": 0.014352129772305489, + "step": 1080, + "token_acc": 0.8445942044328101 + }, + { + "epoch": 0.25138738738738736, + "grad_norm": 148.5345458984375, + "learning_rate": 4.669547078371503e-06, + "loss": 8.134896850585937, + "loss_idm": 0.453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01767578125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4617579817771912, + "loss_sft": 0.01423180066049099, + "step": 1090, + "token_acc": 0.8462392484597084 + }, + { + "epoch": 0.2536936936936937, + "grad_norm": 99.68243408203125, + "learning_rate": 4.587103272638339e-06, + "loss": 6.537295532226563, + "loss_idm": 0.363671875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01689453125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3626400887966156, + "loss_sft": 0.014130813628435135, + "step": 1100, + "token_acc": 0.847496286866115 + }, + { + "epoch": 0.256, + "grad_norm": 106.85647583007812, + "learning_rate": 4.504772348747687e-06, + "loss": 7.332121276855469, + "loss_idm": 0.40234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017724609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.41151710152626036, + "loss_sft": 0.014223045855760574, + "step": 1110, + "token_acc": 0.8462705823615773 + }, + { + "epoch": 0.2583063063063063, + "grad_norm": 107.44613647460938, + "learning_rate": 4.42257681515207e-06, + "loss": 6.910240173339844, + "loss_idm": 0.381640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.018505859375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3840563654899597, + "loss_sft": 0.014243298768997192, + "step": 1120, + "token_acc": 0.8465594450949473 + }, + { + "epoch": 0.2606126126126126, + "grad_norm": 107.26278686523438, + "learning_rate": 4.340539143289655e-06, + "loss": 7.2168525695800785, + "loss_idm": 0.403125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0174560546875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.40406553745269774, + "loss_sft": 0.014284095168113709, + "step": 1130, + "token_acc": 0.8454709336305773 + }, + { + "epoch": 0.26291891891891894, + "grad_norm": 116.3701400756836, + "learning_rate": 4.25868176144079e-06, + "loss": 7.558566284179688, + "loss_idm": 0.419921875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177978515625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4247333943843842, + "loss_sft": 0.014513048902153968, + "step": 1140, + "token_acc": 0.8439526822829935 + }, + { + "epoch": 0.26522522522522524, + "grad_norm": 106.37321472167969, + "learning_rate": 4.17702704859633e-06, + "loss": 7.477565765380859, + "loss_idm": 0.41640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0197021484375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4179769098758698, + "loss_sft": 0.01451987624168396, + "step": 1150, + "token_acc": 0.8434178920433081 + }, + { + "epoch": 0.26753153153153153, + "grad_norm": 115.76299285888672, + "learning_rate": 4.0955973283394525e-06, + "loss": 7.254231262207031, + "loss_idm": 0.402734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0183837890625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.405047333240509, + "loss_sft": 0.014345861971378326, + "step": 1160, + "token_acc": 0.8443389144238174 + }, + { + "epoch": 0.2698378378378378, + "grad_norm": 158.42250061035156, + "learning_rate": 4.0144148627426e-06, + "loss": 7.521492004394531, + "loss_idm": 0.41796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0178955078125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4225184381008148, + "loss_sft": 0.01449190303683281, + "step": 1170, + "token_acc": 0.8435488968221455 + }, + { + "epoch": 0.2721441441441441, + "grad_norm": 103.64552307128906, + "learning_rate": 3.9335018462812664e-06, + "loss": 8.284562683105468, + "loss_idm": 0.45546875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0184814453125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.46931086778640746, + "loss_sft": 0.014623204618692398, + "step": 1180, + "token_acc": 0.8422325811350072 + }, + { + "epoch": 0.27445045045045047, + "grad_norm": 90.13056945800781, + "learning_rate": 3.852880399766243e-06, + "loss": 7.880992126464844, + "loss_idm": 0.440234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0189208984375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.44379000663757323, + "loss_sft": 0.014453979581594468, + "step": 1190, + "token_acc": 0.8441096901131333 + }, + { + "epoch": 0.27675675675675676, + "grad_norm": 104.42198944091797, + "learning_rate": 3.7725725642960047e-06, + "loss": 6.547203826904297, + "loss_idm": 0.3640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01826171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.36079212427139284, + "loss_sft": 0.014533436298370362, + "step": 1200, + "token_acc": 0.8435185033768867 + }, + { + "epoch": 0.27906306306306305, + "grad_norm": 109.83740997314453, + "learning_rate": 3.6926002952309015e-06, + "loss": 7.021842956542969, + "loss_idm": 0.386328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01845703125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.39095706343650816, + "loss_sft": 0.014352500438690186, + "step": 1210, + "token_acc": 0.8457043054504614 + }, + { + "epoch": 0.28136936936936935, + "grad_norm": 92.76325988769531, + "learning_rate": 3.6129854561907786e-06, + "loss": 6.612565612792968, + "loss_idm": 0.363671875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177490234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3655923068523407, + "loss_sft": 0.014593320712447167, + "step": 1220, + "token_acc": 0.8422301537496077 + }, + { + "epoch": 0.2836756756756757, + "grad_norm": 108.94733428955078, + "learning_rate": 3.533749813077677e-06, + "loss": 6.7885498046875, + "loss_idm": 0.3734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01787109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3769371509552002, + "loss_sft": 0.014226069301366806, + "step": 1230, + "token_acc": 0.8456893225753114 + }, + { + "epoch": 0.285981981981982, + "grad_norm": 130.9772186279297, + "learning_rate": 3.4549150281252635e-06, + "loss": 6.741462707519531, + "loss_idm": 0.37109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0176513671875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.37425958514213564, + "loss_sft": 0.014150273799896241, + "step": 1240, + "token_acc": 0.8465476829923336 + }, + { + "epoch": 0.2882882882882883, + "grad_norm": 119.88499450683594, + "learning_rate": 3.3765026539765832e-06, + "loss": 7.194857788085938, + "loss_idm": 0.400390625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.018896484375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4020317316055298, + "loss_sft": 0.013935164362192155, + "step": 1250, + "token_acc": 0.8496373374216677 + }, + { + "epoch": 0.2905945945945946, + "grad_norm": 107.04531860351562, + "learning_rate": 3.298534127791785e-06, + "loss": 7.381626892089844, + "loss_idm": 0.409765625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01904296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.413124680519104, + "loss_sft": 0.014187094569206238, + "step": 1260, + "token_acc": 0.8464285982464 + }, + { + "epoch": 0.2929009009009009, + "grad_norm": 133.18710327148438, + "learning_rate": 3.2210307653874175e-06, + "loss": 6.52490234375, + "loss_idm": 0.360546875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016943359375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.36171150803565977, + "loss_sft": 0.01412605307996273, + "step": 1270, + "token_acc": 0.8482812389612749 + }, + { + "epoch": 0.2952072072072072, + "grad_norm": 133.87335205078125, + "learning_rate": 3.1440137554088957e-06, + "loss": 7.00994873046875, + "loss_idm": 0.389453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0179931640625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.39080272912979125, + "loss_sft": 0.014256376028060912, + "step": 1280, + "token_acc": 0.8455885394740364 + }, + { + "epoch": 0.2975135135135135, + "grad_norm": 118.04527282714844, + "learning_rate": 3.06750415353774e-06, + "loss": 6.74976806640625, + "loss_idm": 0.37421875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017529296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3751899302005768, + "loss_sft": 0.014237936958670616, + "step": 1290, + "token_acc": 0.8457528871889156 + }, + { + "epoch": 0.2998198198198198, + "grad_norm": 152.24542236328125, + "learning_rate": 2.991522876735154e-06, + "loss": 6.73217544555664, + "loss_idm": 0.375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0180419921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.373495876789093, + "loss_sft": 0.014310174062848091, + "step": 1300, + "token_acc": 0.8459135928839514 + }, + { + "epoch": 0.3021261261261261, + "grad_norm": 143.8121337890625, + "learning_rate": 2.9160906975235493e-06, + "loss": 6.565863037109375, + "loss_idm": 0.365234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.019970703125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.36084928512573244, + "loss_sft": 0.014364153146743774, + "step": 1310, + "token_acc": 0.8448077517079684 + }, + { + "epoch": 0.30443243243243245, + "grad_norm": 115.51368713378906, + "learning_rate": 2.8412282383075362e-06, + "loss": 6.6176704406738285, + "loss_idm": 0.365234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.015966796875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.36805868744850156, + "loss_sft": 0.01430610753595829, + "step": 1320, + "token_acc": 0.8448409814290369 + }, + { + "epoch": 0.30673873873873875, + "grad_norm": 103.80223846435547, + "learning_rate": 2.766955965735968e-06, + "loss": 6.629913330078125, + "loss_idm": 0.36796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.018212890625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3660490334033966, + "loss_sft": 0.014540331438183784, + "step": 1330, + "token_acc": 0.8437379067938862 + }, + { + "epoch": 0.30904504504504504, + "grad_norm": 123.66497039794922, + "learning_rate": 2.693294185106562e-06, + "loss": 6.788928985595703, + "loss_idm": 0.38046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01904296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.37561420202255247, + "loss_sft": 0.014337588101625443, + "step": 1340, + "token_acc": 0.8445305506848307 + }, + { + "epoch": 0.31135135135135134, + "grad_norm": 90.85652160644531, + "learning_rate": 2.6202630348146323e-06, + "loss": 6.739347839355469, + "loss_idm": 0.371484375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016357421875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3759130001068115, + "loss_sft": 0.014109160378575325, + "step": 1350, + "token_acc": 0.846979458728549 + }, + { + "epoch": 0.3136576576576577, + "grad_norm": 120.79927825927734, + "learning_rate": 2.5478824808474613e-06, + "loss": 6.946615600585938, + "loss_idm": 0.388671875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01796875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.38742896914482117, + "loss_sft": 0.014025571942329406, + "step": 1360, + "token_acc": 0.8474245017667137 + }, + { + "epoch": 0.315963963963964, + "grad_norm": 127.58772277832031, + "learning_rate": 2.476172311325783e-06, + "loss": 6.843831634521484, + "loss_idm": 0.379296875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177978515625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.38061107993125914, + "loss_sft": 0.014309322834014893, + "step": 1370, + "token_acc": 0.8454213580875642 + }, + { + "epoch": 0.3182702702702703, + "grad_norm": 151.91925048828125, + "learning_rate": 2.4051521310939258e-06, + "loss": 6.953953552246094, + "loss_idm": 0.3859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016845703125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.38837702870368956, + "loss_sft": 0.014433859288692475, + "step": 1380, + "token_acc": 0.8446607330752671 + }, + { + "epoch": 0.32057657657657657, + "grad_norm": 142.9082489013672, + "learning_rate": 2.3348413563600324e-06, + "loss": 7.533473205566406, + "loss_idm": 0.41640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01787109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.42370715737342834, + "loss_sft": 0.01422720868140459, + "step": 1390, + "token_acc": 0.8468790057744192 + }, + { + "epoch": 0.32288288288288286, + "grad_norm": 87.2136001586914, + "learning_rate": 2.265259209387867e-06, + "loss": 6.789948272705078, + "loss_idm": 0.373046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0173095703125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3781958520412445, + "loss_sft": 0.014054833725094796, + "step": 1400, + "token_acc": 0.8476745137822668 + }, + { + "epoch": 0.3251891891891892, + "grad_norm": 109.24415588378906, + "learning_rate": 2.1964247132416373e-06, + "loss": 6.430654144287109, + "loss_idm": 0.356640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0153076171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3571660578250885, + "loss_sft": 0.01431470774114132, + "step": 1410, + "token_acc": 0.8448665901524277 + }, + { + "epoch": 0.3274954954954955, + "grad_norm": 116.79399871826172, + "learning_rate": 2.1283566865852824e-06, + "loss": 6.299882507324218, + "loss_idm": 0.344140625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017724609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.34696500301361083, + "loss_sft": 0.014206652343273164, + "step": 1420, + "token_acc": 0.846456135209066 + }, + { + "epoch": 0.3298018018018018, + "grad_norm": 149.03395080566406, + "learning_rate": 2.061073738537635e-06, + "loss": 6.599853515625, + "loss_idm": 0.36328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0202392578125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.36307021975517273, + "loss_sft": 0.014167750626802445, + "step": 1430, + "token_acc": 0.8466417422677717 + }, + { + "epoch": 0.3321081081081081, + "grad_norm": 132.9679718017578, + "learning_rate": 1.9945942635848745e-06, + "loss": 6.743374633789062, + "loss_idm": 0.374609375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0191162109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3731849730014801, + "loss_sft": 0.014212041348218917, + "step": 1440, + "token_acc": 0.8470092899704256 + }, + { + "epoch": 0.33441441441441444, + "grad_norm": 119.5426254272461, + "learning_rate": 1.928936436551661e-06, + "loss": 6.725025939941406, + "loss_idm": 0.37265625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017724609375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3735596239566803, + "loss_sft": 0.014224140346050263, + "step": 1450, + "token_acc": 0.8464133441783575 + }, + { + "epoch": 0.33672072072072073, + "grad_norm": 139.33950805664062, + "learning_rate": 1.864118207632315e-06, + "loss": 6.592694091796875, + "loss_idm": 0.36484375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0165771484375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.36636746525764463, + "loss_sft": 0.014215507730841636, + "step": 1460, + "token_acc": 0.8457222099999259 + }, + { + "epoch": 0.33902702702702703, + "grad_norm": 126.1017837524414, + "learning_rate": 1.8001572974834169e-06, + "loss": 6.243590545654297, + "loss_idm": 0.34140625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0188232421875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3425869166851044, + "loss_sft": 0.014009960740804673, + "step": 1470, + "token_acc": 0.8481653395222281 + }, + { + "epoch": 0.3413333333333333, + "grad_norm": 143.65353393554688, + "learning_rate": 1.7370711923791567e-06, + "loss": 6.1479026794433596, + "loss_idm": 0.3359375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0189697265625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3357097864151001, + "loss_sft": 0.014131288602948188, + "step": 1480, + "token_acc": 0.8479393838271811 + }, + { + "epoch": 0.3436396396396396, + "grad_norm": 121.66058349609375, + "learning_rate": 1.6748771394307584e-06, + "loss": 6.679605865478516, + "loss_idm": 0.370703125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.37073889970779417, + "loss_sft": 0.01429830901324749, + "step": 1490, + "token_acc": 0.8450069266156981 + }, + { + "epoch": 0.34594594594594597, + "grad_norm": 155.4982452392578, + "learning_rate": 1.6135921418712959e-06, + "loss": 6.242042160034179, + "loss_idm": 0.3453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177490234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.34358363747596743, + "loss_sft": 0.01402926966547966, + "step": 1500, + "token_acc": 0.8482708050343194 + }, + { + "epoch": 0.34594594594594597, + "eval_loss": 0.7831034064292908, + "eval_runtime": 233.7646, + "eval_samples_per_second": 7.465, + "eval_steps_per_second": 1.869, + "eval_token_acc": 0.8332306899706579, + "step": 1500 + }, + { + "epoch": 0.34825225225225226, + "grad_norm": 165.9364776611328, + "learning_rate": 1.5532329544071712e-06, + "loss": 6.288203430175781, + "loss_idm": 0.3484375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0178955078125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.34598525166511535, + "loss_sft": 0.014146770164370537, + "step": 1510, + "token_acc": 0.8467197018422101 + }, + { + "epoch": 0.35055855855855855, + "grad_norm": 122.34672546386719, + "learning_rate": 1.4938160786375571e-06, + "loss": 6.6126861572265625, + "loss_idm": 0.36796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0166748046875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.36726664900779726, + "loss_sft": 0.014271648973226548, + "step": 1520, + "token_acc": 0.8459442144687561 + }, + { + "epoch": 0.35286486486486485, + "grad_norm": 117.75894927978516, + "learning_rate": 1.4353577585430152e-06, + "loss": 6.7383270263671875, + "loss_idm": 0.374609375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0175048828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.37455575466156005, + "loss_sft": 0.014166099578142166, + "step": 1530, + "token_acc": 0.8471839671377358 + }, + { + "epoch": 0.3551711711711712, + "grad_norm": 131.8136749267578, + "learning_rate": 1.3778739760445552e-06, + "loss": 5.911068725585937, + "loss_idm": 0.327734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0180419921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.32219248414039614, + "loss_sft": 0.014310959726572037, + "step": 1540, + "token_acc": 0.845075592022814 + }, + { + "epoch": 0.3574774774774775, + "grad_norm": 109.31105041503906, + "learning_rate": 1.321380446634342e-06, + "loss": 6.490184020996094, + "loss_idm": 0.3640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01689453125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3597273170948029, + "loss_sft": 0.014035460725426673, + "step": 1550, + "token_acc": 0.8480154337268537 + }, + { + "epoch": 0.3597837837837838, + "grad_norm": 137.36572265625, + "learning_rate": 1.2658926150792321e-06, + "loss": 6.138148498535156, + "loss_idm": 0.337109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016943359375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3371982276439667, + "loss_sft": 0.014372247084975243, + "step": 1560, + "token_acc": 0.8449733133737845 + }, + { + "epoch": 0.3620900900900901, + "grad_norm": 105.96505737304688, + "learning_rate": 1.2114256511983274e-06, + "loss": 5.880138778686524, + "loss_idm": 0.319140625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017529296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.32088192105293273, + "loss_sft": 0.014091270789504052, + "step": 1570, + "token_acc": 0.8476390387833302 + }, + { + "epoch": 0.36439639639639637, + "grad_norm": 122.6405029296875, + "learning_rate": 1.157994445715706e-06, + "loss": 5.690219879150391, + "loss_idm": 0.3109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0180419921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.30828317403793337, + "loss_sft": 0.014205870404839515, + "step": 1580, + "token_acc": 0.8467056621660582 + }, + { + "epoch": 0.3667027027027027, + "grad_norm": 143.67105102539062, + "learning_rate": 1.1056136061894386e-06, + "loss": 6.784506225585938, + "loss_idm": 0.376953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0180908203125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3764196753501892, + "loss_sft": 0.014350217953324317, + "step": 1590, + "token_acc": 0.8455792785275132 + }, + { + "epoch": 0.369009009009009, + "grad_norm": 158.58287048339844, + "learning_rate": 1.0542974530180327e-06, + "loss": 6.497494506835937, + "loss_idm": 0.361328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017529296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.35958645343780515, + "loss_sft": 0.01420062966644764, + "step": 1600, + "token_acc": 0.8465405585403796 + }, + { + "epoch": 0.3713153153153153, + "grad_norm": 127.48925018310547, + "learning_rate": 1.0040600155253766e-06, + "loss": 6.240549850463867, + "loss_idm": 0.34296875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01611328125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.34473289251327516, + "loss_sft": 0.014279846474528313, + "step": 1610, + "token_acc": 0.8457844387450508 + }, + { + "epoch": 0.3736216216216216, + "grad_norm": 130.4225616455078, + "learning_rate": 9.549150281252633e-07, + "loss": 6.62506332397461, + "loss_idm": 0.36796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017529296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3672377049922943, + "loss_sft": 0.01430715098977089, + "step": 1620, + "token_acc": 0.8453087239015887 + }, + { + "epoch": 0.37592792792792795, + "grad_norm": 155.16566467285156, + "learning_rate": 9.068759265665384e-07, + "loss": 7.453045654296875, + "loss_idm": 0.414453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01591796875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.4206882119178772, + "loss_sft": 0.014313184469938279, + "step": 1630, + "token_acc": 0.8454000441147678 + }, + { + "epoch": 0.37823423423423425, + "grad_norm": 128.35125732421875, + "learning_rate": 8.599558442598998e-07, + "loss": 6.106525802612305, + "loss_idm": 0.3390625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016943359375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.335712730884552, + "loss_sft": 0.014124685525894165, + "step": 1640, + "token_acc": 0.8472327403858708 + }, + { + "epoch": 0.38054054054054054, + "grad_norm": 115.0096664428711, + "learning_rate": 8.141676086873574e-07, + "loss": 6.469194030761718, + "loss_idm": 0.354296875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0168701171875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.35816329121589663, + "loss_sft": 0.014040127769112587, + "step": 1650, + "token_acc": 0.8477684774822888 + }, + { + "epoch": 0.38284684684684683, + "grad_norm": 118.82299041748047, + "learning_rate": 7.695237378953224e-07, + "loss": 6.429816436767578, + "loss_idm": 0.350390625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01669921875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3561391234397888, + "loss_sft": 0.014136588945984841, + "step": 1660, + "token_acc": 0.8471164172513124 + }, + { + "epoch": 0.3851531531531531, + "grad_norm": 137.1096954345703, + "learning_rate": 7.260364370723044e-07, + "loss": 6.1461631774902346, + "loss_idm": 0.333984375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0161376953125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3382636547088623, + "loss_sft": 0.01443476863205433, + "step": 1670, + "token_acc": 0.8449713322908712 + }, + { + "epoch": 0.3874594594594595, + "grad_norm": 136.0753631591797, + "learning_rate": 6.837175952121305e-07, + "loss": 7.131587219238281, + "loss_idm": 0.39453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01707763671875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3994451344013214, + "loss_sft": 0.014256933704018593, + "step": 1680, + "token_acc": 0.8466482946884202 + }, + { + "epoch": 0.38976576576576577, + "grad_norm": 97.39081573486328, + "learning_rate": 6.425787818636131e-07, + "loss": 5.294178009033203, + "loss_idm": 0.28515625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.014306640625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.2873238503932953, + "loss_sft": 0.014306297898292542, + "step": 1690, + "token_acc": 0.8460799868429963 + }, + { + "epoch": 0.39207207207207206, + "grad_norm": 125.00518798828125, + "learning_rate": 6.026312439675553e-07, + "loss": 5.605047225952148, + "loss_idm": 0.305078125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01572265625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3053286850452423, + "loss_sft": 0.01424778886139393, + "step": 1700, + "token_acc": 0.8460457677165354 + }, + { + "epoch": 0.39437837837837836, + "grad_norm": 159.38278198242188, + "learning_rate": 5.63885902781941e-07, + "loss": 5.934845733642578, + "loss_idm": 0.323046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0159423828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3253439128398895, + "loss_sft": 0.014488279446959495, + "step": 1710, + "token_acc": 0.8434498032190264 + }, + { + "epoch": 0.3966846846846847, + "grad_norm": 130.9042205810547, + "learning_rate": 5.263533508961827e-07, + "loss": 5.979878997802734, + "loss_idm": 0.326953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.015283203125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3295415222644806, + "loss_sft": 0.014136172831058502, + "step": 1720, + "token_acc": 0.847153387628129 + }, + { + "epoch": 0.398990990990991, + "grad_norm": 219.98841857910156, + "learning_rate": 4.900438493352056e-07, + "loss": 6.236768341064453, + "loss_idm": 0.33984375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016357421875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.34439743757247926, + "loss_sft": 0.014121268689632416, + "step": 1730, + "token_acc": 0.847836894722829 + }, + { + "epoch": 0.4012972972972973, + "grad_norm": 94.14291381835938, + "learning_rate": 4.549673247541875e-07, + "loss": 5.808061981201172, + "loss_idm": 0.31328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0172607421875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.31636481285095214, + "loss_sft": 0.014235127717256546, + "step": 1740, + "token_acc": 0.8463776556318166 + }, + { + "epoch": 0.4036036036036036, + "grad_norm": 141.72726440429688, + "learning_rate": 4.211333667247125e-07, + "loss": 5.389980316162109, + "loss_idm": 0.2953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.015869140625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.2917319655418396, + "loss_sft": 0.014310146868228912, + "step": 1750, + "token_acc": 0.8454241937340762 + }, + { + "epoch": 0.4059099099099099, + "grad_norm": 99.67980194091797, + "learning_rate": 3.885512251130763e-07, + "loss": 5.464812469482422, + "loss_idm": 0.294921875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.015234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.2968344271183014, + "loss_sft": 0.014310889691114426, + "step": 1760, + "token_acc": 0.8457676402329125 + }, + { + "epoch": 0.40821621621621623, + "grad_norm": 111.7325668334961, + "learning_rate": 3.572298075514652e-07, + "loss": 5.1314746856689455, + "loss_idm": 0.276171875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.015869140625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.275756174325943, + "loss_sft": 0.014253809675574302, + "step": 1770, + "token_acc": 0.8462664212015208 + }, + { + "epoch": 0.4105225225225225, + "grad_norm": 145.02139282226562, + "learning_rate": 3.271776770026963e-07, + "loss": 6.446458435058593, + "loss_idm": 0.352734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0159912109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.35726869106292725, + "loss_sft": 0.014391064271330833, + "step": 1780, + "token_acc": 0.8451213642302139 + }, + { + "epoch": 0.4128288288288288, + "grad_norm": 117.78968048095703, + "learning_rate": 2.984030494191942e-07, + "loss": 5.586392974853515, + "loss_idm": 0.303515625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.014208984375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.30569988489151, + "loss_sft": 0.01434679925441742, + "step": 1790, + "token_acc": 0.8450751532454446 + }, + { + "epoch": 0.4151351351351351, + "grad_norm": 172.14227294921875, + "learning_rate": 2.7091379149682683e-07, + "loss": 6.588227844238281, + "loss_idm": 0.364453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177734375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.365203458070755, + "loss_sft": 0.014121705666184425, + "step": 1800, + "token_acc": 0.847520602220616 + }, + { + "epoch": 0.41744144144144146, + "grad_norm": 117.77503967285156, + "learning_rate": 2.447174185242324e-07, + "loss": 6.192712020874024, + "loss_idm": 0.3421875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0173828125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.34057103395462035, + "loss_sft": 0.014220455661416054, + "step": 1810, + "token_acc": 0.8467520968098974 + }, + { + "epoch": 0.41974774774774776, + "grad_norm": 145.01576232910156, + "learning_rate": 2.198210923282118e-07, + "loss": 6.494686126708984, + "loss_idm": 0.3609375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0170654296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3598536133766174, + "loss_sft": 0.014018755406141281, + "step": 1820, + "token_acc": 0.8479920807445692 + }, + { + "epoch": 0.42205405405405405, + "grad_norm": 145.6577911376953, + "learning_rate": 1.962316193157593e-07, + "loss": 6.524070739746094, + "loss_idm": 0.3609375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01572265625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3622725188732147, + "loss_sft": 0.01448512151837349, + "step": 1830, + "token_acc": 0.8438664155522041 + }, + { + "epoch": 0.42436036036036034, + "grad_norm": 141.86961364746094, + "learning_rate": 1.7395544861325718e-07, + "loss": 5.844393920898438, + "loss_idm": 0.318359375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0162841796875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.31986252069473264, + "loss_sft": 0.014221835136413574, + "step": 1840, + "token_acc": 0.846834097995988 + }, + { + "epoch": 0.4266666666666667, + "grad_norm": 114.36815643310547, + "learning_rate": 1.5299867030334815e-07, + "loss": 6.8152618408203125, + "loss_idm": 0.376953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01689453125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3801488935947418, + "loss_sft": 0.014056991413235664, + "step": 1850, + "token_acc": 0.847726296440595 + }, + { + "epoch": 0.428972972972973, + "grad_norm": 118.01188659667969, + "learning_rate": 1.333670137599713e-07, + "loss": 5.871685791015625, + "loss_idm": 0.323046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016650390625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3209952533245087, + "loss_sft": 0.014230089634656906, + "step": 1860, + "token_acc": 0.8468013427269361 + }, + { + "epoch": 0.4312792792792793, + "grad_norm": 138.60671997070312, + "learning_rate": 1.1506584608200366e-07, + "loss": 6.1166221618652346, + "loss_idm": 0.334765625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01787109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3347655713558197, + "loss_sft": 0.014273864589631558, + "step": 1870, + "token_acc": 0.8467156937566779 + }, + { + "epoch": 0.4335855855855856, + "grad_norm": 119.0372314453125, + "learning_rate": 9.810017062595322e-08, + "loss": 5.563272094726562, + "loss_idm": 0.30234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.017333984375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3010917007923126, + "loss_sft": 0.014137681201100349, + "step": 1880, + "token_acc": 0.8471984635388708 + }, + { + "epoch": 0.43589189189189187, + "grad_norm": 109.44342041015625, + "learning_rate": 8.247462563808816e-08, + "loss": 6.801327514648437, + "loss_idm": 0.378125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0158203125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.37939977645874023, + "loss_sft": 0.014424377679824829, + "step": 1890, + "token_acc": 0.8440372646223568 + }, + { + "epoch": 0.4381981981981982, + "grad_norm": 108.73650360107422, + "learning_rate": 6.819348298638839e-08, + "loss": 6.3186492919921875, + "loss_idm": 0.348828125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0177490234375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3471773624420166, + "loss_sft": 0.014505642279982568, + "step": 1900, + "token_acc": 0.8439981339604765 + }, + { + "epoch": 0.4405045045045045, + "grad_norm": 115.30857849121094, + "learning_rate": 5.526064699265754e-08, + "loss": 5.467123413085938, + "loss_idm": 0.30078125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.016845703125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.29553488492965696, + "loss_sft": 0.014253704994916915, + "step": 1910, + "token_acc": 0.8457813861881992 + }, + { + "epoch": 0.4428108108108108, + "grad_norm": 121.6728286743164, + "learning_rate": 4.367965336512403e-08, + "loss": 6.475829315185547, + "loss_idm": 0.35703125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0151123046875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3601944029331207, + "loss_sft": 0.014359761029481888, + "step": 1920, + "token_acc": 0.8448007428830144 + }, + { + "epoch": 0.4451171171171171, + "grad_norm": 146.66650390625, + "learning_rate": 3.345366823180929e-08, + "loss": 5.640755081176758, + "loss_idm": 0.30625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0173095703125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3060153484344482, + "loss_sft": 0.014264202490448952, + "step": 1930, + "token_acc": 0.8459558061543763 + }, + { + "epoch": 0.44742342342342345, + "grad_norm": 171.3626251220703, + "learning_rate": 2.4585487274942922e-08, + "loss": 6.090188980102539, + "loss_idm": 0.331640625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01572265625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.33536694645881654, + "loss_sft": 0.014434127882122994, + "step": 1940, + "token_acc": 0.8448119228601985 + }, + { + "epoch": 0.44972972972972974, + "grad_norm": 142.61944580078125, + "learning_rate": 1.7077534966650767e-08, + "loss": 6.4600830078125, + "loss_idm": 0.353125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0164306640625, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3574811339378357, + "loss_sft": 0.01442413292825222, + "step": 1950, + "token_acc": 0.8443659020218682 + }, + { + "epoch": 0.45203603603603604, + "grad_norm": 189.20147705078125, + "learning_rate": 1.0931863906127327e-08, + "loss": 6.146803283691407, + "loss_idm": 0.337109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0154296875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3394834816455841, + "loss_sft": 0.014248322322964668, + "step": 1960, + "token_acc": 0.846120984278879 + }, + { + "epoch": 0.45434234234234233, + "grad_norm": 101.11814880371094, + "learning_rate": 6.150154258476315e-09, + "loss": 6.392897033691407, + "loss_idm": 0.34765625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0174560546875, + "loss_idm_text": 0.0, + "loss_idm_type": 0.35239015221595765, + "loss_sft": 0.014375448971986771, + "step": 1970, + "token_acc": 0.8447592518506186 + }, + { + "epoch": 0.4566486486486486, + "grad_norm": 126.89629364013672, + "learning_rate": 2.7337132953697555e-09, + "loss": 5.897433853149414, + "loss_idm": 0.318359375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0161376953125, + "loss_idm_text": 0.0, + "loss_idm_type": 0.3226142168045044, + "loss_sft": 0.014501743763685227, + "step": 1980, + "token_acc": 0.8442813153244006 + }, + { + "epoch": 0.458954954954955, + "grad_norm": 149.4822998046875, + "learning_rate": 6.834750376549793e-10, + "loss": 5.743742752075195, + "loss_idm": 0.3140625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0159912109375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.31345025897026063, + "loss_sft": 0.014464382827281953, + "step": 1990, + "token_acc": 0.8448808726071968 + }, + { + "epoch": 0.46126126126126127, + "grad_norm": 147.2666473388672, + "learning_rate": 0.0, + "loss": 6.437743377685547, + "loss_idm": 0.35078125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01507568359375, + "loss_idm_text": 0.0, + "loss_idm_type": 0.35724995136260984, + "loss_sft": 0.014660027623176575, + "step": 2000, + "token_acc": 0.8419923088632739 + }, + { + "epoch": 0.46126126126126127, + "eval_loss": 0.7817181944847107, + "eval_runtime": 237.6265, + "eval_samples_per_second": 7.343, + "eval_steps_per_second": 1.839, + "eval_token_acc": 0.8335030896639121, + "step": 2000 + } + ], + "logging_steps": 10, + "max_steps": 2000, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 2.5018639729226154e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/training_args.bin b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..45af7236f9876dbcf3ea319b060e84035fd55948 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10959bc6c05fa08bc09a65c47dfe805dc395511f5874237c6b74d928524bf80c +size 9041 diff --git a/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/zero_to_fp32.py b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694 --- /dev/null +++ b/qwen-cua__runs__probe2000_idm_code_coord__v5-20260622-101413__checkpoint-2000/zero_to_fp32.py @@ -0,0 +1,760 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: +# python zero_to_fp32.py . output_dir/ +# or +# python zero_to_fp32.py . output_dir/ --safe_serialization + +import argparse +import torch +import glob +import math +import os +import re +import gc +import json +import numpy as np +from tqdm import tqdm +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device, weights_only=False) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + total_files = len(files) + state_dicts = [] + for f in tqdm(files, desc='Loading checkpoint shards'): + state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +class GatheredTensor: + """ + A pseudo tensor that collects partitioned weights. + It is more memory efficient when there are multiple groups. + """ + + def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape): + self.flat_groups = flat_groups + self.flat_groups_offset = flat_groups_offset + self.offset = offset + self.partitioned_numel = partitioned_numel + self.shape = shape + self.dtype = self.flat_groups[0][0].dtype + + def contiguous(self): + """ + Merge partitioned weights from flat_groups into a single tensor. + """ + end_idx = self.offset + self.partitioned_numel + world_size = len(self.flat_groups) + pad_flat_param_chunks = [] + + for rank_i in range(world_size): + # for each rank, we need to collect weights from related group/groups + flat_groups_at_rank_i = self.flat_groups[rank_i] + start_group_id = None + end_group_id = None + for group_id in range(len(self.flat_groups_offset)): + if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]: + start_group_id = group_id + if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]: + end_group_id = group_id + break + # collect weights from related group/groups + for group_id in range(start_group_id, end_group_id + 1): + flat_tensor = flat_groups_at_rank_i[group_id] + start_offset = self.offset - self.flat_groups_offset[group_id] + end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id] + pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset]) + + # collect weights from all ranks + pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0) + param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous() + return param + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size + + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]])) + for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'): + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # memory efficient tensor + tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape) + state_dict[name] = tensor + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def to_torch_tensor(state_dict, return_empty_tensor=False): + """ + Convert state_dict of GatheredTensor to torch tensor + """ + torch_state_dict = {} + converted_tensors = {} + for name, tensor in state_dict.items(): + tensor_id = id(tensor) + if tensor_id in converted_tensors: # shared tensors + shared_tensor = torch_state_dict[converted_tensors[tensor_id]] + torch_state_dict[name] = shared_tensor + else: + converted_tensors[tensor_id] = name + if return_empty_tensor: + torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype) + else: + torch_state_dict[name] = tensor.contiguous() + return torch_state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag=None, + exclude_frozen_parameters=False, + lazy_mode=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient. + Convert the pesduo tensor to torch tensor by ``.contiguous()`` + + Returns: + - pytorch ``state_dict`` + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + Note: the above usage may not work if your application doesn't have sufficient free CPU memory. + You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. Or you can load state_dict in lazy mode :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu + for name, lazy_tensor in state_dict.item(): + tensor = lazy_tensor.contiguous() # to cpu + print(name, tensor) + # del tensor to release memory if it no longer in use + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + if lazy_mode: + return state_dict + else: + return to_torch_tensor(state_dict) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, + output_dir, + max_shard_size="5GB", + safe_serialization=False, + tag=None, + exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_dir``: directory to the pytorch fp32 state_dict output files + - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB + - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`). + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + # Dependency pre-check + if safe_serialization: + try: + from safetensors.torch import save_file + except ImportError: + print('If you want to use `safe_serialization`, please `pip install safetensors`') + raise + if max_shard_size is not None: + try: + from huggingface_hub import split_torch_state_dict_into_shards + except ImportError: + print('If you want to use `max_shard_size`, please `pip install huggingface_hub`') + raise + + # Convert zero checkpoint to state_dict + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag, + exclude_frozen_parameters, + lazy_mode=True) + + # Shard the model if it is too big. + weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin" + if max_shard_size is not None: + filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors") + # an memory-efficient approach for sharding + empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True) + state_dict_split = split_torch_state_dict_into_shards(empty_state_dict, + filename_pattern=filename_pattern, + max_shard_size=max_shard_size) + else: + from collections import namedtuple + StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"]) + state_dict_split = StateDictSplit(is_sharded=False, + filename_to_tensors={weights_name: list(state_dict.keys())}) + + # Save the model by shard + os.makedirs(output_dir, exist_ok=True) + filename_to_tensors = state_dict_split.filename_to_tensors.items() + for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"): + shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors} + shard_state_dict = to_torch_tensor(shard_state_dict) + output_path = os.path.join(output_dir, shard_file) + if safe_serialization: + save_file(shard_state_dict, output_path, metadata={"format": "pt"}) + else: + torch.save(shard_state_dict, output_path) + # release the memory of current shard + for tensor_name in list(shard_state_dict.keys()): + del state_dict[tensor_name] + del shard_state_dict[tensor_name] + del shard_state_dict + gc.collect() + + # Save index if sharded + if state_dict_split.is_sharded: + index = { + "metadata": state_dict_split.metadata, + "weight_map": state_dict_split.tensor_to_filename, + } + save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json" + save_index_file = os.path.join(output_dir, save_index_file) + with open(save_index_file, "w", encoding="utf-8") as f: + content = json.dumps(index, indent=2, sort_keys=True) + "\n" + f.write(content) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info("Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info("Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument("output_dir", + type=str, + help="directory to the pytorch fp32 state_dict output files" + "(e.g. path/checkpoint-12-output/)") + parser.add_argument( + "--max_shard_size", + type=str, + default="5GB", + help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size" + "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`" + "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances" + "without CPU OOM issues.") + parser.add_argument( + "--safe_serialization", + default=False, + action='store_true', + help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_dir, + max_shard_size=args.max_shard_size, + safe_serialization=args.safe_serialization, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/args.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/args.json new file mode 100644 index 0000000000000000000000000000000000000000..3c5cdee99d0390fce2c39b3ced2d2b34eb9388e5 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/args.json @@ -0,0 +1,382 @@ +{ + "output_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe_idm_code_full/v0-20260620-203709", + "per_device_train_batch_size": 1, + "num_train_epochs": 3.0, + "max_steps": 500, + "learning_rate": 1e-05, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_steps": 30.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "optim_target_modules": null, + "gradient_accumulation_steps": 32, + "average_tokens_across_devices": true, + "max_grad_norm": 1.0, + "label_smoothing_factor": 0.0, + "bf16": true, + "fp16": false, + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": true, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "use_liger_kernel": false, + "liger_kernel_config": null, + "use_cache": false, + "neftune_noise_alpha": null, + "torch_empty_cache_steps": null, + "auto_find_batch_size": false, + "logging_strategy": "steps", + "logging_steps": 10, + "logging_first_step": true, + "log_on_each_node": true, + "logging_nan_inf_filter": true, + "include_num_input_tokens_seen": false, + "log_level": "passive", + "log_level_replica": "warning", + "disable_tqdm": null, + "report_to": [ + "wandb" + ], + "run_name": "probe_idm_code_full", + "project": "huggingface", + "trackio_space_id": null, + "trackio_bucket_id": null, + "trackio_static_space_id": null, + "eval_strategy": "steps", + "eval_steps": 500.0, + "eval_delay": 0, + "per_device_eval_batch_size": 1, + "prediction_loss_only": false, + "eval_on_start": false, + "eval_do_concat_batches": true, + "eval_use_gather_object": false, + "eval_accumulation_steps": null, + "include_for_metrics": [], + "batch_eval_metrics": false, + "save_only_model": false, + "save_strategy": "steps", + "save_steps": 100.0, + "save_on_each_node": false, + "save_total_limit": 3, + "enable_jit_checkpoint": false, + "push_to_hub": false, + "hub_token": null, + "hub_private_repo": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_always_push": false, + "hub_revision": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "restore_callback_states_from_checkpoint": false, + "full_determinism": false, + "seed": 42, + "data_seed": 42, + "use_cpu": false, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "dataloader_drop_last": false, + "dataloader_num_workers": 8, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 4, + "remove_unused_columns": true, + "label_names": null, + "train_sampling_strategy": "random", + "length_column_name": "length", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "ddp_static_graph": null, + "ddp_backend": null, + "ddp_timeout": 18000000, + "fsdp": [], + "fsdp_config": null, + "deepspeed": { + "bf16": { + "enabled": true + }, + "fp16": { + "enabled": false + }, + "zero_optimization": { + "stage": 2, + "overlap_comm": true, + "contiguous_gradients": true, + "reduce_bucket_size": 500000000.0, + "allgather_bucket_size": 500000000.0, + "reduce_scatter": true, + "round_robin_gradients": true + }, + "gradient_clipping": "auto", + "gradient_accumulation_steps": "auto", + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "steps_per_print": 100, + "wall_clock_breakdown": false, + "_comment_bf16_safety": "DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives." + }, + "debug": null, + "skip_memory_metrics": true, + "do_train": false, + "do_eval": false, + "do_predict": false, + "resume_from_checkpoint": null, + "warmup_ratio": null, + "logging_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe_idm_code_full/v0-20260620-203709/runs", + "local_rank": 0, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "group_by_length": false, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "model_type": "qwen3_5", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attention_2", + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_5", + "system": null, + "max_length": 12288, + "truncation_strategy": "delete", + "max_pixels": 16777216, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": false, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/ubuntu/train_wm.jsonl#35000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/win_mac/train_wm.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l1/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l2/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l3/train_wm.jsonl#50000" + ], + "val_dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/val_wm.jsonl" + ], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 4, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "use_hf": false, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual", + "model.visual.merger" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "rank": 0, + "global_world_size": 4, + "local_world_size": 4, + "model_suffix": "checkpoint-4336", + "model_info": "ModelInfo(model_type='qwen3_5', model_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling=None, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_5', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.5-0.8B', hf_model_id='Qwen/Qwen3.5-0.8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B', hf_model_id='Qwen/Qwen3.5-2B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B', hf_model_id='Qwen/Qwen3.5-4B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B', hf_model_id='Qwen/Qwen3.5-9B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B', hf_model_id='Qwen/Qwen3.5-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B-FP8', hf_model_id='Qwen/Qwen3.5-27B-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-0.8B-Base', hf_model_id='Qwen/Qwen3.5-0.8B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B-Base', hf_model_id='Qwen/Qwen3.5-2B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B-Base', hf_model_id='Qwen/Qwen3.5-4B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B-Base', hf_model_id='Qwen/Qwen3.5-9B-Base', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.6-27B', hf_model_id='Qwen/Qwen3.6-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.6-27B-FP8', hf_model_id='Qwen/Qwen3.6-27B-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[])], loader=, template=None, model_arch=MultiModelKeys(arch_name='qwen2_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3_5ForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=5.0.0.dev', 'qwen_vl_utils>=0.0.14', 'decord'], tags=['vision', 'video'])", + "model_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/ocu_ex_base_exact_all_12k_nopack/v0-20260612-174123/checkpoint-4336", + "template_meta": "QwenTemplateMeta(template_type='qwen3_5', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='qwen3_5', is_thinking=True, thinking_prefix='\\n', non_thinking_prefix='\\n\\n\\n\\n', history_thinking_prefix='')", + "_val_dataset_exists": true, + "hub": "", + "evaluation_strategy": "steps", + "training_args": "Seq2SeqTrainingArguments(output_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/probe_idm_code_full/v0-20260620-203709', per_device_train_batch_size=1, num_train_epochs=3.0, max_steps=500, learning_rate=1e-05, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_steps=30.0, optim=, optim_args=None, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=32, average_tokens_across_devices=None, max_grad_norm=1.0, label_smoothing_factor=0.0, bf16=True, fp16=False, bf16_full_eval=False, fp16_full_eval=False, tf32=True, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, use_liger_kernel=False, liger_kernel_config=None, use_cache=False, neftune_noise_alpha=None, torch_empty_cache_steps=None, auto_find_batch_size=False, logging_strategy=, logging_steps=10, logging_first_step=True, log_on_each_node=True, logging_nan_inf_filter=True, include_num_input_tokens_seen=None, log_level='passive', log_level_replica='warning', disable_tqdm=False, report_to=['wandb'], run_name='probe_idm_code_full', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=, eval_steps=500, eval_delay=0, per_device_eval_batch_size=1, prediction_loss_only=False, eval_on_start=False, eval_do_concat_batches=True, eval_use_gather_object=False, eval_accumulation_steps=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=, save_steps=100, save_on_each_node=False, save_total_limit=3, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=, hub_always_push=False, hub_revision=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, restore_callback_states_from_checkpoint=False, full_determinism=False, seed=42, data_seed=42, use_cpu=False, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, dataloader_drop_last=False, dataloader_num_workers=8, dataloader_pin_memory=True, dataloader_persistent_workers=True, dataloader_prefetch_factor=4, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, deepspeed={'bf16': {'enabled': True}, 'fp16': {'enabled': False}, 'zero_optimization': {'stage': 2, 'overlap_comm': True, 'contiguous_gradients': True, 'reduce_bucket_size': 500000000.0, 'allgather_bucket_size': 500000000.0, 'reduce_scatter': True, 'round_robin_gradients': True}, 'gradient_clipping': 'auto', 'gradient_accumulation_steps': 'auto', 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'steps_per_print': 100, 'wall_clock_breakdown': False, '_comment_bf16_safety': 'DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives.'}, debug=[], skip_memory_metrics=True, do_train=False, do_eval=True, do_predict=False, resume_from_checkpoint=None, warmup_ratio=None, logging_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/probe_idm_code_full/v0-20260620-203709/runs', local_rank=0, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, group_by_length=False, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/chat_template.jinja b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..a585dec894e63da457d9440ec6aa7caa16d20860 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/chat_template.jinja @@ -0,0 +1,154 @@ +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- macro render_content(content, do_vision_count, is_system_content=false) %} + {%- if content is string %} + {{- content }} + {%- elif content is iterable and content is not mapping %} + {%- for item in content %} + {%- if 'image' in item or 'image_url' in item or item.type == 'image' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain images.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set image_count.value = image_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Picture ' ~ image_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|image_pad|><|vision_end|>' }} + {%- elif 'video' in item or item.type == 'video' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain videos.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set video_count.value = video_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Video ' ~ video_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|video_pad|><|vision_end|>' }} + {%- elif 'text' in item %} + {{- item.text }} + {%- else %} + {{- raise_exception('Unexpected item type in content.') }} + {%- endif %} + {%- endfor %} + {%- elif content is none or content is undefined %} + {{- '' }} + {%- else %} + {{- raise_exception('Unexpected content type.') }} + {%- endif %} +{%- endmacro %} +{%- if not messages %} + {{- raise_exception('No messages provided.') }} +{%- endif %} +{%- if tools and tools is iterable and tools is not mapping %} + {{- '<|im_start|>system\n' }} + {{- "# Tools\n\nYou have access to the following functions:\n\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n" }} + {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {%- if content %} + {{- '\n\n' + content }} + {%- endif %} + {%- endif %} + {{- '<|im_end|>\n' }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" %} + {%- set content = render_content(message.content, false)|trim %} + {%- if not(content.startswith('') and content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if ns.multi_step_tool %} + {{- raise_exception('No user query found in messages.') }} +{%- endif %} +{%- for message in messages %} + {%- set content = render_content(message.content, true)|trim %} + {%- if message.role == "system" %} + {%- if not loop.first %} + {{- raise_exception('System message must be at the beginning.') }} + {%- endif %} + {%- elif message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = '' %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- set reasoning_content = reasoning_content|trim %} + {%- if loop.index0 > ns.last_query_index %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {%- if loop.first %} + {%- if content|trim %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n\n' }} + {%- endif %} + {%- else %} + {{- '\n\n\n' }} + {%- endif %} + {%- if tool_call.arguments is defined %} + {%- for args_name, args_value in tool_call.arguments|items %} + {{- '\n' }} + {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %} + {{- args_value }} + {{- '\n\n' }} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.previtem and loop.previtem.role != "tool" %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if not loop.last and loop.nextitem.role != "tool" %} + {{- '<|im_end|>\n' }} + {%- elif loop.last %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- else %} + {{- raise_exception('Unexpected message role.') }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n' }} + {%- endif %} +{%- endif %} \ No newline at end of file diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/config.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b1ade33543aaa3d299835283e0e20968262a8a0e --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/config.json @@ -0,0 +1,113 @@ +{ + "architectures": [ + "Qwen3_5ForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 248046, + "hidden_size": 2560, + "image_token_id": 248056, + "model_type": "qwen3_5", + "pad_token_id": 248044, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "attn_output_gate": true, + "bos_token_id": null, + "dtype": "bfloat16", + "eos_token_id": 248044, + "full_attention_interval": 4, + "head_dim": 256, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9216, + "layer_types": [ + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention" + ], + "linear_conv_kernel_dim": 4, + "linear_key_head_dim": 128, + "linear_num_key_heads": 16, + "linear_num_value_heads": 32, + "linear_value_head_dim": 128, + "mamba_ssm_dtype": "float32", + "max_position_embeddings": 262144, + "mlp_only_layers": [], + "model_type": "qwen3_5_text", + "mtp_num_hidden_layers": 1, + "mtp_use_dedicated_embeddings": false, + "num_attention_heads": 16, + "num_hidden_layers": 32, + "num_key_value_heads": 4, + "pad_token_id": 248044, + "partial_rotary_factor": 0.25, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "mrope_interleaved": true, + "mrope_section": [ + 11, + 11, + 10 + ], + "partial_rotary_factor": 0.25, + "rope_theta": 10000000, + "rope_type": "default" + }, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 248320 + }, + "tie_word_embeddings": true, + "transformers_version": "5.8.1", + "use_cache": false, + "video_token_id": 248057, + "vision_config": { + "deepstack_visual_indexes": [], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_5_vision", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 248054, + "vision_start_token_id": 248053 +} diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/generation_config.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1f1ab970fed81cb788c2945bed33833d247b9091 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "eos_token_id": [ + 248044, + 248046 + ], + "transformers_version": "5.8.1", + "use_cache": true +} diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..941a1e7a4c2758fc0394e627bb159acf64666597 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a2a6292e9e27fd50e06ccfadfe303401ecf2a14d6a72899f826553496b6fe4b +size 12675600417 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..62d1f9e0d613eae4c196e94d7f3ed13afa919be3 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91ad0527bc674a31e962c9c940a9277b8afc59864ba25cab263dce5bc7578b22 +size 12675605601 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..c16fb4b3a917cbbdac8ca63dbdaac1d2d3359222 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:948f457cafbd4c691f7f9cc98810caec01b630ec1376444eea239a521e6c14ab +size 12675605729 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..5c2ba0ba0a8d3964ca1f6e17d94e283557293e79 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3ba4c7d413ee8b3d61c63356a5ef117ad2b4795dda752db4e687ba43793fddf +size 12675604577 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/mp_rank_00_model_states.pt b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/mp_rank_00_model_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3774ba08df4c6bd13aeefddd1e77ceea6b69f14 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/global_step500/mp_rank_00_model_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b0ccf2b2493796bdf4fb4361ab9e25b471eeb8c93b056197d232aafb0588eb1 +size 9784824803 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/latest b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/latest new file mode 100644 index 0000000000000000000000000000000000000000..f0b47ce15fff9a01b2a416a473b2148085048a50 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/latest @@ -0,0 +1 @@ +global_step500 \ No newline at end of file diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00001-of-00003.safetensors b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a7bb419c12893f1b14170d3bbc2b052d3cfd92f1 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5e6ca91193052c6a1c4fcc61ebc45bf09917599e9e61344f7d96e87ffcc4b0c3 +size 4997083614 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00002-of-00003.safetensors b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..de41dda4175f5502b56b71df8d7fa3d52eb72160 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6a15980b3f3249f184ffc3cf60419ca9c6698cecc5fd604b9af37af7f00221aa +size 4993498424 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00003-of-00003.safetensors b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ab8023a740649342f1bc7e224320898de73054c --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9132529e66cac3079a828b70ff720888e735b62dcc8a839f7c42730ef896d42c +size 398327248 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model.safetensors.index.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..a138254f39ab95cffdb02f5f8564913a9c1cf4cb --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/model.safetensors.index.json @@ -0,0 +1,748 @@ +{ + "metadata": { + "total_parameters": 4558710159, + "total_size": 10388818718 + }, + "weight_map": { + "_idm_head.anl_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.anl_head.weight": "model-00001-of-00003.safetensors", + "_idm_head.coord_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.coord_head.weight": "model-00001-of-00003.safetensors", + "_idm_head.ln_next.bias": "model-00001-of-00003.safetensors", + "_idm_head.ln_next.weight": "model-00001-of-00003.safetensors", + "_idm_head.ln_state.bias": "model-00001-of-00003.safetensors", + "_idm_head.ln_state.weight": "model-00001-of-00003.safetensors", + "_idm_head.text_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.text_head.weight": "model-00001-of-00003.safetensors", + "_idm_head.trunk.0.bias": "model-00001-of-00003.safetensors", + "_idm_head.trunk.0.weight": "model-00001-of-00003.safetensors", + "_idm_head.trunk.2.bias": "model-00001-of-00003.safetensors", + "_idm_head.trunk.2.weight": "model-00001-of-00003.safetensors", + "_idm_head.type_head.bias": "model-00001-of-00003.safetensors", + "_idm_head.type_head.weight": "model-00001-of-00003.safetensors", + "lm_head.weight": "model-00001-of-00003.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00003.safetensors" + } +} diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/preprocessor_config.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/processor_config.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..33818c7f9e991ad735fd240209f4fa73e6c28c50 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/processor_config.json @@ -0,0 +1,60 @@ +{ + "image_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_0.pth b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..135dde79611a3dda98b1bd506a4f1d6b1b50dde2 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c425809aa076b2609bcf5f99afbf73a5f91a188ae8aab6150bc8282d69773bdd +size 15429 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_1.pth b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..a289a044b5cc8d717f223cc78a9b9fe3fef7e17c --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dda6ee8d221917fcb7c8e14c911ce427255df3b1fae8030edcdef8869b293cad +size 15429 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_2.pth b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..a7596bd36e77028450fde266968a047457a90461 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2588237e47dbb9a64d7996882e99fe405f58d29dacc1566191a9ab4fb05709c7 +size 15429 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_3.pth b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..1ee52a14b56538e98ef25dab6aa61e6f6fcf7e69 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1baa5871b9b10462c7f0f5f50db8115f45eb15f05e58d6a011790a83081a84ef +size 15429 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/scheduler.pt b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..a28f9a36d84e914a6dcd17d3354aaadecfe24f02 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:397b049f60afa9ebb2caf64ba5b428310a5e9d01385c6359e7192e85daef1fd8 +size 1465 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/tokenizer.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523 +size 19989325 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/tokenizer_config.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1d134cd298be1e3be25db393d93a1cefe80e3214 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/tokenizer_config.json @@ -0,0 +1,33 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "image_token": "<|image_pad|>", + "is_local": true, + "local_files_only": false, + "model_max_length": 262144, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +} diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/trainer_state.json b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..fe9003b1f357171924dd9739c43c52eccf65dd14 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/trainer_state.json @@ -0,0 +1,757 @@ +{ + "best_global_step": 500, + "best_metric": 1.27375484, + "best_model_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe_idm_code_full/v0-20260620-203709/checkpoint-500", + "epoch": 0.11531531531531532, + "eval_steps": 500, + "global_step": 500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00023063063063063062, + "grad_norm": 1592.4708251953125, + "learning_rate": 3.3333333333333335e-07, + "loss": 79.31526184082031, + "loss_idm": 4.953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0621337890625, + "loss_idm_text": 1.8660085201263428, + "loss_idm_type": 3.003211200237274, + "loss_sft": 0.013178251218050718, + "step": 1, + "token_acc": 0.855064946292334 + }, + { + "epoch": 0.0023063063063063064, + "grad_norm": 496.6201171875, + "learning_rate": 3.3333333333333333e-06, + "loss": 66.08669704861111, + "loss_idm": 4.048611111111111, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.060112847222222224, + "loss_idm_text": 1.7702105310228136, + "loss_idm_type": 2.2718035909864636, + "loss_sft": 0.013172145105070539, + "step": 10, + "token_acc": 0.8558691023074346 + }, + { + "epoch": 0.004612612612612613, + "grad_norm": 92.3178939819336, + "learning_rate": 6.666666666666667e-06, + "loss": 43.398077392578124, + "loss_idm": 2.525, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.053515625, + "loss_idm_text": 1.5655737400054932, + "loss_idm_type": 1.0649426221847533, + "loss_sft": 0.013627538271248341, + "step": 20, + "token_acc": 0.8530130423933372 + }, + { + "epoch": 0.006918918918918919, + "grad_norm": 194.6232147216797, + "learning_rate": 1e-05, + "loss": 36.92853698730469, + "loss_idm": 2.1125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.04853515625, + "loss_idm_text": 1.3756094217300414, + "loss_idm_type": 0.8548006534576416, + "loss_sft": 0.014491007849574088, + "step": 30, + "token_acc": 0.8464742555502733 + }, + { + "epoch": 0.009225225225225226, + "grad_norm": 399.7798156738281, + "learning_rate": 9.988834393115768e-06, + "loss": 31.787188720703124, + "loss_idm": 1.85, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0451171875, + "loss_idm_text": 1.170969557762146, + "loss_idm_type": 0.7399456858634949, + "loss_sft": 0.015156799182295799, + "step": 40, + "token_acc": 0.8414492656686929 + }, + { + "epoch": 0.011531531531531532, + "grad_norm": 298.2238464355469, + "learning_rate": 9.955387440773902e-06, + "loss": 28.66724853515625, + "loss_idm": 1.659375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.04189453125, + "loss_idm_text": 1.0696237325668334, + "loss_idm_type": 0.6480883717536926, + "loss_sft": 0.01556096114218235, + "step": 50, + "token_acc": 0.8379309061589034 + }, + { + "epoch": 0.013837837837837838, + "grad_norm": 212.14698791503906, + "learning_rate": 9.899808525182935e-06, + "loss": 24.614469909667967, + "loss_idm": 1.446875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.04228515625, + "loss_idm_text": 0.8591981172561646, + "loss_idm_type": 0.6060697555541992, + "loss_sft": 0.015211768448352814, + "step": 60, + "token_acc": 0.839478816922426 + }, + { + "epoch": 0.016144144144144144, + "grad_norm": 232.3778533935547, + "learning_rate": 9.822345875271884e-06, + "loss": 26.925689697265625, + "loss_idm": 1.5734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.03828125, + "loss_idm_text": 0.9588509798049927, + "loss_idm_type": 0.6548615455627441, + "loss_sft": 0.014741997048258782, + "step": 70, + "token_acc": 0.8432126657383905 + }, + { + "epoch": 0.01845045045045045, + "grad_norm": 120.58707427978516, + "learning_rate": 9.723345458039595e-06, + "loss": 23.952276611328124, + "loss_idm": 1.3875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0416015625, + "loss_idm_text": 0.7988368034362793, + "loss_idm_type": 0.6255433440208436, + "loss_sft": 0.014927534759044648, + "step": 80, + "token_acc": 0.8415277700129072 + }, + { + "epoch": 0.020756756756756756, + "grad_norm": 206.00711059570312, + "learning_rate": 9.603249433382145e-06, + "loss": 23.90917053222656, + "loss_idm": 1.3796875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.03984375, + "loss_idm_text": 0.7747604608535766, + "loss_idm_type": 0.6481955409049988, + "loss_sft": 0.015036704763770104, + "step": 90, + "token_acc": 0.8399600894415618 + }, + { + "epoch": 0.023063063063063063, + "grad_norm": 192.2452392578125, + "learning_rate": 9.462594179299408e-06, + "loss": 25.07174072265625, + "loss_idm": 1.453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.035791015625, + "loss_idm_text": 0.8922490239143371, + "loss_idm_type": 0.6073148608207702, + "loss_sft": 0.015319979563355447, + "step": 100, + "token_acc": 0.8373696958212777 + }, + { + "epoch": 0.025369369369369368, + "grad_norm": 187.56036376953125, + "learning_rate": 9.302007896300697e-06, + "loss": 24.611184692382814, + "loss_idm": 1.4046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.031884765625, + "loss_idm_text": 0.863019871711731, + "loss_idm_type": 0.6118855595588684, + "loss_sft": 0.01517961584031582, + "step": 110, + "token_acc": 0.8388810863092055 + }, + { + "epoch": 0.027675675675675675, + "grad_norm": 212.47923278808594, + "learning_rate": 9.122207801708802e-06, + "loss": 25.484585571289063, + "loss_idm": 1.4734375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.030322265625, + "loss_idm_text": 0.9250306487083435, + "loss_idm_type": 0.6061826467514038, + "loss_sft": 0.014974424988031388, + "step": 120, + "token_acc": 0.8410689656023232 + }, + { + "epoch": 0.029981981981981983, + "grad_norm": 207.89134216308594, + "learning_rate": 8.923996926393306e-06, + "loss": 24.694369506835937, + "loss_idm": 1.4328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0345703125, + "loss_idm_text": 0.8859843015670776, + "loss_idm_type": 0.5919388771057129, + "loss_sft": 0.014933220297098159, + "step": 130, + "token_acc": 0.841529192509836 + }, + { + "epoch": 0.03228828828828829, + "grad_norm": 161.8979034423828, + "learning_rate": 8.708260528239788e-06, + "loss": 23.232698059082033, + "loss_idm": 1.3390625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0294921875, + "loss_idm_text": 0.7606626987457276, + "loss_idm_type": 0.6311262965202331, + "loss_sft": 0.014705052971839905, + "step": 140, + "token_acc": 0.8432037780914124 + }, + { + "epoch": 0.034594594594594595, + "grad_norm": 196.3074951171875, + "learning_rate": 8.475962138373212e-06, + "loss": 25.088485717773438, + "loss_idm": 1.465625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.029248046875, + "loss_idm_text": 0.9114696979522705, + "loss_idm_type": 0.5967441916465759, + "loss_sft": 0.01474086418747902, + "step": 150, + "token_acc": 0.8424649816027868 + }, + { + "epoch": 0.0369009009009009, + "grad_norm": 201.341552734375, + "learning_rate": 8.228139257794012e-06, + "loss": 20.444810485839845, + "loss_idm": 1.1625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0277587890625, + "loss_idm_text": 0.6794226884841919, + "loss_idm_type": 0.5403245210647583, + "loss_sft": 0.014355449751019479, + "step": 160, + "token_acc": 0.8472049995079225 + }, + { + "epoch": 0.03920720720720721, + "grad_norm": 146.19137573242188, + "learning_rate": 7.965898723646777e-06, + "loss": 21.205398559570312, + "loss_idm": 1.2234375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.027685546875, + "loss_idm_text": 0.6731297075748444, + "loss_idm_type": 0.5949551463127136, + "loss_sft": 0.014598514139652252, + "step": 170, + "token_acc": 0.8441934355366634 + }, + { + "epoch": 0.04151351351351351, + "grad_norm": 253.85630798339844, + "learning_rate": 7.690411765816864e-06, + "loss": 21.725579833984376, + "loss_idm": 1.259375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0267578125, + "loss_idm_text": 0.710758113861084, + "loss_idm_type": 0.5906361103057861, + "loss_sft": 0.014163543283939362, + "step": 180, + "token_acc": 0.8481427317929219 + }, + { + "epoch": 0.04381981981981982, + "grad_norm": 202.72634887695312, + "learning_rate": 7.402908775933419e-06, + "loss": 20.416650390625, + "loss_idm": 1.1890625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0257568359375, + "loss_idm_text": 0.6582923889160156, + "loss_idm_type": 0.5625444531440735, + "loss_sft": 0.014225056767463684, + "step": 190, + "token_acc": 0.8473169056836616 + }, + { + "epoch": 0.04612612612612613, + "grad_norm": 140.08740234375, + "learning_rate": 7.104673812141676e-06, + "loss": 20.61493835449219, + "loss_idm": 1.2, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0271240234375, + "loss_idm_text": 0.6250565886497498, + "loss_idm_type": 0.6069200515747071, + "loss_sft": 0.014234534278512002, + "step": 200, + "token_acc": 0.8469481362428138 + }, + { + "epoch": 0.048432432432432435, + "grad_norm": 144.79013061523438, + "learning_rate": 6.797038864187564e-06, + "loss": 21.536167907714844, + "loss_idm": 1.2484375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0248046875, + "loss_idm_text": 0.7416723310947418, + "loss_idm_type": 0.5502459645271301, + "loss_sft": 0.014276015013456345, + "step": 210, + "token_acc": 0.8467435426120589 + }, + { + "epoch": 0.050738738738738735, + "grad_norm": 197.34327697753906, + "learning_rate": 6.481377904428171e-06, + "loss": 20.212065124511717, + "loss_idm": 1.1578125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.027734375, + "loss_idm_text": 0.6414934039115906, + "loss_idm_type": 0.5650732040405273, + "loss_sft": 0.013925277069211007, + "step": 220, + "token_acc": 0.8492174643744967 + }, + { + "epoch": 0.05304504504504504, + "grad_norm": 146.98313903808594, + "learning_rate": 6.1591007513376425e-06, + "loss": 21.818423461914062, + "loss_idm": 1.2765625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0260009765625, + "loss_idm_text": 0.7357961654663085, + "loss_idm_type": 0.5722585082054138, + "loss_sft": 0.013932183012366295, + "step": 230, + "token_acc": 0.8503047074580536 + }, + { + "epoch": 0.05535135135135135, + "grad_norm": 136.5176544189453, + "learning_rate": 5.831646772915651e-06, + "loss": 20.708409118652344, + "loss_idm": 1.2109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.025048828125, + "loss_idm_text": 0.7077920198440552, + "loss_idm_type": 0.5321911096572876, + "loss_sft": 0.014127027615904808, + "step": 240, + "token_acc": 0.8476971707089166 + }, + { + "epoch": 0.05765765765765766, + "grad_norm": 158.37416076660156, + "learning_rate": 5.500478458120493e-06, + "loss": 19.77899169921875, + "loss_idm": 1.146875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.024462890625, + "loss_idm_text": 0.6333002924919129, + "loss_idm_type": 0.5489766359329223, + "loss_sft": 0.01420053131878376, + "step": 250, + "token_acc": 0.8475135369069425 + }, + { + "epoch": 0.059963963963963966, + "grad_norm": 225.21864318847656, + "learning_rate": 5.1670748850383734e-06, + "loss": 19.577163696289062, + "loss_idm": 1.1265625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0234375, + "loss_idm_text": 0.6506115436553955, + "loss_idm_type": 0.5201167941093445, + "loss_sft": 0.01419382393360138, + "step": 260, + "token_acc": 0.847225027215274 + }, + { + "epoch": 0.06227027027027027, + "grad_norm": 150.79400634765625, + "learning_rate": 4.832925114961629e-06, + "loss": 19.28150634765625, + "loss_idm": 1.1046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0267578125, + "loss_idm_text": 0.6314839959144593, + "loss_idm_type": 0.5177404642105102, + "loss_sft": 0.01409110464155674, + "step": 270, + "token_acc": 0.8489457905342311 + }, + { + "epoch": 0.06457657657657657, + "grad_norm": 134.9580535888672, + "learning_rate": 4.499521541879508e-06, + "loss": 20.029010009765624, + "loss_idm": 1.153125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0255859375, + "loss_idm_text": 0.6679429650306702, + "loss_idm_type": 0.5287650108337403, + "loss_sft": 0.014084158837795258, + "step": 280, + "token_acc": 0.8486152363343102 + }, + { + "epoch": 0.06688288288288288, + "grad_norm": 126.87571716308594, + "learning_rate": 4.1683532270843505e-06, + "loss": 17.850544738769532, + "loss_idm": 1.0328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.02548828125, + "loss_idm_text": 0.5736021041870117, + "loss_idm_type": 0.4876852035522461, + "loss_sft": 0.014030075445771218, + "step": 290, + "token_acc": 0.8492478081663718 + }, + { + "epoch": 0.06918918918918919, + "grad_norm": 165.2994842529297, + "learning_rate": 3.840899248662358e-06, + "loss": 19.28271484375, + "loss_idm": 1.1203125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0249755859375, + "loss_idm_text": 0.6873706459999085, + "loss_idm_type": 0.4628427863121033, + "loss_sft": 0.014351583272218704, + "step": 300, + "token_acc": 0.8457350316602694 + }, + { + "epoch": 0.0714954954954955, + "grad_norm": 100.704833984375, + "learning_rate": 3.518622095571831e-06, + "loss": 18.965904235839844, + "loss_idm": 1.1015625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.025, + "loss_idm_text": 0.6294163584709167, + "loss_idm_type": 0.5026254177093505, + "loss_sft": 0.01375502087175846, + "step": 310, + "token_acc": 0.8516751274326874 + }, + { + "epoch": 0.0738018018018018, + "grad_norm": 145.5366973876953, + "learning_rate": 3.202961135812437e-06, + "loss": 17.057415771484376, + "loss_idm": 0.984375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.022412109375, + "loss_idm_text": 0.5393255949020386, + "loss_idm_type": 0.4753493547439575, + "loss_sft": 0.014044289290905, + "step": 320, + "token_acc": 0.8483403071743135 + }, + { + "epoch": 0.07610810810810811, + "grad_norm": 187.4034423828125, + "learning_rate": 2.8953261878583263e-06, + "loss": 20.268019104003905, + "loss_idm": 1.190625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0220703125, + "loss_idm_text": 0.7321646094322205, + "loss_idm_type": 0.4838695049285889, + "loss_sft": 0.013768799416720868, + "step": 330, + "token_acc": 0.8505789174331696 + }, + { + "epoch": 0.07841441441441442, + "grad_norm": 114.28669738769531, + "learning_rate": 2.5970912240665815e-06, + "loss": 20.306488037109375, + "loss_idm": 1.1859375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.022412109375, + "loss_idm_text": 0.6986854553222657, + "loss_idm_type": 0.5198426246643066, + "loss_sft": 0.01369791030883789, + "step": 340, + "token_acc": 0.8512502853816839 + }, + { + "epoch": 0.08072072072072072, + "grad_norm": 140.67515563964844, + "learning_rate": 2.309588234183137e-06, + "loss": 19.12022247314453, + "loss_idm": 1.1046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0212890625, + "loss_idm_text": 0.6607465267181396, + "loss_idm_type": 0.48446189761161806, + "loss_sft": 0.013866173475980759, + "step": 350, + "token_acc": 0.84998152334058 + }, + { + "epoch": 0.08302702702702702, + "grad_norm": 233.16070556640625, + "learning_rate": 2.0341012763532243e-06, + "loss": 17.8349365234375, + "loss_idm": 1.0328125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0210693359375, + "loss_idm_text": 0.6012717366218567, + "loss_idm_type": 0.46430001258850095, + "loss_sft": 0.013462393172085285, + "step": 360, + "token_acc": 0.8536638642293931 + }, + { + "epoch": 0.08533333333333333, + "grad_norm": 126.32984924316406, + "learning_rate": 1.771860742205988e-06, + "loss": 17.884495544433594, + "loss_idm": 1.0453125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.020849609375, + "loss_idm_text": 0.6073294937610626, + "loss_idm_type": 0.4610201001167297, + "loss_sft": 0.013804734125733376, + "step": 370, + "token_acc": 0.8500316903125152 + }, + { + "epoch": 0.08763963963963964, + "grad_norm": 137.39288330078125, + "learning_rate": 1.5240378616267887e-06, + "loss": 18.990934753417967, + "loss_idm": 1.0953125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.021435546875, + "loss_idm_text": 0.6517799496650696, + "loss_idm_type": 0.485338294506073, + "loss_sft": 0.013747426494956017, + "step": 380, + "token_acc": 0.8511524744489672 + }, + { + "epoch": 0.08994594594594595, + "grad_norm": 157.7312469482422, + "learning_rate": 1.2917394717602123e-06, + "loss": 17.10843505859375, + "loss_idm": 0.9875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0224609375, + "loss_idm_text": 0.591307520866394, + "loss_idm_type": 0.4277715265750885, + "loss_sft": 0.013545073568820953, + "step": 390, + "token_acc": 0.8528315369201633 + }, + { + "epoch": 0.09225225225225225, + "grad_norm": 176.7089385986328, + "learning_rate": 1.0760030736066952e-06, + "loss": 18.193414306640626, + "loss_idm": 1.046875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0213134765625, + "loss_idm_text": 0.635960602760315, + "loss_idm_type": 0.45176085233688357, + "loss_sft": 0.013588648289442062, + "step": 400, + "token_acc": 0.8530703600449623 + }, + { + "epoch": 0.09455855855855856, + "grad_norm": 133.8060760498047, + "learning_rate": 8.777921982911996e-07, + "loss": 18.297395324707033, + "loss_idm": 1.0578125, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0218994140625, + "loss_idm_text": 0.6218137860298156, + "loss_idm_type": 0.4721128821372986, + "loss_sft": 0.013110358640551567, + "step": 410, + "token_acc": 0.8572929395285072 + }, + { + "epoch": 0.09686486486486487, + "grad_norm": 136.15188598632812, + "learning_rate": 6.979921036993042e-07, + "loss": 18.296505737304688, + "loss_idm": 1.075, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0208984375, + "loss_idm_text": 0.6396299839019776, + "loss_idm_type": 0.454649555683136, + "loss_sft": 0.013779004663228988, + "step": 420, + "token_acc": 0.8511524995137133 + }, + { + "epoch": 0.09917117117117118, + "grad_norm": 156.84353637695312, + "learning_rate": 5.374058207005945e-07, + "loss": 17.45019073486328, + "loss_idm": 1.015625, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0217529296875, + "loss_idm_text": 0.5926790595054626, + "loss_idm_type": 0.4479467272758484, + "loss_sft": 0.013568659871816635, + "step": 430, + "token_acc": 0.8527494820441989 + }, + { + "epoch": 0.10147747747747747, + "grad_norm": 150.65768432617188, + "learning_rate": 3.9675056661785563e-07, + "loss": 19.32103271484375, + "loss_idm": 1.11875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0222412109375, + "loss_idm_text": 0.7090672850608826, + "loss_idm_type": 0.44831966161727904, + "loss_sft": 0.013689876347780228, + "step": 440, + "token_acc": 0.8510585113542408 + }, + { + "epoch": 0.10378378378378378, + "grad_norm": 123.24799346923828, + "learning_rate": 2.7665454196040665e-07, + "loss": 18.494871520996092, + "loss_idm": 1.06875, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.01982421875, + "loss_idm_text": 0.7104277849197388, + "loss_idm_type": 0.39781501293182375, + "loss_sft": 0.013484922796487808, + "step": 450, + "token_acc": 0.8533654843520049 + }, + { + "epoch": 0.10609009009009009, + "grad_norm": 186.10308837890625, + "learning_rate": 1.776541247281177e-07, + "loss": 17.906594848632814, + "loss_idm": 1.034375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0230712890625, + "loss_idm_text": 0.6649515867233277, + "loss_idm_type": 0.40319021344184874, + "loss_sft": 0.013544189929962158, + "step": 460, + "token_acc": 0.852078970984146 + }, + { + "epoch": 0.1083963963963964, + "grad_norm": 143.83282470703125, + "learning_rate": 1.0019147481706626e-07, + "loss": 19.552313232421874, + "loss_idm": 1.1375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.020166015625, + "loss_idm_text": 0.7164960980415345, + "loss_idm_type": 0.45704705119132993, + "loss_sft": 0.013577612675726413, + "step": 470, + "token_acc": 0.8530463974466188 + }, + { + "epoch": 0.1107027027027027, + "grad_norm": 167.7138671875, + "learning_rate": 4.461255922609986e-08, + "loss": 17.61268768310547, + "loss_idm": 1.025, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.020654296875, + "loss_idm_text": 0.6174419403076172, + "loss_idm_type": 0.4340633749961853, + "loss_sft": 0.013672563619911671, + "step": 480, + "token_acc": 0.8514405777283139 + }, + { + "epoch": 0.11300900900900901, + "grad_norm": 157.22537231445312, + "learning_rate": 1.1165606884234182e-08, + "loss": 18.91051330566406, + "loss_idm": 1.109375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0201416015625, + "loss_idm_text": 0.6843774914741516, + "loss_idm_type": 0.4488487482070923, + "loss_sft": 0.013685198128223419, + "step": 490, + "token_acc": 0.851507725655314 + }, + { + "epoch": 0.11531531531531532, + "grad_norm": 156.47488403320312, + "learning_rate": 0.0, + "loss": 18.66029052734375, + "loss_idm": 1.084375, + "loss_idm_anl": 0.0, + "loss_idm_coord": 0.0232177734375, + "loss_idm_text": 0.6769741654396058, + "loss_idm_type": 0.43823606967926027, + "loss_sft": 0.01359667256474495, + "step": 500, + "token_acc": 0.8519141550512672 + }, + { + "epoch": 0.11531531531531532, + "eval_loss": 1.2737548351287842, + "eval_runtime": 169.3982, + "eval_samples_per_second": 10.301, + "eval_steps_per_second": 2.58, + "eval_token_acc": 0.835355111491418, + "step": 500 + } + ], + "logging_steps": 10, + "max_steps": 500, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6.251724037081268e+18, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/training_args.bin b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..cd255967dce5ed68765de655f5d804e7e95450ee --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:185919749beef1744d3a367c95bd30c26d67842ec01833229165fd939ab2baf2 +size 8913 diff --git a/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/zero_to_fp32.py b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694 --- /dev/null +++ b/qwen-cua__runs__probe_idm_code_full__v0-20260620-203709__checkpoint-500/zero_to_fp32.py @@ -0,0 +1,760 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: +# python zero_to_fp32.py . output_dir/ +# or +# python zero_to_fp32.py . output_dir/ --safe_serialization + +import argparse +import torch +import glob +import math +import os +import re +import gc +import json +import numpy as np +from tqdm import tqdm +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device, weights_only=False) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + total_files = len(files) + state_dicts = [] + for f in tqdm(files, desc='Loading checkpoint shards'): + state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +class GatheredTensor: + """ + A pseudo tensor that collects partitioned weights. + It is more memory efficient when there are multiple groups. + """ + + def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape): + self.flat_groups = flat_groups + self.flat_groups_offset = flat_groups_offset + self.offset = offset + self.partitioned_numel = partitioned_numel + self.shape = shape + self.dtype = self.flat_groups[0][0].dtype + + def contiguous(self): + """ + Merge partitioned weights from flat_groups into a single tensor. + """ + end_idx = self.offset + self.partitioned_numel + world_size = len(self.flat_groups) + pad_flat_param_chunks = [] + + for rank_i in range(world_size): + # for each rank, we need to collect weights from related group/groups + flat_groups_at_rank_i = self.flat_groups[rank_i] + start_group_id = None + end_group_id = None + for group_id in range(len(self.flat_groups_offset)): + if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]: + start_group_id = group_id + if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]: + end_group_id = group_id + break + # collect weights from related group/groups + for group_id in range(start_group_id, end_group_id + 1): + flat_tensor = flat_groups_at_rank_i[group_id] + start_offset = self.offset - self.flat_groups_offset[group_id] + end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id] + pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset]) + + # collect weights from all ranks + pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0) + param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous() + return param + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size + + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]])) + for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'): + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # memory efficient tensor + tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape) + state_dict[name] = tensor + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def to_torch_tensor(state_dict, return_empty_tensor=False): + """ + Convert state_dict of GatheredTensor to torch tensor + """ + torch_state_dict = {} + converted_tensors = {} + for name, tensor in state_dict.items(): + tensor_id = id(tensor) + if tensor_id in converted_tensors: # shared tensors + shared_tensor = torch_state_dict[converted_tensors[tensor_id]] + torch_state_dict[name] = shared_tensor + else: + converted_tensors[tensor_id] = name + if return_empty_tensor: + torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype) + else: + torch_state_dict[name] = tensor.contiguous() + return torch_state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag=None, + exclude_frozen_parameters=False, + lazy_mode=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient. + Convert the pesduo tensor to torch tensor by ``.contiguous()`` + + Returns: + - pytorch ``state_dict`` + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + Note: the above usage may not work if your application doesn't have sufficient free CPU memory. + You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. Or you can load state_dict in lazy mode :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu + for name, lazy_tensor in state_dict.item(): + tensor = lazy_tensor.contiguous() # to cpu + print(name, tensor) + # del tensor to release memory if it no longer in use + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + if lazy_mode: + return state_dict + else: + return to_torch_tensor(state_dict) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, + output_dir, + max_shard_size="5GB", + safe_serialization=False, + tag=None, + exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_dir``: directory to the pytorch fp32 state_dict output files + - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB + - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`). + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + # Dependency pre-check + if safe_serialization: + try: + from safetensors.torch import save_file + except ImportError: + print('If you want to use `safe_serialization`, please `pip install safetensors`') + raise + if max_shard_size is not None: + try: + from huggingface_hub import split_torch_state_dict_into_shards + except ImportError: + print('If you want to use `max_shard_size`, please `pip install huggingface_hub`') + raise + + # Convert zero checkpoint to state_dict + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag, + exclude_frozen_parameters, + lazy_mode=True) + + # Shard the model if it is too big. + weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin" + if max_shard_size is not None: + filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors") + # an memory-efficient approach for sharding + empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True) + state_dict_split = split_torch_state_dict_into_shards(empty_state_dict, + filename_pattern=filename_pattern, + max_shard_size=max_shard_size) + else: + from collections import namedtuple + StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"]) + state_dict_split = StateDictSplit(is_sharded=False, + filename_to_tensors={weights_name: list(state_dict.keys())}) + + # Save the model by shard + os.makedirs(output_dir, exist_ok=True) + filename_to_tensors = state_dict_split.filename_to_tensors.items() + for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"): + shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors} + shard_state_dict = to_torch_tensor(shard_state_dict) + output_path = os.path.join(output_dir, shard_file) + if safe_serialization: + save_file(shard_state_dict, output_path, metadata={"format": "pt"}) + else: + torch.save(shard_state_dict, output_path) + # release the memory of current shard + for tensor_name in list(shard_state_dict.keys()): + del state_dict[tensor_name] + del shard_state_dict[tensor_name] + del shard_state_dict + gc.collect() + + # Save index if sharded + if state_dict_split.is_sharded: + index = { + "metadata": state_dict_split.metadata, + "weight_map": state_dict_split.tensor_to_filename, + } + save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json" + save_index_file = os.path.join(output_dir, save_index_file) + with open(save_index_file, "w", encoding="utf-8") as f: + content = json.dumps(index, indent=2, sort_keys=True) + "\n" + f.write(content) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info("Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info("Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument("output_dir", + type=str, + help="directory to the pytorch fp32 state_dict output files" + "(e.g. path/checkpoint-12-output/)") + parser.add_argument( + "--max_shard_size", + type=str, + default="5GB", + help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size" + "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`" + "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances" + "without CPU OOM issues.") + parser.add_argument( + "--safe_serialization", + default=False, + action='store_true', + help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_dir, + max_shard_size=args.max_shard_size, + safe_serialization=args.safe_serialization, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/args.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/args.json new file mode 100644 index 0000000000000000000000000000000000000000..65c8e7e80789999ae1f21c0384230b0a843d4a66 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/args.json @@ -0,0 +1,394 @@ +{ + "output_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_wm_phase1-20260529-173944/v0-20260529-174010", + "per_device_train_batch_size": 1, + "num_train_epochs": 1.0, + "max_steps": -1, + "learning_rate": 1e-05, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_steps": 200.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "optim_target_modules": null, + "gradient_accumulation_steps": 16, + "average_tokens_across_devices": true, + "max_grad_norm": 1.0, + "label_smoothing_factor": 0.0, + "bf16": true, + "fp16": false, + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": true, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "use_liger_kernel": false, + "liger_kernel_config": null, + "use_cache": false, + "neftune_noise_alpha": null, + "torch_empty_cache_steps": null, + "auto_find_batch_size": false, + "logging_strategy": "steps", + "logging_steps": 10, + "logging_first_step": true, + "log_on_each_node": true, + "logging_nan_inf_filter": true, + "include_num_input_tokens_seen": false, + "log_level": "passive", + "log_level_replica": "warning", + "disable_tqdm": null, + "report_to": [ + "wandb" + ], + "run_name": "qwen35_4b_wm_phase1-20260529-173944", + "project": "huggingface", + "trackio_space_id": null, + "trackio_bucket_id": null, + "trackio_static_space_id": null, + "eval_strategy": "steps", + "eval_steps": 500.0, + "eval_delay": 0, + "per_device_eval_batch_size": 1, + "prediction_loss_only": false, + "eval_on_start": false, + "eval_do_concat_batches": true, + "eval_use_gather_object": false, + "eval_accumulation_steps": null, + "include_for_metrics": [], + "batch_eval_metrics": false, + "save_only_model": false, + "save_strategy": "steps", + "save_steps": 1000.0, + "save_on_each_node": false, + "save_total_limit": 3, + "enable_jit_checkpoint": false, + "push_to_hub": false, + "hub_token": null, + "hub_private_repo": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_always_push": false, + "hub_revision": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "restore_callback_states_from_checkpoint": false, + "full_determinism": false, + "seed": 42, + "data_seed": 42, + "use_cpu": false, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "dataloader_drop_last": false, + "dataloader_num_workers": 8, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 4, + "remove_unused_columns": true, + "label_names": null, + "train_sampling_strategy": "random", + "length_column_name": "length", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "ddp_static_graph": null, + "ddp_backend": null, + "ddp_timeout": 18000000, + "fsdp": [], + "fsdp_config": null, + "deepspeed": { + "bf16": { + "enabled": true + }, + "fp16": { + "enabled": false + }, + "zero_optimization": { + "stage": 2, + "overlap_comm": true, + "contiguous_gradients": true, + "reduce_bucket_size": 500000000.0, + "allgather_bucket_size": 500000000.0, + "reduce_scatter": true, + "round_robin_gradients": true + }, + "gradient_clipping": "auto", + "gradient_accumulation_steps": "auto", + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "steps_per_print": 100, + "wall_clock_breakdown": false, + "_comment_bf16_safety": "DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives." + }, + "debug": null, + "skip_memory_metrics": true, + "do_train": false, + "do_eval": false, + "do_predict": false, + "resume_from_checkpoint": null, + "warmup_ratio": null, + "logging_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_wm_phase1-20260529-173944/v0-20260529-174010/runs", + "local_rank": 0, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "group_by_length": false, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", + "model_type": "qwen3_5", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attention_2", + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_5", + "system": null, + "max_length": 32768, + "truncation_strategy": "delete", + "max_pixels": 16777216, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": true, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/ubuntu/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/ubuntu/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l1/win_mac/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/win_mac/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l3/win_mac/train_wm.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l1/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l2/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GUI-360/l3/train_wm.jsonl#30000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/OS-Atlas/linux/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/OS-Atlas/macos/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/OS-Atlas/windows/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/ShowUI/desktop/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/ShowUI-web/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/GroundCUA/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/ScreenSpot/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/UGround/web_direct_150k_description_filtered/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/UGround/web_direct_258k_function_filtered/train.jsonl#50000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/UGround/web_hybrid_773k_max_25qa_filtered/train.jsonl#100000", + "/NHNHOME/data/sanghyeok/qwen-cua/data/Multimodal-Mind2Web/train.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/data/RICO-ScreenQA/train.jsonl#5000" + ], + "val_dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/data/AgentNet/l2/ubuntu/val_wm.jsonl" + ], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 4, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": true, + "packing_length": 32768, + "packing_num_proc": 1, + "lazy_tokenize": false, + "use_hf": false, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual", + "model.visual.merger" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3.5-4B", + "model_info": "ModelInfo(model_type='qwen3_5', model_dir='/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling=None, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_5', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.5-0.8B', hf_model_id='Qwen/Qwen3.5-0.8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B', hf_model_id='Qwen/Qwen3.5-2B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B', hf_model_id='Qwen/Qwen3.5-4B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B', hf_model_id='Qwen/Qwen3.5-9B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B', hf_model_id='Qwen/Qwen3.5-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B-FP8', hf_model_id='Qwen/Qwen3.5-27B-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-0.8B-Base', hf_model_id='Qwen/Qwen3.5-0.8B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B-Base', hf_model_id='Qwen/Qwen3.5-2B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B-Base', hf_model_id='Qwen/Qwen3.5-4B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B-Base', hf_model_id='Qwen/Qwen3.5-9B-Base', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.6-27B', hf_model_id='Qwen/Qwen3.6-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.6-27B-FP8', hf_model_id='Qwen/Qwen3.6-27B-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[])], loader=, template='qwen3_5', model_arch=MultiModelKeys(arch_name='qwen2_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3_5ForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=5.0.0.dev', 'qwen_vl_utils>=0.0.14', 'decord'], tags=[])", + "model_dir": "/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", + "template_meta": "QwenTemplateMeta(template_type='qwen3_5', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='qwen3_5', is_thinking=True, thinking_prefix='\\n', non_thinking_prefix='\\n\\n\\n\\n', history_thinking_prefix='')", + "_val_dataset_exists": true, + "hub": "", + "evaluation_strategy": "steps", + "training_args": "Seq2SeqTrainingArguments(output_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_wm_phase1-20260529-173944/v0-20260529-174010', per_device_train_batch_size=1, num_train_epochs=1.0, max_steps=-1, learning_rate=1e-05, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_steps=200.0, optim=, optim_args=None, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=16, average_tokens_across_devices=None, max_grad_norm=1.0, label_smoothing_factor=0.0, bf16=True, fp16=False, bf16_full_eval=False, fp16_full_eval=False, tf32=True, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, use_liger_kernel=False, liger_kernel_config=None, use_cache=False, neftune_noise_alpha=None, torch_empty_cache_steps=None, auto_find_batch_size=False, logging_strategy=, logging_steps=10, logging_first_step=True, log_on_each_node=True, logging_nan_inf_filter=True, include_num_input_tokens_seen=None, log_level='passive', log_level_replica='warning', disable_tqdm=False, report_to=['wandb'], run_name='qwen35_4b_wm_phase1-20260529-173944', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=, eval_steps=500, eval_delay=0, per_device_eval_batch_size=1, prediction_loss_only=False, eval_on_start=False, eval_do_concat_batches=True, eval_use_gather_object=False, eval_accumulation_steps=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=, save_steps=1000, save_on_each_node=False, save_total_limit=3, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=, hub_always_push=False, hub_revision=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, restore_callback_states_from_checkpoint=False, full_determinism=False, seed=42, data_seed=42, use_cpu=False, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, dataloader_drop_last=False, dataloader_num_workers=8, dataloader_pin_memory=True, dataloader_persistent_workers=True, dataloader_prefetch_factor=4, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, deepspeed={'bf16': {'enabled': True}, 'fp16': {'enabled': False}, 'zero_optimization': {'stage': 2, 'overlap_comm': True, 'contiguous_gradients': True, 'reduce_bucket_size': 500000000.0, 'allgather_bucket_size': 500000000.0, 'reduce_scatter': True, 'round_robin_gradients': True}, 'gradient_clipping': 'auto', 'gradient_accumulation_steps': 'auto', 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'steps_per_print': 100, 'wall_clock_breakdown': False, '_comment_bf16_safety': 'DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives.'}, debug=[], skip_memory_metrics=True, do_train=False, do_eval=True, do_predict=False, resume_from_checkpoint=None, warmup_ratio=None, logging_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_wm_phase1-20260529-173944/v0-20260529-174010/runs', local_rank=0, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, group_by_length=False, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/chat_template.jinja b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..a585dec894e63da457d9440ec6aa7caa16d20860 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/chat_template.jinja @@ -0,0 +1,154 @@ +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- macro render_content(content, do_vision_count, is_system_content=false) %} + {%- if content is string %} + {{- content }} + {%- elif content is iterable and content is not mapping %} + {%- for item in content %} + {%- if 'image' in item or 'image_url' in item or item.type == 'image' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain images.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set image_count.value = image_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Picture ' ~ image_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|image_pad|><|vision_end|>' }} + {%- elif 'video' in item or item.type == 'video' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain videos.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set video_count.value = video_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Video ' ~ video_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|video_pad|><|vision_end|>' }} + {%- elif 'text' in item %} + {{- item.text }} + {%- else %} + {{- raise_exception('Unexpected item type in content.') }} + {%- endif %} + {%- endfor %} + {%- elif content is none or content is undefined %} + {{- '' }} + {%- else %} + {{- raise_exception('Unexpected content type.') }} + {%- endif %} +{%- endmacro %} +{%- if not messages %} + {{- raise_exception('No messages provided.') }} +{%- endif %} +{%- if tools and tools is iterable and tools is not mapping %} + {{- '<|im_start|>system\n' }} + {{- "# Tools\n\nYou have access to the following functions:\n\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n" }} + {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {%- if content %} + {{- '\n\n' + content }} + {%- endif %} + {%- endif %} + {{- '<|im_end|>\n' }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" %} + {%- set content = render_content(message.content, false)|trim %} + {%- if not(content.startswith('') and content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if ns.multi_step_tool %} + {{- raise_exception('No user query found in messages.') }} +{%- endif %} +{%- for message in messages %} + {%- set content = render_content(message.content, true)|trim %} + {%- if message.role == "system" %} + {%- if not loop.first %} + {{- raise_exception('System message must be at the beginning.') }} + {%- endif %} + {%- elif message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = '' %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- set reasoning_content = reasoning_content|trim %} + {%- if loop.index0 > ns.last_query_index %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {%- if loop.first %} + {%- if content|trim %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n\n' }} + {%- endif %} + {%- else %} + {{- '\n\n\n' }} + {%- endif %} + {%- if tool_call.arguments is defined %} + {%- for args_name, args_value in tool_call.arguments|items %} + {{- '\n' }} + {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %} + {{- args_value }} + {{- '\n\n' }} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.previtem and loop.previtem.role != "tool" %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if not loop.last and loop.nextitem.role != "tool" %} + {{- '<|im_end|>\n' }} + {%- elif loop.last %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- else %} + {{- raise_exception('Unexpected message role.') }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n' }} + {%- endif %} +{%- endif %} \ No newline at end of file diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/config.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b1ade33543aaa3d299835283e0e20968262a8a0e --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/config.json @@ -0,0 +1,113 @@ +{ + "architectures": [ + "Qwen3_5ForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 248046, + "hidden_size": 2560, + "image_token_id": 248056, + "model_type": "qwen3_5", + "pad_token_id": 248044, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "attn_output_gate": true, + "bos_token_id": null, + "dtype": "bfloat16", + "eos_token_id": 248044, + "full_attention_interval": 4, + "head_dim": 256, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9216, + "layer_types": [ + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention" + ], + "linear_conv_kernel_dim": 4, + "linear_key_head_dim": 128, + "linear_num_key_heads": 16, + "linear_num_value_heads": 32, + "linear_value_head_dim": 128, + "mamba_ssm_dtype": "float32", + "max_position_embeddings": 262144, + "mlp_only_layers": [], + "model_type": "qwen3_5_text", + "mtp_num_hidden_layers": 1, + "mtp_use_dedicated_embeddings": false, + "num_attention_heads": 16, + "num_hidden_layers": 32, + "num_key_value_heads": 4, + "pad_token_id": 248044, + "partial_rotary_factor": 0.25, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "mrope_interleaved": true, + "mrope_section": [ + 11, + 11, + 10 + ], + "partial_rotary_factor": 0.25, + "rope_theta": 10000000, + "rope_type": "default" + }, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 248320 + }, + "tie_word_embeddings": true, + "transformers_version": "5.8.1", + "use_cache": false, + "video_token_id": 248057, + "vision_config": { + "deepstack_visual_indexes": [], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_5_vision", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 248054, + "vision_start_token_id": 248053 +} diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/generation_config.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1f1ab970fed81cb788c2945bed33833d247b9091 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "eos_token_id": [ + 248044, + 248046 + ], + "transformers_version": "5.8.1", + "use_cache": true +} diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..6c015488cd07552e3b3da2cf90d4a3bd777f0467 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9d7f97be6ac0353dd1933b26d0b93bbe2fb886454d411dbdc2911e6eee16bce3 +size 6330672305 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..35f1a4cfb979996ad5553d919b5fc47dc2147c72 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e407ef2da7c796622a73b701c8b134d0b4aff99290ba179b2dd79b0211cf5b96 +size 6330675505 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..d3667961bac24f676a0af5f20014226a37cd0208 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5ee2b78bd99e3d67ca2f0eb03873c44cf3acfc759d5e9f1fc0f8ce19e14ebb6a +size 6330676593 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..658dec82d6cfd019eb064f401101067c88c0c20a --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:873e46398c18578d10222cba37367d5d25f175bdab311a108cd8e3443d5b4a8b +size 6330676209 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..a7d597ef73bf8ee8e1e1fe1ecab672a5e8fc7af9 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e3b2f956a0f51fbc349bd90d98b5412b599b5958f218f933d0df0bdbda7172f +size 6330676273 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..427644627dc051d2e331272d8aa988db248f94b5 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d4b27bc9caac1de46f3fcfee39a65687f639d43e5c81cfa11753f7ad83ceab52 +size 6330676721 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..3b29824a3a80bf5ec2e9d696df36d4289b22227e --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c76f45b91bcdd0d1de0aef94905b601c37cdd00fdeb1077f3774ae9f69e58dc +size 6330676145 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..355a283e4a3099a03ce68eb789d0368fe4f95f91 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01b2b075bde9ada082ebea5bd006753b3b2f414461d8b2938a479f3e3fd2fb65 +size 6330675313 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/mp_rank_00_model_states.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/mp_rank_00_model_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..28347567b85004b987e31ff2df785217f440ad93 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/global_step494/mp_rank_00_model_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a4352031fee309cffc43d7e8b72f2585ed5ec23b5d30cd4f5291942988ac70dc +size 9775317667 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/latest b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/latest new file mode 100644 index 0000000000000000000000000000000000000000..f9eebeb526ba45ecc3cc9ddbd9a085e7b67f13f2 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/latest @@ -0,0 +1 @@ +global_step494 \ No newline at end of file diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00001-of-00003.safetensors b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f894d8d27f77ac5f414d2044e142e18fe3b485b6 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6daf811befee85788cafb3da71dc3148344b41845fd1e2c5b35c0a8431916ee +size 4987577312 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00002-of-00003.safetensors b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6075f60bc1546d2500e3cc9d3ceaed755935d679 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:088cce9203f6ad122198b53d2e8f99512358dea5b55eae12dcce077597e87c59 +size 4993498424 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00003-of-00003.safetensors b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ab8023a740649342f1bc7e224320898de73054c --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9132529e66cac3079a828b70ff720888e735b62dcc8a839f7c42730ef896d42c +size 398327248 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model.safetensors.index.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..fa5edcedf9204be7803616f27251bc852462e5f3 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/model.safetensors.index.json @@ -0,0 +1,740 @@ +{ + "metadata": { + "total_parameters": 4553957376, + "total_size": 10379313152 + }, + "weight_map": { + "_wm_projector.net.0.bias": "model-00001-of-00003.safetensors", + "_wm_projector.net.0.weight": "model-00001-of-00003.safetensors", + "_wm_projector.net.2.bias": "model-00001-of-00003.safetensors", + "_wm_projector.net.2.weight": "model-00001-of-00003.safetensors", + "_wm_projector.net.4.bias": "model-00001-of-00003.safetensors", + "_wm_projector.net.4.weight": "model-00001-of-00003.safetensors", + "_wm_projector.norm.bias": "model-00001-of-00003.safetensors", + "_wm_projector.norm.weight": "model-00001-of-00003.safetensors", + "lm_head.weight": "model-00001-of-00003.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00003.safetensors" + } +} diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/preprocessor_config.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/processor_config.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..33818c7f9e991ad735fd240209f4fa73e6c28c50 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/processor_config.json @@ -0,0 +1,60 @@ +{ + "image_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_0.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..f8e3e0bb161cdf9655fb1ca12575ddbca23bcdd8 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc1cb76231a5f4bb8aecfe0c732420d095e0e46bcd6480d8f61382e9842d79b6 +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_1.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..d105b5e5b5dafe3f8650ddae6a4d0904d28785cb --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2bd1684578e956e2a7ed1e7bfbc9ae431af9225476103f19a0f7a728807549b4 +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_2.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..42c7d397153b47c29b5e16018bc68ee6e908956c --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3942144659e2e608c3aa55404be9ae8461aaaefb850dd7dbc8f4c2a7be445fe +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_3.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..ca610b39916aacd8859f3b0cde59a0eb451243ba --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:204bae054fde6522015b429a01f3e09ef9f503b37a906ca81f2ddd4a12c65b30 +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_4.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_4.pth new file mode 100644 index 0000000000000000000000000000000000000000..af6915be1bf1b75b3dc15b4afaf4ce1fc803682f --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_4.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8206ee07ea38ec96536eb9dd30431a13fe96769c27402aefb6e8209756e25b70 +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_5.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_5.pth new file mode 100644 index 0000000000000000000000000000000000000000..93bd622ec4dc74b89bfa607eb00d87c21839677e --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_5.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fad1878e209cc9732947788539239ba11e930acd3a29eb32a68c8ee668424cd3 +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_6.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_6.pth new file mode 100644 index 0000000000000000000000000000000000000000..537bf3ec5faf95013e66c963d4d23e978292a877 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_6.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb002951cabc74c31a5779d8fb288c005415acc14236a51f7e636faf48d0b479 +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_7.pth b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_7.pth new file mode 100644 index 0000000000000000000000000000000000000000..aba80d7eefd11e0210c89d8836543e7ab68576e1 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/rng_state_7.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bedab2f5c48f8d95642ca282841b2079e2643f635cdd563c75b665a8da7891cc +size 16389 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/scheduler.pt b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..ae6bb8b6b62492893b77f294e9bdb3e78f8de450 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b2c132fcf249c22c6a14bab55d7a113e06d85840a51de6612190d1a41f30efc +size 1465 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/tokenizer.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523 +size 19989325 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/tokenizer_config.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1d134cd298be1e3be25db393d93a1cefe80e3214 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/tokenizer_config.json @@ -0,0 +1,33 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "image_token": "<|image_pad|>", + "is_local": true, + "local_files_only": false, + "model_max_length": 262144, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +} diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/trainer_state.json b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..b068ea2a52eba843d897cdf8751bd5da289f397a --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/trainer_state.json @@ -0,0 +1,545 @@ +{ + "best_global_step": 494, + "best_metric": 0.01240529, + "best_model_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/qwen35_4b_wm_phase1-20260529-173944/v0-20260529-174010/checkpoint-494", + "epoch": 1.0, + "eval_steps": 500, + "global_step": 494, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002027112631445585, + "grad_norm": 62.73832321166992, + "learning_rate": 5.0000000000000004e-08, + "loss": 12.4658203125, + "loss_sft": 0.07244362682104111, + "loss_wm": 1.0078125, + "step": 1, + "token_acc": 0.7370004425890262 + }, + { + "epoch": 0.020271126314455847, + "grad_norm": 63.18445587158203, + "learning_rate": 5.000000000000001e-07, + "loss": 12.50439453125, + "loss_sft": 0.04951130226254463, + "loss_wm": 0.99609375, + "step": 10, + "token_acc": 0.7407632974159035 + }, + { + "epoch": 0.040542252628911694, + "grad_norm": 62.66050720214844, + "learning_rate": 1.0000000000000002e-06, + "loss": 11.381005859375, + "loss_sft": 0.14802858233451843, + "loss_wm": 0.8125, + "step": 20, + "token_acc": 0.7422864806145153 + }, + { + "epoch": 0.06081337894336754, + "grad_norm": 30.287220001220703, + "learning_rate": 1.5e-06, + "loss": 9.300146484375, + "loss_sft": 0.01947697438299656, + "loss_wm": 0.65625, + "step": 30, + "token_acc": 0.7414682625149995 + }, + { + "epoch": 0.08108450525782339, + "grad_norm": 21.574708938598633, + "learning_rate": 2.0000000000000003e-06, + "loss": 7.708203125, + "loss_sft": 0.07248757779598236, + "loss_wm": 0.546875, + "step": 40, + "token_acc": 0.739062549778525 + }, + { + "epoch": 0.10135563157227924, + "grad_norm": 25.440834045410156, + "learning_rate": 2.5e-06, + "loss": 6.2340087890625, + "loss_sft": 0.06427811086177826, + "loss_wm": 0.484375, + "step": 50, + "token_acc": 0.7297968391562963 + }, + { + "epoch": 0.12162675788673508, + "grad_norm": 21.481849670410156, + "learning_rate": 3e-06, + "loss": 5.571240234375, + "loss_sft": 0.07587280869483948, + "loss_wm": 0.42578125, + "step": 60, + "token_acc": 0.7263029002205778 + }, + { + "epoch": 0.14189788420119093, + "grad_norm": 30.29012107849121, + "learning_rate": 3.5e-06, + "loss": 4.8462646484375, + "loss_sft": 0.09043002873659134, + "loss_wm": 0.359375, + "step": 70, + "token_acc": 0.7218637126610188 + }, + { + "epoch": 0.16216901051564678, + "grad_norm": 44.21630096435547, + "learning_rate": 4.000000000000001e-06, + "loss": 4.178466796875, + "loss_sft": 0.0188987385481596, + "loss_wm": 0.0, + "step": 80, + "token_acc": 0.7206794664033481 + }, + { + "epoch": 0.18244013683010263, + "grad_norm": 34.9310188293457, + "learning_rate": 4.5e-06, + "loss": 3.53758544921875, + "loss_sft": 0.08267555385828018, + "loss_wm": 0.251953125, + "step": 90, + "token_acc": 0.719468296546468 + }, + { + "epoch": 0.20271126314455848, + "grad_norm": 43.241310119628906, + "learning_rate": 5e-06, + "loss": 2.8167236328125, + "loss_sft": 0.040601231157779694, + "loss_wm": 0.19921875, + "step": 100, + "token_acc": 0.718706493823162 + }, + { + "epoch": 0.22298238945901433, + "grad_norm": 49.01982498168945, + "learning_rate": 5.500000000000001e-06, + "loss": 2.25465087890625, + "loss_sft": 0.05535919964313507, + "loss_wm": 0.154296875, + "step": 110, + "token_acc": 0.7181063176139241 + }, + { + "epoch": 0.24325351577347015, + "grad_norm": 57.312339782714844, + "learning_rate": 6e-06, + "loss": 1.757977294921875, + "loss_sft": 0.04172511398792267, + "loss_wm": 0.12109375, + "step": 120, + "token_acc": 0.7159847342063304 + }, + { + "epoch": 0.26352464208792603, + "grad_norm": 72.91925811767578, + "learning_rate": 6.5000000000000004e-06, + "loss": 1.217291259765625, + "loss_sft": 0.012581109069287777, + "loss_wm": 0.08056640625, + "step": 130, + "token_acc": 0.7134890125264517 + }, + { + "epoch": 0.28379576840238185, + "grad_norm": 65.72998809814453, + "learning_rate": 7e-06, + "loss": 0.8768341064453125, + "loss_sft": 0.03586875647306442, + "loss_wm": 0.05419921875, + "step": 140, + "token_acc": 0.7132199041360262 + }, + { + "epoch": 0.3040668947168377, + "grad_norm": 54.54481887817383, + "learning_rate": 7.500000000000001e-06, + "loss": 0.6329452514648437, + "loss_sft": 0.07855330407619476, + "loss_wm": 0.035888671875, + "step": 150, + "token_acc": 0.7154701897942395 + }, + { + "epoch": 0.32433802103129356, + "grad_norm": 57.35872268676758, + "learning_rate": 8.000000000000001e-06, + "loss": 0.4973175048828125, + "loss_sft": 0.0197242833673954, + "loss_wm": 0.0, + "step": 160, + "token_acc": 0.7134932036348094 + }, + { + "epoch": 0.3446091473457494, + "grad_norm": 45.244773864746094, + "learning_rate": 8.5e-06, + "loss": 0.46766204833984376, + "loss_sft": 0.04410327970981598, + "loss_wm": 0.0, + "step": 170, + "token_acc": 0.7143055232774049 + }, + { + "epoch": 0.36488027366020526, + "grad_norm": 33.39529037475586, + "learning_rate": 9e-06, + "loss": 0.41038970947265624, + "loss_sft": 0.015252730809152126, + "loss_wm": 0.024169921875, + "step": 180, + "token_acc": 0.712828237396499 + }, + { + "epoch": 0.3851513999746611, + "grad_norm": 42.51606750488281, + "learning_rate": 9.5e-06, + "loss": 0.47931976318359376, + "loss_sft": 0.04617125540971756, + "loss_wm": 0.03515625, + "step": 190, + "token_acc": 0.7115349725292256 + }, + { + "epoch": 0.40542252628911696, + "grad_norm": 33.25460433959961, + "learning_rate": 1e-05, + "loss": 0.4097625732421875, + "loss_sft": 0.019754070788621902, + "loss_wm": 0.0, + "step": 200, + "token_acc": 0.7131235679559028 + }, + { + "epoch": 0.4256936526035728, + "grad_norm": 29.123674392700195, + "learning_rate": 9.97148116317027e-06, + "loss": 0.36804046630859377, + "loss_sft": 0.03928299620747566, + "loss_wm": 0.0322265625, + "step": 210, + "token_acc": 0.7120108997890158 + }, + { + "epoch": 0.44596477891802866, + "grad_norm": 22.663774490356445, + "learning_rate": 9.88624998230272e-06, + "loss": 0.3521270751953125, + "loss_sft": 0.1824990212917328, + "loss_wm": 0.03173828125, + "step": 220, + "token_acc": 0.7098435193907559 + }, + { + "epoch": 0.4662359052324845, + "grad_norm": 29.16453742980957, + "learning_rate": 9.745278735053345e-06, + "loss": 0.3382591247558594, + "loss_sft": 0.052088283002376556, + "loss_wm": 0.031982421875, + "step": 230, + "token_acc": 0.7103755155993402 + }, + { + "epoch": 0.4865070315469403, + "grad_norm": 18.598217010498047, + "learning_rate": 9.550175555821333e-06, + "loss": 0.34694442749023435, + "loss_sft": 0.0895623192191124, + "loss_wm": 0.0257568359375, + "step": 240, + "token_acc": 0.7069384950431404 + }, + { + "epoch": 0.5067781578613961, + "grad_norm": 22.804227828979492, + "learning_rate": 9.303166090900082e-06, + "loss": 0.32968978881835936, + "loss_sft": 0.024845661595463753, + "loss_wm": 0.0, + "step": 250, + "token_acc": 0.7053896365418665 + }, + { + "epoch": 0.5270492841758521, + "grad_norm": 18.255435943603516, + "learning_rate": 9.007068109339783e-06, + "loss": 0.3148284912109375, + "loss_sft": 0.07574786990880966, + "loss_wm": 0.0281982421875, + "step": 260, + "token_acc": 0.7006531161430265 + }, + { + "epoch": 0.5473204104903079, + "grad_norm": 14.750127792358398, + "learning_rate": 8.665259359149132e-06, + "loss": 0.3146209716796875, + "loss_sft": 0.16314981877803802, + "loss_wm": 0.025390625, + "step": 270, + "token_acc": 0.6974019306937929 + }, + { + "epoch": 0.5675915368047637, + "grad_norm": 14.330487251281738, + "learning_rate": 8.281639035517591e-06, + "loss": 0.3019264221191406, + "loss_sft": 0.02875220775604248, + "loss_wm": 0.0, + "step": 280, + "token_acc": 0.6919131841283725 + }, + { + "epoch": 0.5878626631192195, + "grad_norm": 12.369770050048828, + "learning_rate": 7.860583300610849e-06, + "loss": 0.28978958129882815, + "loss_sft": 0.017795421183109283, + "loss_wm": 0.020263671875, + "step": 290, + "token_acc": 0.6912881230844372 + }, + { + "epoch": 0.6081337894336754, + "grad_norm": 9.962769508361816, + "learning_rate": 7.406895362348916e-06, + "loss": 0.28767547607421873, + "loss_sft": 0.057277485728263855, + "loss_wm": 0.0281982421875, + "step": 300, + "token_acc": 0.68837586379956 + }, + { + "epoch": 0.6284049157481313, + "grad_norm": 9.016581535339355, + "learning_rate": 6.925750681644954e-06, + "loss": 0.27554855346679685, + "loss_sft": 0.18838967382907867, + "loss_wm": 0.03564453125, + "step": 310, + "token_acc": 0.6840670976523872 + }, + { + "epoch": 0.6486760420625871, + "grad_norm": 11.221012115478516, + "learning_rate": 6.4226379331551625e-06, + "loss": 0.26334228515625, + "loss_sft": 0.0758930966258049, + "loss_wm": 0.018798828125, + "step": 320, + "token_acc": 0.6826579142529379 + }, + { + "epoch": 0.6689471683770429, + "grad_norm": 8.411371231079102, + "learning_rate": 5.903296393031996e-06, + "loss": 0.2581489562988281, + "loss_sft": 0.19933484494686127, + "loss_wm": 0.025146484375, + "step": 330, + "token_acc": 0.6797854826415043 + }, + { + "epoch": 0.6892182946914988, + "grad_norm": 6.862353324890137, + "learning_rate": 5.373650467932122e-06, + "loss": 0.252960205078125, + "loss_sft": 0.11295592784881592, + "loss_wm": 0.0140380859375, + "step": 340, + "token_acc": 0.6727570627079092 + }, + { + "epoch": 0.7094894210059547, + "grad_norm": 7.022729873657227, + "learning_rate": 4.839742112141725e-06, + "loss": 0.24704132080078126, + "loss_sft": 0.0666738897562027, + "loss_wm": 0.0179443359375, + "step": 350, + "token_acc": 0.6695278444010333 + }, + { + "epoch": 0.7297605473204105, + "grad_norm": 8.0825834274292, + "learning_rate": 4.307661903773129e-06, + "loss": 0.23574295043945312, + "loss_sft": 0.06745769083499908, + "loss_wm": 0.0, + "step": 360, + "token_acc": 0.665216758745826 + }, + { + "epoch": 0.7500316736348663, + "grad_norm": 6.737499237060547, + "learning_rate": 3.783479566283457e-06, + "loss": 0.23339004516601564, + "loss_sft": 0.08825641125440598, + "loss_wm": 0.025146484375, + "step": 370, + "token_acc": 0.6645291864987343 + }, + { + "epoch": 0.7703027999493222, + "grad_norm": 5.647401809692383, + "learning_rate": 3.273174727893463e-06, + "loss": 0.23535919189453125, + "loss_sft": 0.11529740691184998, + "loss_wm": 0.018798828125, + "step": 380, + "token_acc": 0.663425675849474 + }, + { + "epoch": 0.790573926263778, + "grad_norm": 3.912385940551758, + "learning_rate": 2.782568708770933e-06, + "loss": 0.23144989013671874, + "loss_sft": 0.08149220049381256, + "loss_wm": 0.0162353515625, + "step": 390, + "token_acc": 0.6631629762328183 + }, + { + "epoch": 0.8108450525782339, + "grad_norm": 2.9518821239471436, + "learning_rate": 2.317258114118686e-06, + "loss": 0.22767715454101561, + "loss_sft": 0.07110083103179932, + "loss_wm": 0.015625, + "step": 400, + "token_acc": 0.6567939574291515 + }, + { + "epoch": 0.8311161788926897, + "grad_norm": 2.9035332202911377, + "learning_rate": 1.8825509907063328e-06, + "loss": 0.21597061157226563, + "loss_sft": 0.08032238483428955, + "loss_wm": 0.0179443359375, + "step": 410, + "token_acc": 0.6536217036570595 + }, + { + "epoch": 0.8513873052071456, + "grad_norm": 2.425269365310669, + "learning_rate": 1.4834062751424018e-06, + "loss": 0.21766357421875, + "loss_sft": 0.10151892155408859, + "loss_wm": 0.01300048828125, + "step": 420, + "token_acc": 0.6554254770227169 + }, + { + "epoch": 0.8716584315216014, + "grad_norm": 2.32000470161438, + "learning_rate": 1.1243772246327416e-06, + "loss": 0.21241455078125, + "loss_sft": 0.07681550830602646, + "loss_wm": 0.02099609375, + "step": 430, + "token_acc": 0.6527156733493962 + }, + { + "epoch": 0.8919295578360573, + "grad_norm": 1.6833781003952026, + "learning_rate": 8.095594755407971e-07, + "loss": 0.2165496826171875, + "loss_sft": 0.06190862879157066, + "loss_wm": 0.020263671875, + "step": 440, + "token_acc": 0.6512058743676058 + }, + { + "epoch": 0.9122006841505131, + "grad_norm": 1.5111762285232544, + "learning_rate": 5.425443222735527e-07, + "loss": 0.21238632202148439, + "loss_sft": 0.09600041061639786, + "loss_wm": 0.01708984375, + "step": 450, + "token_acc": 0.6474550337431504 + }, + { + "epoch": 0.932471810464969, + "grad_norm": 1.3054075241088867, + "learning_rate": 3.263777494658449e-07, + "loss": 0.20197601318359376, + "loss_sft": 0.09694653004407883, + "loss_wm": 0.0107421875, + "step": 460, + "token_acc": 0.6500177115451552 + }, + { + "epoch": 0.9527429367794248, + "grad_norm": 1.002300500869751, + "learning_rate": 1.6352568480485277e-07, + "loss": 0.20771408081054688, + "loss_sft": 0.07341653108596802, + "loss_wm": 0.02294921875, + "step": 470, + "token_acc": 0.6479369963405088 + }, + { + "epoch": 0.9730140630938806, + "grad_norm": 0.816746175289154, + "learning_rate": 5.584586887435739e-08, + "loss": 0.20708198547363282, + "loss_sft": 0.1257670372724533, + "loss_wm": 0.019775390625, + "step": 480, + "token_acc": 0.6475466010868641 + }, + { + "epoch": 0.9932851894083365, + "grad_norm": 0.7419450283050537, + "learning_rate": 4.56666291450858e-09, + "loss": 0.20498123168945312, + "loss_sft": 0.07137016952037811, + "loss_wm": 0.0126953125, + "step": 490, + "token_acc": 0.6488176309237075 + }, + { + "epoch": 1.0, + "eval_loss": 0.012405285611748695, + "eval_runtime": 72.1636, + "eval_samples_per_second": 5.765, + "eval_steps_per_second": 0.721, + "eval_token_acc": 0.6923132349055924, + "loss_sft": 1.3594563007354736, + "loss_wm": 0.01483154296875, + "step": 494 + } + ], + "logging_steps": 10, + "max_steps": 494, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.8277099859373195e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/training_args.bin b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..84333b5f33ac81eaa3ba4317ffb0fe4c73b38df2 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e8f54e70435a3d4c7f49a8bd942b0e952061fb9fc7168f82f909b4480f48defb +size 8913 diff --git a/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/zero_to_fp32.py b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694 --- /dev/null +++ b/qwen-cua__runs__qwen35_4b_wm_phase1-20260529-173944__v0-20260529-174010__checkpoint-494/zero_to_fp32.py @@ -0,0 +1,760 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: +# python zero_to_fp32.py . output_dir/ +# or +# python zero_to_fp32.py . output_dir/ --safe_serialization + +import argparse +import torch +import glob +import math +import os +import re +import gc +import json +import numpy as np +from tqdm import tqdm +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device, weights_only=False) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + total_files = len(files) + state_dicts = [] + for f in tqdm(files, desc='Loading checkpoint shards'): + state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +class GatheredTensor: + """ + A pseudo tensor that collects partitioned weights. + It is more memory efficient when there are multiple groups. + """ + + def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape): + self.flat_groups = flat_groups + self.flat_groups_offset = flat_groups_offset + self.offset = offset + self.partitioned_numel = partitioned_numel + self.shape = shape + self.dtype = self.flat_groups[0][0].dtype + + def contiguous(self): + """ + Merge partitioned weights from flat_groups into a single tensor. + """ + end_idx = self.offset + self.partitioned_numel + world_size = len(self.flat_groups) + pad_flat_param_chunks = [] + + for rank_i in range(world_size): + # for each rank, we need to collect weights from related group/groups + flat_groups_at_rank_i = self.flat_groups[rank_i] + start_group_id = None + end_group_id = None + for group_id in range(len(self.flat_groups_offset)): + if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]: + start_group_id = group_id + if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]: + end_group_id = group_id + break + # collect weights from related group/groups + for group_id in range(start_group_id, end_group_id + 1): + flat_tensor = flat_groups_at_rank_i[group_id] + start_offset = self.offset - self.flat_groups_offset[group_id] + end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id] + pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset]) + + # collect weights from all ranks + pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0) + param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous() + return param + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size + + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]])) + for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'): + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # memory efficient tensor + tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape) + state_dict[name] = tensor + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def to_torch_tensor(state_dict, return_empty_tensor=False): + """ + Convert state_dict of GatheredTensor to torch tensor + """ + torch_state_dict = {} + converted_tensors = {} + for name, tensor in state_dict.items(): + tensor_id = id(tensor) + if tensor_id in converted_tensors: # shared tensors + shared_tensor = torch_state_dict[converted_tensors[tensor_id]] + torch_state_dict[name] = shared_tensor + else: + converted_tensors[tensor_id] = name + if return_empty_tensor: + torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype) + else: + torch_state_dict[name] = tensor.contiguous() + return torch_state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag=None, + exclude_frozen_parameters=False, + lazy_mode=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient. + Convert the pesduo tensor to torch tensor by ``.contiguous()`` + + Returns: + - pytorch ``state_dict`` + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + Note: the above usage may not work if your application doesn't have sufficient free CPU memory. + You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. Or you can load state_dict in lazy mode :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu + for name, lazy_tensor in state_dict.item(): + tensor = lazy_tensor.contiguous() # to cpu + print(name, tensor) + # del tensor to release memory if it no longer in use + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + if lazy_mode: + return state_dict + else: + return to_torch_tensor(state_dict) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, + output_dir, + max_shard_size="5GB", + safe_serialization=False, + tag=None, + exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_dir``: directory to the pytorch fp32 state_dict output files + - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB + - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`). + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + # Dependency pre-check + if safe_serialization: + try: + from safetensors.torch import save_file + except ImportError: + print('If you want to use `safe_serialization`, please `pip install safetensors`') + raise + if max_shard_size is not None: + try: + from huggingface_hub import split_torch_state_dict_into_shards + except ImportError: + print('If you want to use `max_shard_size`, please `pip install huggingface_hub`') + raise + + # Convert zero checkpoint to state_dict + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag, + exclude_frozen_parameters, + lazy_mode=True) + + # Shard the model if it is too big. + weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin" + if max_shard_size is not None: + filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors") + # an memory-efficient approach for sharding + empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True) + state_dict_split = split_torch_state_dict_into_shards(empty_state_dict, + filename_pattern=filename_pattern, + max_shard_size=max_shard_size) + else: + from collections import namedtuple + StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"]) + state_dict_split = StateDictSplit(is_sharded=False, + filename_to_tensors={weights_name: list(state_dict.keys())}) + + # Save the model by shard + os.makedirs(output_dir, exist_ok=True) + filename_to_tensors = state_dict_split.filename_to_tensors.items() + for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"): + shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors} + shard_state_dict = to_torch_tensor(shard_state_dict) + output_path = os.path.join(output_dir, shard_file) + if safe_serialization: + save_file(shard_state_dict, output_path, metadata={"format": "pt"}) + else: + torch.save(shard_state_dict, output_path) + # release the memory of current shard + for tensor_name in list(shard_state_dict.keys()): + del state_dict[tensor_name] + del shard_state_dict[tensor_name] + del shard_state_dict + gc.collect() + + # Save index if sharded + if state_dict_split.is_sharded: + index = { + "metadata": state_dict_split.metadata, + "weight_map": state_dict_split.tensor_to_filename, + } + save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json" + save_index_file = os.path.join(output_dir, save_index_file) + with open(save_index_file, "w", encoding="utf-8") as f: + content = json.dumps(index, indent=2, sort_keys=True) + "\n" + f.write(content) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info("Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info("Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument("output_dir", + type=str, + help="directory to the pytorch fp32 state_dict output files" + "(e.g. path/checkpoint-12-output/)") + parser.add_argument( + "--max_shard_size", + type=str, + default="5GB", + help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size" + "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`" + "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances" + "without CPU OOM issues.") + parser.add_argument( + "--safe_serialization", + default=False, + action='store_true', + help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_dir, + max_shard_size=args.max_shard_size, + safe_serialization=args.safe_serialization, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/args.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/args.json new file mode 100644 index 0000000000000000000000000000000000000000..15f7f6a06f3da9a368f2b1742eef31dd33b7f147 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/args.json @@ -0,0 +1,377 @@ +{ + "output_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/v6-recurtorch-recon-img3-hw2-k16-12k/v0-20260605-222754", + "per_device_train_batch_size": 1, + "num_train_epochs": 1.0, + "max_steps": -1, + "learning_rate": 1e-05, + "lr_scheduler_type": "cosine", + "lr_scheduler_kwargs": null, + "warmup_steps": 200.0, + "optim": "adamw_torch_fused", + "optim_args": null, + "weight_decay": 0.0, + "adam_beta1": 0.9, + "adam_beta2": 0.95, + "adam_epsilon": 1e-08, + "optim_target_modules": null, + "gradient_accumulation_steps": 16, + "average_tokens_across_devices": true, + "max_grad_norm": 1.0, + "label_smoothing_factor": 0.0, + "bf16": true, + "fp16": false, + "bf16_full_eval": false, + "fp16_full_eval": false, + "tf32": true, + "gradient_checkpointing": true, + "gradient_checkpointing_kwargs": null, + "torch_compile": false, + "torch_compile_backend": null, + "torch_compile_mode": null, + "use_liger_kernel": false, + "liger_kernel_config": null, + "use_cache": false, + "neftune_noise_alpha": null, + "torch_empty_cache_steps": null, + "auto_find_batch_size": false, + "logging_strategy": "steps", + "logging_steps": 1, + "logging_first_step": true, + "log_on_each_node": true, + "logging_nan_inf_filter": true, + "include_num_input_tokens_seen": false, + "log_level": "passive", + "log_level_replica": "warning", + "disable_tqdm": null, + "report_to": [ + "wandb" + ], + "run_name": "v6-recurtorch-recon-img3-hw2-k16-12k", + "project": "huggingface", + "trackio_space_id": null, + "trackio_bucket_id": null, + "trackio_static_space_id": null, + "eval_strategy": "steps", + "eval_steps": 100.0, + "eval_delay": 0, + "per_device_eval_batch_size": 1, + "prediction_loss_only": false, + "eval_on_start": false, + "eval_do_concat_batches": true, + "eval_use_gather_object": false, + "eval_accumulation_steps": null, + "include_for_metrics": [], + "batch_eval_metrics": false, + "save_only_model": false, + "save_strategy": "steps", + "save_steps": 100.0, + "save_on_each_node": false, + "save_total_limit": 3, + "enable_jit_checkpoint": false, + "push_to_hub": false, + "hub_token": null, + "hub_private_repo": null, + "hub_model_id": null, + "hub_strategy": "every_save", + "hub_always_push": false, + "hub_revision": null, + "load_best_model_at_end": false, + "metric_for_best_model": "loss", + "greater_is_better": false, + "ignore_data_skip": false, + "restore_callback_states_from_checkpoint": false, + "full_determinism": false, + "seed": 42, + "data_seed": 42, + "use_cpu": false, + "accelerator_config": { + "dispatch_batches": false + }, + "parallelism_config": null, + "dataloader_drop_last": false, + "dataloader_num_workers": 8, + "dataloader_pin_memory": true, + "dataloader_persistent_workers": true, + "dataloader_prefetch_factor": 4, + "remove_unused_columns": true, + "label_names": null, + "train_sampling_strategy": "random", + "length_column_name": "length", + "ddp_find_unused_parameters": null, + "ddp_bucket_cap_mb": null, + "ddp_broadcast_buffers": null, + "ddp_static_graph": null, + "ddp_backend": null, + "ddp_timeout": 18000000, + "fsdp": [], + "fsdp_config": null, + "deepspeed": { + "bf16": { + "enabled": true + }, + "fp16": { + "enabled": false + }, + "zero_optimization": { + "stage": 2, + "overlap_comm": true, + "contiguous_gradients": true, + "reduce_bucket_size": 500000000.0, + "allgather_bucket_size": 500000000.0, + "reduce_scatter": true, + "round_robin_gradients": true + }, + "gradient_clipping": "auto", + "gradient_accumulation_steps": "auto", + "train_batch_size": "auto", + "train_micro_batch_size_per_gpu": "auto", + "steps_per_print": 100, + "wall_clock_breakdown": false, + "_comment_bf16_safety": "DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives." + }, + "debug": null, + "skip_memory_metrics": true, + "do_train": false, + "do_eval": false, + "do_predict": false, + "resume_from_checkpoint": null, + "warmup_ratio": null, + "logging_dir": "/NHNHOME/data/sanghyeok/qwen-cua/runs/v6-recurtorch-recon-img3-hw2-k16-12k/v0-20260605-222754/runs", + "local_rank": 0, + "sortish_sampler": false, + "predict_with_generate": false, + "generation_max_length": null, + "generation_num_beams": null, + "generation_config": null, + "tuner_backend": "peft", + "vit_gradient_checkpointing": false, + "router_aux_loss_coef": 0.0, + "enable_dft_loss": false, + "enable_channel_loss": false, + "safe_serialization": true, + "max_shard_size": "5GB", + "check_model": true, + "acc_strategy": "token", + "train_dataloader_shuffle": true, + "group_by_length": false, + "max_epochs": null, + "aligner_lr": null, + "vit_lr": null, + "use_logits_to_keep": null, + "ds3_gather_for_generation": true, + "resume_only_model": false, + "optimizer": null, + "loss_type": null, + "eval_metric": null, + "callbacks": [], + "early_stop_interval": null, + "eval_use_evalscope": false, + "eval_dataset": [], + "eval_dataset_args": null, + "eval_limit": null, + "eval_generation_config": null, + "extra_eval_args": null, + "tuner_type": "full", + "use_galore": false, + "galore_target_modules": null, + "galore_rank": 128, + "galore_update_proj_gap": 50, + "galore_scale": 1.0, + "galore_proj_type": "std", + "galore_optim_per_parameter": false, + "galore_with_embedding": false, + "galore_quantization": false, + "galore_proj_quant": false, + "galore_proj_bits": 4, + "galore_proj_group_size": 256, + "galore_cos_threshold": 0.4, + "galore_gamma_proj": 2, + "galore_queue_size": 5, + "lisa_activated_layers": 0, + "lisa_step_interval": 20, + "use_flash_ckpt": false, + "use_ray": false, + "ray_exp_name": null, + "device_groups": null, + "model": "/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", + "model_type": "qwen3_5", + "model_revision": null, + "task_type": "causal_lm", + "torch_dtype": "bfloat16", + "attn_impl": "flash_attention_2", + "experts_impl": null, + "new_special_tokens": [], + "num_labels": null, + "problem_type": null, + "rope_scaling": null, + "device_map": null, + "max_memory": {}, + "max_model_len": null, + "local_repo_path": null, + "init_strategy": null, + "template": "qwen3_5", + "system": null, + "max_length": 12288, + "truncation_strategy": "delete", + "max_pixels": 16777216, + "agent_template": null, + "norm_bbox": null, + "use_chat_template": true, + "padding_side": "right", + "padding_free": false, + "loss_scale": "default", + "sequence_parallel_size": 1, + "template_backend": "swift", + "response_prefix": null, + "enable_thinking": null, + "add_non_thinking_prefix": true, + "dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/_traj_expanded_cache/train_traj_a91429cd18dd0934.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/_traj_expanded_cache/train_traj_93f9095470ba8031.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/_traj_expanded_cache/train_traj_b30d45f5dd2f85dd.jsonl", + "/NHNHOME/data/sanghyeok/qwen-cua/_traj_expanded_cache/train_traj_9382fa43e1278ec0.jsonl" + ], + "val_dataset": [ + "/NHNHOME/data/sanghyeok/qwen-cua/_traj_expanded_cache/val_traj_b4ad0016c2d07d39.jsonl" + ], + "cached_dataset": [], + "cached_val_dataset": [], + "split_dataset_ratio": 0.0, + "dataset_num_proc": 4, + "load_from_cache_file": false, + "dataset_shuffle": true, + "val_dataset_shuffle": false, + "streaming": false, + "interleave_prob": null, + "stopping_strategy": "first_exhausted", + "shuffle_buffer_size": 1000, + "download_mode": "reuse_dataset_if_exists", + "columns": {}, + "strict": false, + "disable_auto_column_mapping": false, + "model_name": null, + "model_author": null, + "custom_dataset_info": [], + "quant_method": null, + "quant_bits": null, + "hqq_axis": null, + "bnb_4bit_compute_dtype": "bfloat16", + "bnb_4bit_quant_type": "nf4", + "bnb_4bit_use_double_quant": true, + "bnb_4bit_quant_storage": null, + "max_new_tokens": 64, + "temperature": 0.0, + "top_k": null, + "top_p": null, + "repetition_penalty": null, + "num_beams": 1, + "stream": false, + "stop_words": [], + "logprobs": false, + "top_logprobs": null, + "structured_outputs_regex": null, + "adapters": [], + "external_plugins": [], + "custom_register_path": [], + "model_kwargs": {}, + "enable_npu_model_patch": true, + "load_args": false, + "load_data_args": false, + "packing": false, + "packing_length": null, + "packing_num_proc": 1, + "lazy_tokenize": true, + "use_hf": false, + "ignore_args_error": false, + "use_swift_lora": false, + "freeze_parameters": [ + "model.visual", + "model.visual.merger" + ], + "freeze_parameters_regex": null, + "freeze_parameters_ratio": 0.0, + "trainable_parameters": [], + "trainable_parameters_regex": null, + "freeze_llm": false, + "freeze_vit": true, + "freeze_aligner": true, + "target_modules": [ + "all-linear" + ], + "target_regex": null, + "target_parameters": null, + "modules_to_save": [], + "lora_rank": 8, + "lora_alpha": 32, + "lora_dropout": 0.05, + "lora_bias": "none", + "lora_dtype": null, + "lorap_lr_ratio": null, + "use_rslora": false, + "use_dora": false, + "lora_ga_batch_size": 2, + "lora_ga_iters": 2, + "lora_ga_max_length": 1024, + "lora_ga_direction": "ArB2r", + "lora_ga_scale": "stable", + "lora_ga_stable_gamma": 16, + "init_weights": true, + "fourier_n_frequency": 2000, + "fourier_scaling": 300.0, + "boft_block_size": 4, + "boft_block_num": 0, + "boft_n_butterfly_factor": 1, + "boft_dropout": 0.0, + "vera_rank": 256, + "vera_projection_prng_key": 0, + "vera_dropout": 0.0, + "vera_d_initial": 0.1, + "adapter_act": "gelu", + "adapter_length": 128, + "adalora_target_r": 8, + "adalora_init_r": 12, + "adalora_tinit": 0, + "adalora_tfinal": 0, + "adalora_deltaT": 1, + "adalora_beta1": 0.85, + "adalora_beta2": 0.85, + "adalora_orth_reg_weight": 0.5, + "llamapro_num_new_blocks": 4, + "llamapro_num_groups": null, + "reft_layer_key": null, + "reft_layers": null, + "reft_rank": 4, + "reft_intervention_type": "LoreftIntervention", + "reft_args": null, + "swanlab_token": null, + "swanlab_project": "ms-swift", + "swanlab_workspace": null, + "swanlab_exp_name": null, + "swanlab_notification_method": null, + "swanlab_webhook_url": null, + "swanlab_secret": null, + "swanlab_sender_email": null, + "swanlab_receiver_email": null, + "swanlab_smtp_server": null, + "swanlab_smtp_port": null, + "swanlab_email_language": "zh", + "swanlab_mode": "cloud", + "add_version": true, + "create_checkpoint_symlink": false, + "zero_hpz_partition_size": null, + "deepspeed_autotp_size": null, + "swift_version": "4.2.1", + "ckpt_dir": null, + "rank": 0, + "global_world_size": 8, + "local_world_size": 8, + "model_suffix": "Qwen3.5-4B", + "model_info": "ModelInfo(model_type='qwen3_5', model_dir='/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a', torch_dtype=torch.bfloat16, max_model_len=262144, quant_method=None, quant_bits=None, rope_scaling=None, is_moe_model=False, is_multimodal=True, config=None, task_type='causal_lm', num_labels=None)", + "model_meta": "ModelMeta(model_type='qwen3_5', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.5-0.8B', hf_model_id='Qwen/Qwen3.5-0.8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B', hf_model_id='Qwen/Qwen3.5-2B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B', hf_model_id='Qwen/Qwen3.5-4B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B', hf_model_id='Qwen/Qwen3.5-9B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B', hf_model_id='Qwen/Qwen3.5-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-27B-FP8', hf_model_id='Qwen/Qwen3.5-27B-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-0.8B-Base', hf_model_id='Qwen/Qwen3.5-0.8B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-2B-Base', hf_model_id='Qwen/Qwen3.5-2B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-4B-Base', hf_model_id='Qwen/Qwen3.5-4B-Base', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.5-9B-Base', hf_model_id='Qwen/Qwen3.5-9B-Base', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='Qwen/Qwen3.6-27B', hf_model_id='Qwen/Qwen3.6-27B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen3.6-27B-FP8', hf_model_id='Qwen/Qwen3.6-27B-FP8', model_path=None, ms_revision=None, hf_revision=None)], template='qwen3_5', ignore_patterns=None, requires=None, tags=[])], loader=, template='qwen3_5', model_arch=MultiModelKeys(arch_name='qwen2_vl', embedding=None, module_list=None, lm_head=None, q_proj=None, k_proj=None, v_proj=None, o_proj=None, attention=None, mlp=None, down_proj=None, qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None, language_model=['model.language_model', 'lm_head'], aligner=['model.visual.merger'], vision_tower=['model.visual'], generator=[]), mcore_model_type=None, architectures=['Qwen3_5ForConditionalGeneration'], additional_saved_files=[], torch_dtype=None, is_multimodal=True, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=5.0.0.dev', 'qwen_vl_utils>=0.0.14', 'decord'], tags=[])", + "model_dir": "/NHNHOME/data/sanghyeok/qwen-cua/hf_cache/hub/models--Qwen--Qwen3.5-4B/snapshots/851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", + "template_meta": "QwenTemplateMeta(template_type='qwen3_5', prefix=[], prompt=['<|im_start|>user\\n{{QUERY}}<|im_end|>\\n<|im_start|>assistant\\n'], chat_sep=['<|im_end|>\\n'], suffix=['<|im_end|>\\n'], template_cls=, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='qwen3_5', is_thinking=True, thinking_prefix='\\n', non_thinking_prefix='\\n\\n\\n\\n', history_thinking_prefix='')", + "_val_dataset_exists": true, + "hub": "", + "evaluation_strategy": "steps", + "training_args": "Seq2SeqTrainingArguments(output_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/v6-recurtorch-recon-img3-hw2-k16-12k/v0-20260605-222754', per_device_train_batch_size=1, num_train_epochs=1.0, max_steps=-1, learning_rate=1e-05, lr_scheduler_type=, lr_scheduler_kwargs=None, warmup_steps=200.0, optim=, optim_args=None, weight_decay=0.0, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=16, average_tokens_across_devices=None, max_grad_norm=1.0, label_smoothing_factor=0.0, bf16=True, fp16=False, bf16_full_eval=False, fp16_full_eval=False, tf32=True, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, use_liger_kernel=False, liger_kernel_config=None, use_cache=False, neftune_noise_alpha=None, torch_empty_cache_steps=None, auto_find_batch_size=False, logging_strategy=, logging_steps=1, logging_first_step=True, log_on_each_node=True, logging_nan_inf_filter=True, include_num_input_tokens_seen=None, log_level='passive', log_level_replica='warning', disable_tqdm=False, report_to=['wandb'], run_name='v6-recurtorch-recon-img3-hw2-k16-12k', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=, eval_steps=100, eval_delay=0, per_device_eval_batch_size=1, prediction_loss_only=False, eval_on_start=False, eval_do_concat_batches=True, eval_use_gather_object=False, eval_accumulation_steps=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=, save_steps=100, save_on_each_node=False, save_total_limit=3, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=, hub_always_push=False, hub_revision=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, restore_callback_states_from_checkpoint=False, full_determinism=False, seed=42, data_seed=42, use_cpu=False, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), parallelism_config=None, dataloader_drop_last=False, dataloader_num_workers=8, dataloader_pin_memory=True, dataloader_persistent_workers=True, dataloader_prefetch_factor=4, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, deepspeed={'bf16': {'enabled': True}, 'fp16': {'enabled': False}, 'zero_optimization': {'stage': 2, 'overlap_comm': True, 'contiguous_gradients': True, 'reduce_bucket_size': 500000000.0, 'allgather_bucket_size': 500000000.0, 'reduce_scatter': True, 'round_robin_gradients': True}, 'gradient_clipping': 'auto', 'gradient_accumulation_steps': 'auto', 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'steps_per_print': 100, 'wall_clock_breakdown': False, '_comment_bf16_safety': 'DeepSpeed 0.19 BF16_Optimizer keeps fp32 master grad partitions (fp32_groups_flat_partition + accumulate_hp_grads_and_remove_lp hook), so gradient accumulation across many micro-batches is done in fp32. bf16 is only used for the param master copy and forward/backward activations. This is safe at grad_accum=16+; no need to force fp32 collectives.'}, debug=[], skip_memory_metrics=True, do_train=False, do_eval=True, do_predict=False, resume_from_checkpoint=None, warmup_ratio=None, logging_dir='/NHNHOME/data/sanghyeok/qwen-cua/runs/v6-recurtorch-recon-img3-hw2-k16-12k/v0-20260605-222754/runs', local_rank=0, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, tuner_backend='peft', vit_gradient_checkpointing=False, router_aux_loss_coef=0.0, enable_dft_loss=False, enable_channel_loss=False, safe_serialization=True, max_shard_size='5GB', check_model=True, acc_strategy='token', train_dataloader_shuffle=True, group_by_length=False, max_epochs=None, aligner_lr=None, vit_lr=None, use_logits_to_keep=None, ds3_gather_for_generation=True, resume_only_model=False, optimizer=None, loss_type=None, eval_metric=None, callbacks=[], early_stop_interval=None, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, extra_eval_args=None, tuner_type='full', use_galore=False, galore_target_modules=None, galore_rank=128, galore_update_proj_gap=50, galore_scale=1.0, galore_proj_type='std', galore_optim_per_parameter=False, galore_with_embedding=False, galore_quantization=False, galore_proj_quant=False, galore_proj_bits=4, galore_proj_group_size=256, galore_cos_threshold=0.4, galore_gamma_proj=2, galore_queue_size=5, lisa_activated_layers=0, lisa_step_interval=20, use_flash_ckpt=False)" +} \ No newline at end of file diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/chat_template.jinja b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/chat_template.jinja new file mode 100644 index 0000000000000000000000000000000000000000..a585dec894e63da457d9440ec6aa7caa16d20860 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/chat_template.jinja @@ -0,0 +1,154 @@ +{%- set image_count = namespace(value=0) %} +{%- set video_count = namespace(value=0) %} +{%- macro render_content(content, do_vision_count, is_system_content=false) %} + {%- if content is string %} + {{- content }} + {%- elif content is iterable and content is not mapping %} + {%- for item in content %} + {%- if 'image' in item or 'image_url' in item or item.type == 'image' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain images.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set image_count.value = image_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Picture ' ~ image_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|image_pad|><|vision_end|>' }} + {%- elif 'video' in item or item.type == 'video' %} + {%- if is_system_content %} + {{- raise_exception('System message cannot contain videos.') }} + {%- endif %} + {%- if do_vision_count %} + {%- set video_count.value = video_count.value + 1 %} + {%- endif %} + {%- if add_vision_id %} + {{- 'Video ' ~ video_count.value ~ ': ' }} + {%- endif %} + {{- '<|vision_start|><|video_pad|><|vision_end|>' }} + {%- elif 'text' in item %} + {{- item.text }} + {%- else %} + {{- raise_exception('Unexpected item type in content.') }} + {%- endif %} + {%- endfor %} + {%- elif content is none or content is undefined %} + {{- '' }} + {%- else %} + {{- raise_exception('Unexpected content type.') }} + {%- endif %} +{%- endmacro %} +{%- if not messages %} + {{- raise_exception('No messages provided.') }} +{%- endif %} +{%- if tools and tools is iterable and tools is not mapping %} + {{- '<|im_start|>system\n' }} + {{- "# Tools\n\nYou have access to the following functions:\n\n" }} + {%- for tool in tools %} + {{- "\n" }} + {{- tool | tojson }} + {%- endfor %} + {{- "\n" }} + {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {%- if content %} + {{- '\n\n' + content }} + {%- endif %} + {%- endif %} + {{- '<|im_end|>\n' }} +{%- else %} + {%- if messages[0].role == 'system' %} + {%- set content = render_content(messages[0].content, false, true)|trim %} + {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }} + {%- endif %} +{%- endif %} +{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %} +{%- for message in messages[::-1] %} + {%- set index = (messages|length - 1) - loop.index0 %} + {%- if ns.multi_step_tool and message.role == "user" %} + {%- set content = render_content(message.content, false)|trim %} + {%- if not(content.startswith('') and content.endswith('')) %} + {%- set ns.multi_step_tool = false %} + {%- set ns.last_query_index = index %} + {%- endif %} + {%- endif %} +{%- endfor %} +{%- if ns.multi_step_tool %} + {{- raise_exception('No user query found in messages.') }} +{%- endif %} +{%- for message in messages %} + {%- set content = render_content(message.content, true)|trim %} + {%- if message.role == "system" %} + {%- if not loop.first %} + {{- raise_exception('System message must be at the beginning.') }} + {%- endif %} + {%- elif message.role == "user" %} + {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }} + {%- elif message.role == "assistant" %} + {%- set reasoning_content = '' %} + {%- if message.reasoning_content is string %} + {%- set reasoning_content = message.reasoning_content %} + {%- else %} + {%- if '' in content %} + {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %} + {%- set content = content.split('')[-1].lstrip('\n') %} + {%- endif %} + {%- endif %} + {%- set reasoning_content = reasoning_content|trim %} + {%- if loop.index0 > ns.last_query_index %} + {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }} + {%- else %} + {{- '<|im_start|>' + message.role + '\n' + content }} + {%- endif %} + {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %} + {%- for tool_call in message.tool_calls %} + {%- if tool_call.function is defined %} + {%- set tool_call = tool_call.function %} + {%- endif %} + {%- if loop.first %} + {%- if content|trim %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n\n' }} + {%- endif %} + {%- else %} + {{- '\n\n\n' }} + {%- endif %} + {%- if tool_call.arguments is defined %} + {%- for args_name, args_value in tool_call.arguments|items %} + {{- '\n' }} + {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %} + {{- args_value }} + {{- '\n\n' }} + {%- endfor %} + {%- endif %} + {{- '\n' }} + {%- endfor %} + {%- endif %} + {{- '<|im_end|>\n' }} + {%- elif message.role == "tool" %} + {%- if loop.previtem and loop.previtem.role != "tool" %} + {{- '<|im_start|>user' }} + {%- endif %} + {{- '\n\n' }} + {{- content }} + {{- '\n' }} + {%- if not loop.last and loop.nextitem.role != "tool" %} + {{- '<|im_end|>\n' }} + {%- elif loop.last %} + {{- '<|im_end|>\n' }} + {%- endif %} + {%- else %} + {{- raise_exception('Unexpected message role.') }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|im_start|>assistant\n' }} + {%- if enable_thinking is defined and enable_thinking is false %} + {{- '\n\n\n\n' }} + {%- else %} + {{- '\n' }} + {%- endif %} +{%- endif %} \ No newline at end of file diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/config.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/config.json new file mode 100644 index 0000000000000000000000000000000000000000..314297e9ae3a6f7ca90b01d29ee0a46cfbf53c2a --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/config.json @@ -0,0 +1,119 @@ +{ + "architectures": [ + "Qwen35VLLatentForConditionalGeneration" + ], + "dtype": "bfloat16", + "eos_token_id": 248046, + "hidden_size": 2560, + "image_token_id": 248056, + "latent_end_id": 248078, + "latent_pad_id": 248079, + "latent_start_id": 248077, + "model_type": "qwen3_5", + "pad_token_id": 248044, + "text_config": { + "attention_bias": false, + "attention_dropout": 0.0, + "attn_output_gate": true, + "bos_token_id": null, + "dtype": "bfloat16", + "eos_token_id": 248044, + "full_attention_interval": 4, + "head_dim": 256, + "hidden_act": "silu", + "hidden_size": 2560, + "initializer_range": 0.02, + "intermediate_size": 9216, + "latent_end_id": 248078, + "latent_pad_id": 248079, + "latent_start_id": 248077, + "layer_types": [ + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention", + "linear_attention", + "linear_attention", + "linear_attention", + "full_attention" + ], + "linear_conv_kernel_dim": 4, + "linear_key_head_dim": 128, + "linear_num_key_heads": 16, + "linear_num_value_heads": 32, + "linear_value_head_dim": 128, + "mamba_ssm_dtype": "float32", + "max_position_embeddings": 262144, + "mlp_only_layers": [], + "model_type": "qwen3_5_text", + "mtp_num_hidden_layers": 1, + "mtp_use_dedicated_embeddings": false, + "num_attention_heads": 16, + "num_hidden_layers": 32, + "num_key_value_heads": 4, + "pad_token_id": 248044, + "partial_rotary_factor": 0.25, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "mrope_interleaved": true, + "mrope_section": [ + 11, + 11, + 10 + ], + "partial_rotary_factor": 0.25, + "rope_theta": 10000000, + "rope_type": "default" + }, + "tie_word_embeddings": true, + "use_cache": false, + "vocab_size": 248080 + }, + "tie_word_embeddings": true, + "transformers_version": "5.8.1", + "use_cache": false, + "video_token_id": 248057, + "vision_config": { + "deepstack_visual_indexes": [], + "depth": 24, + "dtype": "bfloat16", + "hidden_act": "gelu_pytorch_tanh", + "hidden_size": 1024, + "in_channels": 3, + "initializer_range": 0.02, + "intermediate_size": 4096, + "model_type": "qwen3_5_vision", + "num_heads": 16, + "num_position_embeddings": 2304, + "out_hidden_size": 2560, + "patch_size": 16, + "spatial_merge_size": 2, + "temporal_patch_size": 2 + }, + "vision_end_token_id": 248054, + "vision_start_token_id": 248053 +} diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/generation_config.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1f1ab970fed81cb788c2945bed33833d247b9091 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/generation_config.json @@ -0,0 +1,9 @@ +{ + "_from_model_config": true, + "eos_token_id": [ + 248044, + 248046 + ], + "transformers_version": "5.8.1", + "use_cache": true +} diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..c5be8e3939027d5c499d1d9362de359d72b93209 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5db192fbcc9de1547285f2a4e99ae35f6e75783df5e579b5d5b323a712e60e28 +size 6359675185 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..83b4c1d53f0421594d845fe893075509de43caec --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3c828e6f5347aee87073bb4f1b5b6a61540475978d7eecb3989d45a23d2a30d +size 6359678641 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..3dfa131fd60f9318b20f4f9cce632a6b184b9190 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bca95de33321dce804becc35abccc163e834a9b1379cef6039074dd68c96a6c2 +size 6359679473 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..d6cc2b0483beee5a9e20e8bf793ff6e359c10a0c --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6557d2b2266ea84db3e6305f3722c23f0f7ce2e16f841c26c56c9e78313aa57 +size 6359679345 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..4b971c45619a68871419564b4e50873b69118398 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bfa82b2943c831eb02e569dd129eabe2026c47ebb70478e2d9fb59e32fff9a85 +size 6359679345 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..06284e4213047429227ef27be44b70b817b9a04e --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:196593b2f52d58c1eb8a40c3d427cbfa7fbcc7cef431aba2315c271ffc3db2e2 +size 6359679281 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..9ba6e69a4bb7400b072a0d98c628aca31d2ac0d5 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bba5210c1c30584b218ce481c191ca192f3a548d155b245eac018c796f70f72d +size 6359678641 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..1467ba5f6bd91161ff84c7546e38f37c9dd0e1bb --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d492d9d3b55bedc4d1897c41c9b0151171e5a29643e8af9d5e970cd7dfd7a9b4 +size 6359678257 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/mp_rank_00_model_states.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/mp_rank_00_model_states.pt new file mode 100644 index 0000000000000000000000000000000000000000..8ae842ec57fcdfaeff77a67a2e5146dfb31e54a3 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/global_step1500/mp_rank_00_model_states.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d246350388cefbb57ee456d792a5edae046c1db1463d5c4a60a1ad854d849571 +size 9813994319 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/latest b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/latest new file mode 100644 index 0000000000000000000000000000000000000000..c56ff7708f44fb7928fea2f70d6d7342ce0d5b67 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/latest @@ -0,0 +1 @@ +global_step1500 \ No newline at end of file diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00001-of-00003.safetensors b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e751563d208d0e5d3592f0a85bac5bfc16e065aa --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a21cd5dadc19defa093dfc67e004a53453c84257982ceec02c730539c366cd55 +size 4977834400 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00002-of-00003.safetensors b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..dc449bb59de56b35ce66df91afdb2039c2774f07 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6db06ee33e93edebe929212ddd64d125c6f03c25be92ceb93dba9e3311188b6 +size 4998696368 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00003-of-00003.safetensors b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e5ce1b5f512214102c971ca03c94fbf8474673a9 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5afcd66ba37e50fa1299af80572a55f8885c0161a67e33930b5148606815bcfb +size 440315320 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model.safetensors.index.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..bf74ca7d81215784b7c8ad30d70c2fa0982f5f2e --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/model.safetensors.index.json @@ -0,0 +1,755 @@ +{ + "metadata": { + "total_parameters": 4573292544, + "total_size": 10416754688 + }, + "weight_map": { + "_wm_projector_recon.net.0.bias": "model-00001-of-00003.safetensors", + "_wm_projector_recon.net.0.weight": "model-00001-of-00003.safetensors", + "_wm_projector_recon.net.2.bias": "model-00001-of-00003.safetensors", + "_wm_projector_recon.net.2.weight": "model-00001-of-00003.safetensors", + "_wm_projector_recon.net.4.bias": "model-00001-of-00003.safetensors", + "_wm_projector_recon.net.4.weight": "model-00001-of-00003.safetensors", + "_wm_projector_recon.norm.bias": "model-00001-of-00003.safetensors", + "_wm_projector_recon.norm.weight": "model-00001-of-00003.safetensors", + "alignment_projector.action_proj.bias": "model-00001-of-00003.safetensors", + "alignment_projector.action_proj.weight": "model-00001-of-00003.safetensors", + "alignment_projector.ln_action.bias": "model-00001-of-00003.safetensors", + "alignment_projector.ln_action.weight": "model-00001-of-00003.safetensors", + "alignment_projector.ln_state.bias": "model-00001-of-00003.safetensors", + "alignment_projector.ln_state.weight": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.1.bias": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.1.weight": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.3.bias": "model-00001-of-00003.safetensors", + "alignment_projector.mlp.3.weight": "model-00001-of-00003.safetensors", + "alignment_projector.norm.bias": "model-00001-of-00003.safetensors", + "alignment_projector.norm.weight": "model-00001-of-00003.safetensors", + "alignment_projector.state_proj.bias": "model-00001-of-00003.safetensors", + "alignment_projector.state_proj.weight": "model-00001-of-00003.safetensors", + "latent_init_embedding": "model-00001-of-00003.safetensors", + "lm_head.weight": "model-00001-of-00003.safetensors", + "model.language_model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.10.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.A_log": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.conv1d.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.dt_bias": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_a.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_b.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_qkv.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.in_proj_z.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.linear_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.language_model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.A_log": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.conv1d.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.dt_bias": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_a.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_b.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_qkv.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.in_proj_z.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.norm.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.linear_attn.out_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors", + "model.language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.language_model.norm.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.0.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.1.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.10.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.11.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.12.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.13.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.14.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.15.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.16.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.17.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.18.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.19.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.2.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.20.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.21.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.22.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.23.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.3.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.4.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.5.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.6.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.7.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.proj.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.attn.qkv.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.mlp.linear_fc2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm1.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.8.norm2.weight": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.bias": "model-00002-of-00003.safetensors", + "model.visual.blocks.9.attn.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.attn.qkv.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.mlp.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.norm1.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.norm1.weight": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.norm2.bias": "model-00003-of-00003.safetensors", + "model.visual.blocks.9.norm2.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc1.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc1.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.linear_fc2.weight": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.bias": "model-00003-of-00003.safetensors", + "model.visual.merger.norm.weight": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.bias": "model-00003-of-00003.safetensors", + "model.visual.patch_embed.proj.weight": "model-00003-of-00003.safetensors", + "model.visual.pos_embed.weight": "model-00003-of-00003.safetensors" + } +} diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/preprocessor_config.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/preprocessor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea84a437d448ff71b08df68fdd949d5cc4ebb64 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/preprocessor_config.json @@ -0,0 +1,21 @@ +{ + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "patch_size": 16, + "temporal_patch_size": 2, + "merge_size": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "processor_class": "Qwen3VLProcessor", + "image_processor_type": "Qwen2VLImageProcessorFast" +} \ No newline at end of file diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/processor_config.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/processor_config.json new file mode 100644 index 0000000000000000000000000000000000000000..33818c7f9e991ad735fd240209f4fa73e6c28c50 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/processor_config.json @@ -0,0 +1,60 @@ +{ + "image_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_processor_type": "Qwen2VLImageProcessor", + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "merge_size": 2, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "size": { + "longest_edge": 16777216, + "shortest_edge": 65536 + }, + "temporal_patch_size": 2 + }, + "processor_class": "Qwen3VLProcessor", + "video_processor": { + "do_convert_rgb": true, + "do_normalize": true, + "do_rescale": true, + "do_resize": true, + "do_sample_frames": true, + "fps": 2, + "image_mean": [ + 0.5, + 0.5, + 0.5 + ], + "image_std": [ + 0.5, + 0.5, + 0.5 + ], + "max_frames": 768, + "merge_size": 2, + "min_frames": 4, + "patch_size": 16, + "resample": 3, + "rescale_factor": 0.00392156862745098, + "return_metadata": false, + "size": { + "longest_edge": 25165824, + "shortest_edge": 4096 + }, + "temporal_patch_size": 2, + "video_processor_type": "Qwen3VLVideoProcessor" + } +} diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_0.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..9e9efed881b4aea3ce1fbacfbecc1c93783c89f5 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:51b8b905495903b77622c38443251f860b6a68e8360a7b0de2df71870ea4c164 +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_1.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..f665bebbed1c3acd979edbb73b858af5169b45d7 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:585da800b166f6a63ad6b885dbc8ddc7855580aad792e9046fc60a3cf05502db +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_2.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..f290295d1a399de2f0114e3255adf78283c9cef9 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d2bfcdfc2ba7d71b50238ab84c656fb2fec7d3289d2696cd7b580ccf4ed2b074 +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_3.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..5996764f80b8b8b675ba96726b07c7c773e1f436 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:978de33bd7b14f6449b997a18c1ef26323787eccf2f7f1745f8dfd7f4210b3a8 +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_4.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_4.pth new file mode 100644 index 0000000000000000000000000000000000000000..e72ce506516497380e6ad49c78d27694582989f4 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_4.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f460a7f977d96d18c3de69a7d7dcac5ccff2287ef425ab77d75c8ddbd9b6fd35 +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_5.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_5.pth new file mode 100644 index 0000000000000000000000000000000000000000..f8477263379ba534e41803acef1bf61dcf210de4 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_5.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5690a9b5c7a273c21d1a5e62f72105eed2b88f310755cceb45261aefbe280863 +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_6.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_6.pth new file mode 100644 index 0000000000000000000000000000000000000000..8d7a1b7fd176779fcb70a63f3929ee18a04c9b73 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_6.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bbbf85b6cc4906d9edac58c51912bf9c841ca93292e7bd660405a4e3b8c9d369 +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_7.pth b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_7.pth new file mode 100644 index 0000000000000000000000000000000000000000..b2a8c7587835cbbf473739fc0542f21191bc84cd --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/rng_state_7.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f27c88004c0582e5e409b3a286d5c7c6258c603dfcda6efaf3741fbd8276443 +size 16389 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/scheduler.pt b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..d4b7e488bfa24a9c7680a3ba105fcade57c48cbb --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1825dc5eb0e39ce8b4f2f0bbb2c7a4aa27ccba972a812fa0ac4e17a9865bb99a +size 1465 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/tokenizer.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..9e8ac8634b7c90a05a65182d43516ddf7371c521 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4012eb888c8edaca05866241a3a8319ca02d45c367eef3da25b8ba4ef002f5b3 +size 19989885 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/tokenizer_config.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1d134cd298be1e3be25db393d93a1cefe80e3214 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/tokenizer_config.json @@ -0,0 +1,33 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "image_token": "<|image_pad|>", + "is_local": true, + "local_files_only": false, + "model_max_length": 262144, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +} diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/trainer_state.json b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..63c56f5a647056580f0f555bab688ba4ce0c68c7 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/trainer_state.json @@ -0,0 +1,19744 @@ +{ + "best_global_step": 1500, + "best_metric": 0.62629604, + "best_model_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/v6-recurtorch-recon-img3-hw2-k16-12k/v0-20260605-222754/checkpoint-1500", + "epoch": 0.5590756615728661, + "eval_steps": 100, + "global_step": 1500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0003727171077152441, + "grad_norm": 642.1688842773438, + "learning_rate": 5.0000000000000004e-08, + "loss": 29.254833221435547, + "loss_alignment": 0.98828125, + "loss_alignment_w": 0.494140625, + "loss_sft": 0.8918107151985168, + "loss_total_v6": 1.634974718093872, + "loss_wm_recon": 0.99609375, + "loss_wm_recon_w": 0.2490234375, + "step": 1 + }, + { + "epoch": 0.0007454342154304882, + "grad_norm": 738.6651000976562, + "learning_rate": 1.0000000000000001e-07, + "loss": 29.71014976501465, + "loss_alignment": 0.98828125, + "loss_alignment_w": 0.494140625, + "loss_sft": 0.9509366154670715, + "loss_total_v6": 1.6950771808624268, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 2 + }, + { + "epoch": 0.0011181513231457323, + "grad_norm": 685.7134399414062, + "learning_rate": 1.5000000000000002e-07, + "loss": 29.454927444458008, + "loss_alignment": 0.9921875, + "loss_alignment_w": 0.49609375, + "loss_sft": 0.9922879338264465, + "loss_total_v6": 1.7383816242218018, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 3 + }, + { + "epoch": 0.0014908684308609765, + "grad_norm": 675.60302734375, + "learning_rate": 2.0000000000000002e-07, + "loss": 30.340919494628906, + "loss_alignment": 0.9921875, + "loss_alignment_w": 0.49609375, + "loss_sft": 1.160660743713379, + "loss_total_v6": 1.906754493713379, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 4 + }, + { + "epoch": 0.0018635855385762206, + "grad_norm": 595.6573486328125, + "learning_rate": 2.5000000000000004e-07, + "loss": 29.0760498046875, + "loss_alignment": 0.9921875, + "loss_alignment_w": 0.49609375, + "loss_sft": 1.3701270818710327, + "loss_total_v6": 2.1162209510803223, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 5 + }, + { + "epoch": 0.0022363026462914647, + "grad_norm": 616.6759643554688, + "learning_rate": 3.0000000000000004e-07, + "loss": 30.312149047851562, + "loss_alignment": 0.984375, + "loss_alignment_w": 0.4921875, + "loss_sft": 1.3790565729141235, + "loss_total_v6": 2.123197078704834, + "loss_wm_recon": 1.0078125, + "loss_wm_recon_w": 0.251953125, + "step": 6 + }, + { + "epoch": 0.002609019754006709, + "grad_norm": 556.392822265625, + "learning_rate": 3.5000000000000004e-07, + "loss": 28.869979858398438, + "loss_alignment": 0.98828125, + "loss_alignment_w": 0.494140625, + "loss_sft": 1.40814208984375, + "loss_total_v6": 2.15228271484375, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 7 + }, + { + "epoch": 0.002981736861721953, + "grad_norm": 422.8194580078125, + "learning_rate": 4.0000000000000003e-07, + "loss": 28.033710479736328, + "loss_alignment": 0.97265625, + "loss_alignment_w": 0.486328125, + "loss_sft": 0.6725314855575562, + "loss_total_v6": 1.4088596105575562, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 8 + }, + { + "epoch": 0.003354453969437197, + "grad_norm": 415.63348388671875, + "learning_rate": 4.5000000000000003e-07, + "loss": 28.14672088623047, + "loss_alignment": 1.0078125, + "loss_alignment_w": 0.50390625, + "loss_sft": 1.596529483795166, + "loss_total_v6": 2.349459171295166, + "loss_wm_recon": 0.99609375, + "loss_wm_recon_w": 0.2490234375, + "step": 9 + }, + { + "epoch": 0.003727171077152441, + "grad_norm": 399.5170593261719, + "learning_rate": 5.000000000000001e-07, + "loss": 27.577014923095703, + "loss_alignment": 0.9921875, + "loss_alignment_w": 0.49609375, + "loss_sft": 0.9909589886665344, + "loss_total_v6": 1.7370526790618896, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 10 + }, + { + "epoch": 0.004099888184867686, + "grad_norm": 245.6530303955078, + "learning_rate": 5.5e-07, + "loss": 27.60774803161621, + "loss_alignment": 0.9921875, + "loss_alignment_w": 0.49609375, + "loss_sft": 1.1845998764038086, + "loss_total_v6": 1.9306936264038086, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 11 + }, + { + "epoch": 0.004472605292582929, + "grad_norm": 242.03900146484375, + "learning_rate": 6.000000000000001e-07, + "loss": 27.530597686767578, + "loss_alignment": 0.9921875, + "loss_alignment_w": 0.49609375, + "loss_sft": 0.6820294260978699, + "loss_total_v6": 1.4281232357025146, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 12 + }, + { + "epoch": 0.004845322400298174, + "grad_norm": 198.661865234375, + "learning_rate": 6.5e-07, + "loss": 26.214059829711914, + "loss_alignment": 0.984375, + "loss_alignment_w": 0.4921875, + "loss_sft": 0.9469002485275269, + "loss_total_v6": 1.6890877485275269, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 13 + }, + { + "epoch": 0.005218039508013418, + "grad_norm": 149.8998565673828, + "learning_rate": 7.000000000000001e-07, + "loss": 25.257862091064453, + "loss_alignment": 0.984375, + "loss_alignment_w": 0.4921875, + "loss_sft": 0.4753841161727905, + "loss_total_v6": 1.2195247411727905, + "loss_wm_recon": 1.0078125, + "loss_wm_recon_w": 0.251953125, + "step": 14 + }, + { + "epoch": 0.005590756615728662, + "grad_norm": 141.48898315429688, + "learning_rate": 7.5e-07, + "loss": 25.45758628845215, + "loss_alignment": 0.97265625, + "loss_alignment_w": 0.486328125, + "loss_sft": 1.1171826124191284, + "loss_total_v6": 1.8535107374191284, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 15 + }, + { + "epoch": 0.005963473723443906, + "grad_norm": 150.29283142089844, + "learning_rate": 8.000000000000001e-07, + "loss": 25.914649963378906, + "loss_alignment": 0.9921875, + "loss_alignment_w": 0.49609375, + "loss_sft": 0.7353274822235107, + "loss_total_v6": 1.4804446697235107, + "loss_wm_recon": 0.99609375, + "loss_wm_recon_w": 0.2490234375, + "step": 16 + }, + { + "epoch": 0.00633619083115915, + "grad_norm": 118.89230346679688, + "learning_rate": 8.500000000000001e-07, + "loss": 24.914047241210938, + "loss_alignment": 0.98046875, + "loss_alignment_w": 0.490234375, + "loss_sft": 0.49088242650032043, + "loss_total_v6": 1.231116771697998, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 17 + }, + { + "epoch": 0.006708907938874394, + "grad_norm": 120.16749572753906, + "learning_rate": 9.000000000000001e-07, + "loss": 25.31186866760254, + "loss_alignment": 0.98828125, + "loss_alignment_w": 0.494140625, + "loss_sft": 1.020512342453003, + "loss_total_v6": 1.764652967453003, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 18 + }, + { + "epoch": 0.007081625046589639, + "grad_norm": 107.98271179199219, + "learning_rate": 9.500000000000001e-07, + "loss": 25.900774002075195, + "loss_alignment": 0.9765625, + "loss_alignment_w": 0.48828125, + "loss_sft": 0.6567914485931396, + "loss_total_v6": 1.3950726985931396, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 19 + }, + { + "epoch": 0.007454342154304882, + "grad_norm": 99.65218353271484, + "learning_rate": 1.0000000000000002e-06, + "loss": 24.562515258789062, + "loss_alignment": 0.9765625, + "loss_alignment_w": 0.48828125, + "loss_sft": 0.7712084054946899, + "loss_total_v6": 1.50851309299469, + "loss_wm_recon": 0.99609375, + "loss_wm_recon_w": 0.2490234375, + "step": 20 + }, + { + "epoch": 0.007827059262020127, + "grad_norm": 84.11001586914062, + "learning_rate": 1.0500000000000001e-06, + "loss": 23.813358306884766, + "loss_alignment": 0.96875, + "loss_alignment_w": 0.484375, + "loss_sft": 0.7313810586929321, + "loss_total_v6": 1.4657560586929321, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 21 + }, + { + "epoch": 0.008199776369735371, + "grad_norm": 72.41863250732422, + "learning_rate": 1.1e-06, + "loss": 24.43708038330078, + "loss_alignment": 0.96875, + "loss_alignment_w": 0.484375, + "loss_sft": 0.6998303532600403, + "loss_total_v6": 1.4342052936553955, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 22 + }, + { + "epoch": 0.008572493477450614, + "grad_norm": 64.80335235595703, + "learning_rate": 1.1500000000000002e-06, + "loss": 24.200511932373047, + "loss_alignment": 0.97265625, + "loss_alignment_w": 0.486328125, + "loss_sft": 0.70793217420578, + "loss_total_v6": 1.4442603588104248, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 23 + }, + { + "epoch": 0.008945210585165859, + "grad_norm": 59.48184585571289, + "learning_rate": 1.2000000000000002e-06, + "loss": 23.15682601928711, + "loss_alignment": 0.97265625, + "loss_alignment_w": 0.486328125, + "loss_sft": 0.7386725544929504, + "loss_total_v6": 1.4740240573883057, + "loss_wm_recon": 0.99609375, + "loss_wm_recon_w": 0.2490234375, + "step": 24 + }, + { + "epoch": 0.009317927692881103, + "grad_norm": 69.02259826660156, + "learning_rate": 1.25e-06, + "loss": 23.559423446655273, + "loss_alignment": 0.9765625, + "loss_alignment_w": 0.48828125, + "loss_sft": 0.5090436339378357, + "loss_total_v6": 1.2463483810424805, + "loss_wm_recon": 0.99609375, + "loss_wm_recon_w": 0.2490234375, + "step": 25 + }, + { + "epoch": 0.009690644800596348, + "grad_norm": 59.536685943603516, + "learning_rate": 1.3e-06, + "loss": 23.0986328125, + "loss_alignment": 0.96875, + "loss_alignment_w": 0.484375, + "loss_sft": 0.5611676573753357, + "loss_total_v6": 1.2955427169799805, + "loss_wm_recon": 1.0, + "loss_wm_recon_w": 0.25, + "step": 26 + }, + { + "epoch": 0.01006336190831159, + "grad_norm": 68.37198638916016, + "learning_rate": 1.3500000000000002e-06, + "loss": 22.796478271484375, + "loss_alignment": 0.95703125, + "loss_alignment_w": 0.478515625, + "loss_sft": 0.5158475637435913, + "loss_total_v6": 1.2433866262435913, + "loss_wm_recon": 0.99609375, + "loss_wm_recon_w": 0.2490234375, + "step": 27 + }, + { + "epoch": 0.010436079016026835, + "grad_norm": 56.29936599731445, + "learning_rate": 1.4000000000000001e-06, + "loss": 22.65923500061035, + "loss_alignment": 0.94921875, + "loss_alignment_w": 0.474609375, + "loss_sft": 0.3271106779575348, + "loss_total_v6": 1.0487903356552124, + "loss_wm_recon": 0.98828125, + "loss_wm_recon_w": 0.2470703125, + "step": 28 + }, + { + "epoch": 0.01080879612374208, + "grad_norm": 61.71649932861328, + "learning_rate": 1.45e-06, + "loss": 23.245519638061523, + "loss_alignment": 0.9609375, + "loss_alignment_w": 0.48046875, + "loss_sft": 0.648211658000946, + "loss_total_v6": 1.3767273426055908, + "loss_wm_recon": 0.9921875, + "loss_wm_recon_w": 0.248046875, + "step": 29 + }, + { + "epoch": 0.011181513231457324, + "grad_norm": 53.57901382446289, + "learning_rate": 1.5e-06, + "loss": 22.961509704589844, + "loss_alignment": 0.953125, + "loss_alignment_w": 0.4765625, + "loss_sft": 0.6306349039077759, + "loss_total_v6": 1.3552442789077759, + "loss_wm_recon": 0.9921875, + "loss_wm_recon_w": 0.248046875, + "step": 30 + }, + { + "epoch": 0.011554230339172567, + "grad_norm": 47.52519226074219, + "learning_rate": 1.5500000000000002e-06, + "loss": 22.148696899414062, + "loss_alignment": 0.94140625, + "loss_alignment_w": 0.470703125, + "loss_sft": 0.5542641878128052, + "loss_total_v6": 1.2710610628128052, + "loss_wm_recon": 0.984375, + "loss_wm_recon_w": 0.24609375, + "step": 31 + }, + { + "epoch": 0.011926947446887812, + "grad_norm": 44.574832916259766, + "learning_rate": 1.6000000000000001e-06, + "loss": 22.291624069213867, + "loss_alignment": 0.94140625, + "loss_alignment_w": 0.470703125, + "loss_sft": 0.6441787481307983, + "loss_total_v6": 1.3619521856307983, + "loss_wm_recon": 0.98828125, + "loss_wm_recon_w": 0.2470703125, + "step": 32 + }, + { + "epoch": 0.012299664554603056, + "grad_norm": 47.90296936035156, + "learning_rate": 1.6500000000000003e-06, + "loss": 22.021961212158203, + "loss_alignment": 0.92578125, + "loss_alignment_w": 0.462890625, + "loss_sft": 0.8582122921943665, + "loss_total_v6": 1.5681731700897217, + "loss_wm_recon": 0.98828125, + "loss_wm_recon_w": 0.2470703125, + "step": 33 + }, + { + "epoch": 0.0126723816623183, + "grad_norm": 45.98907470703125, + "learning_rate": 1.7000000000000002e-06, + "loss": 21.881267547607422, + "loss_alignment": 0.9375, + "loss_alignment_w": 0.46875, + "loss_sft": 0.6044222116470337, + "loss_total_v6": 1.3202425241470337, + "loss_wm_recon": 0.98828125, + "loss_wm_recon_w": 0.2470703125, + "step": 34 + }, + { + "epoch": 0.013045098770033545, + "grad_norm": 40.1224250793457, + "learning_rate": 1.75e-06, + "loss": 21.513507843017578, + "loss_alignment": 0.92578125, + "loss_alignment_w": 0.462890625, + "loss_sft": 0.5146201848983765, + "loss_total_v6": 1.2236045598983765, + "loss_wm_recon": 0.984375, + "loss_wm_recon_w": 0.24609375, + "step": 35 + }, + { + "epoch": 0.013417815877748788, + "grad_norm": 44.25279998779297, + "learning_rate": 1.8000000000000001e-06, + "loss": 20.998172760009766, + "loss_alignment": 0.8984375, + "loss_alignment_w": 0.44921875, + "loss_sft": 0.6891136169433594, + "loss_total_v6": 1.3814964294433594, + "loss_wm_recon": 0.97265625, + "loss_wm_recon_w": 0.2431640625, + "step": 36 + }, + { + "epoch": 0.013790532985464033, + "grad_norm": 82.62953186035156, + "learning_rate": 1.85e-06, + "loss": 21.191898345947266, + "loss_alignment": 0.8828125, + "loss_alignment_w": 0.44140625, + "loss_sft": 0.5984556078910828, + "loss_total_v6": 1.2810728549957275, + "loss_wm_recon": 0.96484375, + "loss_wm_recon_w": 0.2412109375, + "step": 37 + }, + { + "epoch": 0.014163250093179277, + "grad_norm": 44.74763870239258, + "learning_rate": 1.9000000000000002e-06, + "loss": 21.128612518310547, + "loss_alignment": 0.8671875, + "loss_alignment_w": 0.43359375, + "loss_sft": 0.6013138294219971, + "loss_total_v6": 1.276118516921997, + "loss_wm_recon": 0.96484375, + "loss_wm_recon_w": 0.2412109375, + "step": 38 + }, + { + "epoch": 0.014535967200894522, + "grad_norm": 78.59943389892578, + "learning_rate": 1.9500000000000004e-06, + "loss": 20.944820404052734, + "loss_alignment": 0.875, + "loss_alignment_w": 0.4375, + "loss_sft": 0.5633572340011597, + "loss_total_v6": 1.2430447340011597, + "loss_wm_recon": 0.96875, + "loss_wm_recon_w": 0.2421875, + "step": 39 + }, + { + "epoch": 0.014908684308609765, + "grad_norm": 46.90171813964844, + "learning_rate": 2.0000000000000003e-06, + "loss": 20.90599822998047, + "loss_alignment": 0.87109375, + "loss_alignment_w": 0.435546875, + "loss_sft": 0.74976646900177, + "loss_total_v6": 1.42847740650177, + "loss_wm_recon": 0.97265625, + "loss_wm_recon_w": 0.2431640625, + "step": 40 + }, + { + "epoch": 0.01528140141632501, + "grad_norm": 46.4991455078125, + "learning_rate": 2.05e-06, + "loss": 20.766246795654297, + "loss_alignment": 0.84765625, + "loss_alignment_w": 0.423828125, + "loss_sft": 0.649825394153595, + "loss_total_v6": 1.3129112720489502, + "loss_wm_recon": 0.95703125, + "loss_wm_recon_w": 0.2392578125, + "step": 41 + }, + { + "epoch": 0.015654118524040254, + "grad_norm": 43.30699920654297, + "learning_rate": 2.1000000000000002e-06, + "loss": 20.881547927856445, + "loss_alignment": 0.8515625, + "loss_alignment_w": 0.42578125, + "loss_sft": 0.7401037216186523, + "loss_total_v6": 1.4061193466186523, + "loss_wm_recon": 0.9609375, + "loss_wm_recon_w": 0.240234375, + "step": 42 + }, + { + "epoch": 0.0160268356317555, + "grad_norm": 37.661712646484375, + "learning_rate": 2.15e-06, + "loss": 20.126901626586914, + "loss_alignment": 0.8515625, + "loss_alignment_w": 0.42578125, + "loss_sft": 0.5750482678413391, + "loss_total_v6": 1.2400872707366943, + "loss_wm_recon": 0.95703125, + "loss_wm_recon_w": 0.2392578125, + "step": 43 + }, + { + "epoch": 0.016399552739470743, + "grad_norm": 39.36619186401367, + "learning_rate": 2.2e-06, + "loss": 20.30915069580078, + "loss_alignment": 0.82421875, + "loss_alignment_w": 0.412109375, + "loss_sft": 0.6749017834663391, + "loss_total_v6": 1.3252923488616943, + "loss_wm_recon": 0.953125, + "loss_wm_recon_w": 0.23828125, + "step": 44 + }, + { + "epoch": 0.016772269847185987, + "grad_norm": 38.01599884033203, + "learning_rate": 2.25e-06, + "loss": 20.259538650512695, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.48017412424087524, + "loss_total_v6": 0.7165022492408752, + "loss_wm_recon": 0.9453125, + "loss_wm_recon_w": 0.236328125, + "step": 45 + }, + { + "epoch": 0.01714498695490123, + "grad_norm": 38.2237663269043, + "learning_rate": 2.3000000000000004e-06, + "loss": 20.099788665771484, + "loss_alignment": 0.80859375, + "loss_alignment_w": 0.404296875, + "loss_sft": 0.5222454071044922, + "loss_total_v6": 1.1628704071044922, + "loss_wm_recon": 0.9453125, + "loss_wm_recon_w": 0.236328125, + "step": 46 + }, + { + "epoch": 0.017517704062616473, + "grad_norm": 37.573158264160156, + "learning_rate": 2.35e-06, + "loss": 20.50567054748535, + "loss_alignment": 0.8046875, + "loss_alignment_w": 0.40234375, + "loss_sft": 0.5515639781951904, + "loss_total_v6": 1.1892592906951904, + "loss_wm_recon": 0.94140625, + "loss_wm_recon_w": 0.2353515625, + "step": 47 + }, + { + "epoch": 0.017890421170331718, + "grad_norm": 39.46200942993164, + "learning_rate": 2.4000000000000003e-06, + "loss": 20.206918716430664, + "loss_alignment": 0.80078125, + "loss_alignment_w": 0.400390625, + "loss_sft": 0.5059560537338257, + "loss_total_v6": 1.1416982412338257, + "loss_wm_recon": 0.94140625, + "loss_wm_recon_w": 0.2353515625, + "step": 48 + }, + { + "epoch": 0.018263138278046962, + "grad_norm": 33.98234939575195, + "learning_rate": 2.4500000000000003e-06, + "loss": 20.2667236328125, + "loss_alignment": 0.78125, + "loss_alignment_w": 0.390625, + "loss_sft": 0.7308799028396606, + "loss_total_v6": 1.3549033403396606, + "loss_wm_recon": 0.93359375, + "loss_wm_recon_w": 0.2333984375, + "step": 49 + }, + { + "epoch": 0.018635855385762207, + "grad_norm": 42.92646026611328, + "learning_rate": 2.5e-06, + "loss": 19.69300651550293, + "loss_alignment": 0.76171875, + "loss_alignment_w": 0.380859375, + "loss_sft": 0.7078077793121338, + "loss_total_v6": 1.3093702793121338, + "loss_wm_recon": 0.8828125, + "loss_wm_recon_w": 0.220703125, + "step": 50 + }, + { + "epoch": 0.01900857249347745, + "grad_norm": 38.12112808227539, + "learning_rate": 2.55e-06, + "loss": 18.84749984741211, + "loss_alignment": 0.7578125, + "loss_alignment_w": 0.37890625, + "loss_sft": 0.7261800765991211, + "loss_total_v6": 1.325789451599121, + "loss_wm_recon": 0.8828125, + "loss_wm_recon_w": 0.220703125, + "step": 51 + }, + { + "epoch": 0.019381289601192696, + "grad_norm": 36.03596115112305, + "learning_rate": 2.6e-06, + "loss": 19.576507568359375, + "loss_alignment": 0.765625, + "loss_alignment_w": 0.3828125, + "loss_sft": 0.6533637046813965, + "loss_total_v6": 1.2588324546813965, + "loss_wm_recon": 0.890625, + "loss_wm_recon_w": 0.22265625, + "step": 52 + }, + { + "epoch": 0.01975400670890794, + "grad_norm": 32.860687255859375, + "learning_rate": 2.6500000000000005e-06, + "loss": 18.979429244995117, + "loss_alignment": 0.734375, + "loss_alignment_w": 0.3671875, + "loss_sft": 0.7619298100471497, + "loss_total_v6": 1.3468906879425049, + "loss_wm_recon": 0.87109375, + "loss_wm_recon_w": 0.2177734375, + "step": 53 + }, + { + "epoch": 0.02012672381662318, + "grad_norm": 36.121639251708984, + "learning_rate": 2.7000000000000004e-06, + "loss": 19.047441482543945, + "loss_alignment": 0.7421875, + "loss_alignment_w": 0.37109375, + "loss_sft": 0.8256629705429077, + "loss_total_v6": 1.4145301580429077, + "loss_wm_recon": 0.87109375, + "loss_wm_recon_w": 0.2177734375, + "step": 54 + }, + { + "epoch": 0.020499440924338426, + "grad_norm": 33.843685150146484, + "learning_rate": 2.7500000000000004e-06, + "loss": 19.496036529541016, + "loss_alignment": 0.71875, + "loss_alignment_w": 0.359375, + "loss_sft": 0.6369810700416565, + "loss_total_v6": 1.2141294479370117, + "loss_wm_recon": 0.87109375, + "loss_wm_recon_w": 0.2177734375, + "step": 55 + }, + { + "epoch": 0.02087215803205367, + "grad_norm": 32.9994010925293, + "learning_rate": 2.8000000000000003e-06, + "loss": 18.61590003967285, + "loss_alignment": 0.69921875, + "loss_alignment_w": 0.349609375, + "loss_sft": 0.5701775550842285, + "loss_total_v6": 1.1336541175842285, + "loss_wm_recon": 0.85546875, + "loss_wm_recon_w": 0.2138671875, + "step": 56 + }, + { + "epoch": 0.021244875139768915, + "grad_norm": 34.55649185180664, + "learning_rate": 2.85e-06, + "loss": 18.727394104003906, + "loss_alignment": 0.71484375, + "loss_alignment_w": 0.357421875, + "loss_sft": 0.7254595160484314, + "loss_total_v6": 1.2977252006530762, + "loss_wm_recon": 0.859375, + "loss_wm_recon_w": 0.21484375, + "step": 57 + }, + { + "epoch": 0.02161759224748416, + "grad_norm": 32.64362335205078, + "learning_rate": 2.9e-06, + "loss": 19.117477416992188, + "loss_alignment": 0.703125, + "loss_alignment_w": 0.3515625, + "loss_sft": 0.58753502368927, + "loss_total_v6": 1.15101158618927, + "loss_wm_recon": 0.84765625, + "loss_wm_recon_w": 0.2119140625, + "step": 58 + }, + { + "epoch": 0.021990309355199404, + "grad_norm": 32.187198638916016, + "learning_rate": 2.95e-06, + "loss": 18.518455505371094, + "loss_alignment": 0.69140625, + "loss_alignment_w": 0.345703125, + "loss_sft": 0.691379964351654, + "loss_total_v6": 1.2489972114562988, + "loss_wm_recon": 0.84765625, + "loss_wm_recon_w": 0.2119140625, + "step": 59 + }, + { + "epoch": 0.02236302646291465, + "grad_norm": 33.08318328857422, + "learning_rate": 3e-06, + "loss": 18.246620178222656, + "loss_alignment": 0.671875, + "loss_alignment_w": 0.3359375, + "loss_sft": 0.6046364307403564, + "loss_total_v6": 1.1485817432403564, + "loss_wm_recon": 0.83203125, + "loss_wm_recon_w": 0.2080078125, + "step": 60 + }, + { + "epoch": 0.022735743570629893, + "grad_norm": 34.76655578613281, + "learning_rate": 3.05e-06, + "loss": 17.95541000366211, + "loss_alignment": 0.6875, + "loss_alignment_w": 0.34375, + "loss_sft": 0.5211333632469177, + "loss_total_v6": 1.0719146728515625, + "loss_wm_recon": 0.828125, + "loss_wm_recon_w": 0.20703125, + "step": 61 + }, + { + "epoch": 0.023108460678345134, + "grad_norm": 32.69984817504883, + "learning_rate": 3.1000000000000004e-06, + "loss": 18.06586456298828, + "loss_alignment": 0.6875, + "loss_alignment_w": 0.34375, + "loss_sft": 0.6461591720581055, + "loss_total_v6": 1.1949872970581055, + "loss_wm_recon": 0.8203125, + "loss_wm_recon_w": 0.205078125, + "step": 62 + }, + { + "epoch": 0.02348117778606038, + "grad_norm": 33.31916809082031, + "learning_rate": 3.1500000000000003e-06, + "loss": 18.22872543334961, + "loss_alignment": 0.6640625, + "loss_alignment_w": 0.33203125, + "loss_sft": 0.4926115870475769, + "loss_total_v6": 1.0297210216522217, + "loss_wm_recon": 0.8203125, + "loss_wm_recon_w": 0.205078125, + "step": 63 + }, + { + "epoch": 0.023853894893775623, + "grad_norm": 33.18378829956055, + "learning_rate": 3.2000000000000003e-06, + "loss": 17.90694808959961, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5716736316680908, + "loss_total_v6": 0.7777283191680908, + "loss_wm_recon": 0.82421875, + "loss_wm_recon_w": 0.2060546875, + "step": 64 + }, + { + "epoch": 0.024226612001490868, + "grad_norm": 32.72041320800781, + "learning_rate": 3.2500000000000002e-06, + "loss": 17.95746612548828, + "loss_alignment": 0.6484375, + "loss_alignment_w": 0.32421875, + "loss_sft": 0.7222312688827515, + "loss_total_v6": 1.2466453313827515, + "loss_wm_recon": 0.80078125, + "loss_wm_recon_w": 0.2001953125, + "step": 65 + }, + { + "epoch": 0.024599329109206113, + "grad_norm": 32.7794303894043, + "learning_rate": 3.3000000000000006e-06, + "loss": 17.887977600097656, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7914941310882568, + "loss_total_v6": 0.9887597560882568, + "loss_wm_recon": 0.7890625, + "loss_wm_recon_w": 0.197265625, + "step": 66 + }, + { + "epoch": 0.024972046216921357, + "grad_norm": 31.044544219970703, + "learning_rate": 3.3500000000000005e-06, + "loss": 17.685394287109375, + "loss_alignment": 0.64453125, + "loss_alignment_w": 0.322265625, + "loss_sft": 0.6898031830787659, + "loss_total_v6": 1.210310935974121, + "loss_wm_recon": 0.79296875, + "loss_wm_recon_w": 0.1982421875, + "step": 67 + }, + { + "epoch": 0.0253447633246366, + "grad_norm": 32.697731018066406, + "learning_rate": 3.4000000000000005e-06, + "loss": 17.823057174682617, + "loss_alignment": 0.6328125, + "loss_alignment_w": 0.31640625, + "loss_sft": 0.5551432967185974, + "loss_total_v6": 1.0688152313232422, + "loss_wm_recon": 0.7890625, + "loss_wm_recon_w": 0.197265625, + "step": 68 + }, + { + "epoch": 0.025717480432351846, + "grad_norm": 32.440032958984375, + "learning_rate": 3.45e-06, + "loss": 17.755449295043945, + "loss_alignment": 0.625, + "loss_alignment_w": 0.3125, + "loss_sft": 0.5545801520347595, + "loss_total_v6": 1.0594630241394043, + "loss_wm_recon": 0.76953125, + "loss_wm_recon_w": 0.1923828125, + "step": 69 + }, + { + "epoch": 0.02609019754006709, + "grad_norm": 37.678382873535156, + "learning_rate": 3.5e-06, + "loss": 17.182287216186523, + "loss_alignment": 0.625, + "loss_alignment_w": 0.3125, + "loss_sft": 0.7881507277488708, + "loss_total_v6": 1.2881507873535156, + "loss_wm_recon": 0.75, + "loss_wm_recon_w": 0.1875, + "step": 70 + }, + { + "epoch": 0.026462914647782332, + "grad_norm": 31.004871368408203, + "learning_rate": 3.5500000000000003e-06, + "loss": 17.25444793701172, + "loss_alignment": 0.6328125, + "loss_alignment_w": 0.31640625, + "loss_sft": 0.5467127561569214, + "loss_total_v6": 1.0525721311569214, + "loss_wm_recon": 0.7578125, + "loss_wm_recon_w": 0.189453125, + "step": 71 + }, + { + "epoch": 0.026835631755497576, + "grad_norm": 30.946535110473633, + "learning_rate": 3.6000000000000003e-06, + "loss": 17.528636932373047, + "loss_alignment": 0.625, + "loss_alignment_w": 0.3125, + "loss_sft": 0.5252281427383423, + "loss_total_v6": 1.0222984552383423, + "loss_wm_recon": 0.73828125, + "loss_wm_recon_w": 0.1845703125, + "step": 72 + }, + { + "epoch": 0.02720834886321282, + "grad_norm": 31.828222274780273, + "learning_rate": 3.65e-06, + "loss": 17.45116424560547, + "loss_alignment": 0.59375, + "loss_alignment_w": 0.296875, + "loss_sft": 0.7261127233505249, + "loss_total_v6": 1.201698660850525, + "loss_wm_recon": 0.71484375, + "loss_wm_recon_w": 0.1787109375, + "step": 73 + }, + { + "epoch": 0.027581065970928065, + "grad_norm": 32.40294647216797, + "learning_rate": 3.7e-06, + "loss": 17.132556915283203, + "loss_alignment": 0.62890625, + "loss_alignment_w": 0.314453125, + "loss_sft": 0.774795413017273, + "loss_total_v6": 1.273818850517273, + "loss_wm_recon": 0.73828125, + "loss_wm_recon_w": 0.1845703125, + "step": 74 + }, + { + "epoch": 0.02795378307864331, + "grad_norm": 31.096494674682617, + "learning_rate": 3.7500000000000005e-06, + "loss": 17.182418823242188, + "loss_alignment": 0.625, + "loss_alignment_w": 0.3125, + "loss_sft": 0.9863027930259705, + "loss_total_v6": 1.4784903526306152, + "loss_wm_recon": 0.71875, + "loss_wm_recon_w": 0.1796875, + "step": 75 + }, + { + "epoch": 0.028326500186358555, + "grad_norm": 32.63114547729492, + "learning_rate": 3.8000000000000005e-06, + "loss": 17.397735595703125, + "loss_alignment": 0.58203125, + "loss_alignment_w": 0.291015625, + "loss_sft": 0.8287179470062256, + "loss_total_v6": 1.2964913845062256, + "loss_wm_recon": 0.70703125, + "loss_wm_recon_w": 0.1767578125, + "step": 76 + }, + { + "epoch": 0.0286992172940738, + "grad_norm": 31.019302368164062, + "learning_rate": 3.85e-06, + "loss": 17.06117057800293, + "loss_alignment": 0.5859375, + "loss_alignment_w": 0.29296875, + "loss_sft": 0.49188143014907837, + "loss_total_v6": 0.9596548676490784, + "loss_wm_recon": 0.69921875, + "loss_wm_recon_w": 0.1748046875, + "step": 77 + }, + { + "epoch": 0.029071934401789044, + "grad_norm": 29.835783004760742, + "learning_rate": 3.900000000000001e-06, + "loss": 17.007251739501953, + "loss_alignment": 0.609375, + "loss_alignment_w": 0.3046875, + "loss_sft": 0.6450256705284119, + "loss_total_v6": 1.1245179176330566, + "loss_wm_recon": 0.69921875, + "loss_wm_recon_w": 0.1748046875, + "step": 78 + }, + { + "epoch": 0.029444651509504285, + "grad_norm": 31.35733985900879, + "learning_rate": 3.95e-06, + "loss": 16.63011932373047, + "loss_alignment": 0.6171875, + "loss_alignment_w": 0.30859375, + "loss_sft": 0.5934960246086121, + "loss_total_v6": 1.0798242092132568, + "loss_wm_recon": 0.7109375, + "loss_wm_recon_w": 0.177734375, + "step": 79 + }, + { + "epoch": 0.02981736861721953, + "grad_norm": 29.776811599731445, + "learning_rate": 4.000000000000001e-06, + "loss": 16.907699584960938, + "loss_alignment": 0.6171875, + "loss_alignment_w": 0.30859375, + "loss_sft": 0.6803476810455322, + "loss_total_v6": 1.1666758060455322, + "loss_wm_recon": 0.7109375, + "loss_wm_recon_w": 0.177734375, + "step": 80 + }, + { + "epoch": 0.030190085724934774, + "grad_norm": 29.670822143554688, + "learning_rate": 4.05e-06, + "loss": 16.842918395996094, + "loss_alignment": 0.59375, + "loss_alignment_w": 0.296875, + "loss_sft": 0.49154040217399597, + "loss_total_v6": 0.9563841819763184, + "loss_wm_recon": 0.671875, + "loss_wm_recon_w": 0.16796875, + "step": 81 + }, + { + "epoch": 0.03056280283265002, + "grad_norm": 32.5367431640625, + "learning_rate": 4.1e-06, + "loss": 16.300546646118164, + "loss_alignment": 0.578125, + "loss_alignment_w": 0.2890625, + "loss_sft": 0.4871807098388672, + "loss_total_v6": 0.9461650848388672, + "loss_wm_recon": 0.6796875, + "loss_wm_recon_w": 0.169921875, + "step": 82 + }, + { + "epoch": 0.030935519940365263, + "grad_norm": 36.84134292602539, + "learning_rate": 4.15e-06, + "loss": 16.58626365661621, + "loss_alignment": 0.5546875, + "loss_alignment_w": 0.27734375, + "loss_sft": 0.33545881509780884, + "loss_total_v6": 0.7768650650978088, + "loss_wm_recon": 0.65625, + "loss_wm_recon_w": 0.1640625, + "step": 83 + }, + { + "epoch": 0.03130823704808051, + "grad_norm": 30.133636474609375, + "learning_rate": 4.2000000000000004e-06, + "loss": 16.455913543701172, + "loss_alignment": 0.55078125, + "loss_alignment_w": 0.275390625, + "loss_sft": 0.6900305151939392, + "loss_total_v6": 1.129483699798584, + "loss_wm_recon": 0.65625, + "loss_wm_recon_w": 0.1640625, + "step": 84 + }, + { + "epoch": 0.03168095415579575, + "grad_norm": 29.700756072998047, + "learning_rate": 4.25e-06, + "loss": 16.23173713684082, + "loss_alignment": 0.55078125, + "loss_alignment_w": 0.275390625, + "loss_sft": 0.2694815397262573, + "loss_total_v6": 0.7001456022262573, + "loss_wm_recon": 0.62109375, + "loss_wm_recon_w": 0.1552734375, + "step": 85 + }, + { + "epoch": 0.032053671263511, + "grad_norm": 30.17758560180664, + "learning_rate": 4.3e-06, + "loss": 16.336673736572266, + "loss_alignment": 0.5859375, + "loss_alignment_w": 0.29296875, + "loss_sft": 0.7935765385627747, + "loss_total_v6": 1.2506077289581299, + "loss_wm_recon": 0.65625, + "loss_wm_recon_w": 0.1640625, + "step": 86 + }, + { + "epoch": 0.03242638837122624, + "grad_norm": 30.66332244873047, + "learning_rate": 4.350000000000001e-06, + "loss": 16.00567054748535, + "loss_alignment": 0.546875, + "loss_alignment_w": 0.2734375, + "loss_sft": 0.7761882543563843, + "loss_total_v6": 1.2078288793563843, + "loss_wm_recon": 0.6328125, + "loss_wm_recon_w": 0.158203125, + "step": 87 + }, + { + "epoch": 0.032799105478941486, + "grad_norm": 29.200092315673828, + "learning_rate": 4.4e-06, + "loss": 16.253841400146484, + "loss_alignment": 0.53125, + "loss_alignment_w": 0.265625, + "loss_sft": 0.4773183763027191, + "loss_total_v6": 0.8923574686050415, + "loss_wm_recon": 0.59765625, + "loss_wm_recon_w": 0.1494140625, + "step": 88 + }, + { + "epoch": 0.03317182258665673, + "grad_norm": 30.03227424621582, + "learning_rate": 4.450000000000001e-06, + "loss": 16.36832046508789, + "loss_alignment": 0.53125, + "loss_alignment_w": 0.265625, + "loss_sft": 0.6827205419540405, + "loss_total_v6": 1.1065486669540405, + "loss_wm_recon": 0.6328125, + "loss_wm_recon_w": 0.158203125, + "step": 89 + }, + { + "epoch": 0.033544539694371975, + "grad_norm": 27.389554977416992, + "learning_rate": 4.5e-06, + "loss": 16.011043548583984, + "loss_alignment": 0.515625, + "loss_alignment_w": 0.2578125, + "loss_sft": 0.7472290396690369, + "loss_total_v6": 1.1485962867736816, + "loss_wm_recon": 0.57421875, + "loss_wm_recon_w": 0.1435546875, + "step": 90 + }, + { + "epoch": 0.033917256802087216, + "grad_norm": 27.727907180786133, + "learning_rate": 4.5500000000000005e-06, + "loss": 15.599132537841797, + "loss_alignment": 0.515625, + "loss_alignment_w": 0.2578125, + "loss_sft": 0.5225967764854431, + "loss_total_v6": 0.9298233389854431, + "loss_wm_recon": 0.59765625, + "loss_wm_recon_w": 0.1494140625, + "step": 91 + }, + { + "epoch": 0.03428997390980246, + "grad_norm": 29.753156661987305, + "learning_rate": 4.600000000000001e-06, + "loss": 15.709417343139648, + "loss_alignment": 0.54296875, + "loss_alignment_w": 0.271484375, + "loss_sft": 0.38304412364959717, + "loss_total_v6": 0.8019894361495972, + "loss_wm_recon": 0.58984375, + "loss_wm_recon_w": 0.1474609375, + "step": 92 + }, + { + "epoch": 0.034662691017517705, + "grad_norm": 29.418603897094727, + "learning_rate": 4.65e-06, + "loss": 15.756609916687012, + "loss_alignment": 0.5, + "loss_alignment_w": 0.25, + "loss_sft": 0.39105331897735596, + "loss_total_v6": 0.782654881477356, + "loss_wm_recon": 0.56640625, + "loss_wm_recon_w": 0.1416015625, + "step": 93 + }, + { + "epoch": 0.035035408125232946, + "grad_norm": 30.482667922973633, + "learning_rate": 4.7e-06, + "loss": 15.664122581481934, + "loss_alignment": 0.5, + "loss_alignment_w": 0.25, + "loss_sft": 0.7323926091194153, + "loss_total_v6": 1.1269237995147705, + "loss_wm_recon": 0.578125, + "loss_wm_recon_w": 0.14453125, + "step": 94 + }, + { + "epoch": 0.035408125232948194, + "grad_norm": 29.620349884033203, + "learning_rate": 4.75e-06, + "loss": 16.080751419067383, + "loss_alignment": 0.5234375, + "loss_alignment_w": 0.26171875, + "loss_sft": 0.7650907635688782, + "loss_total_v6": 1.1772000789642334, + "loss_wm_recon": 0.6015625, + "loss_wm_recon_w": 0.150390625, + "step": 95 + }, + { + "epoch": 0.035780842340663435, + "grad_norm": 29.19498634338379, + "learning_rate": 4.800000000000001e-06, + "loss": 16.066802978515625, + "loss_alignment": 0.5234375, + "loss_alignment_w": 0.26171875, + "loss_sft": 0.7074967622756958, + "loss_total_v6": 1.1156998872756958, + "loss_wm_recon": 0.5859375, + "loss_wm_recon_w": 0.146484375, + "step": 96 + }, + { + "epoch": 0.03615355944837868, + "grad_norm": 27.1844539642334, + "learning_rate": 4.85e-06, + "loss": 15.241634368896484, + "loss_alignment": 0.546875, + "loss_alignment_w": 0.2734375, + "loss_sft": 0.6870298981666565, + "loss_total_v6": 1.1069517135620117, + "loss_wm_recon": 0.5859375, + "loss_wm_recon_w": 0.146484375, + "step": 97 + }, + { + "epoch": 0.036526276556093924, + "grad_norm": 29.397022247314453, + "learning_rate": 4.9000000000000005e-06, + "loss": 15.190471649169922, + "loss_alignment": 0.49609375, + "loss_alignment_w": 0.248046875, + "loss_sft": 0.5463375449180603, + "loss_total_v6": 0.9340328574180603, + "loss_wm_recon": 0.55859375, + "loss_wm_recon_w": 0.1396484375, + "step": 98 + }, + { + "epoch": 0.03689899366380917, + "grad_norm": 28.19446563720703, + "learning_rate": 4.95e-06, + "loss": 15.637191772460938, + "loss_alignment": 0.48828125, + "loss_alignment_w": 0.244140625, + "loss_sft": 0.776823103427887, + "loss_total_v6": 1.1547527313232422, + "loss_wm_recon": 0.53515625, + "loss_wm_recon_w": 0.1337890625, + "step": 99 + }, + { + "epoch": 0.03727171077152441, + "grad_norm": 28.929712295532227, + "learning_rate": 5e-06, + "loss": 15.70954704284668, + "loss_alignment": 0.48828125, + "loss_alignment_w": 0.244140625, + "loss_sft": 0.6243842840194702, + "loss_total_v6": 1.0071967840194702, + "loss_wm_recon": 0.5546875, + "loss_wm_recon_w": 0.138671875, + "step": 100 + }, + { + "epoch": 0.03727171077152441, + "eval_loss": 0.8844190835952759, + "eval_loss_alignment": 0.4595874232104701, + "eval_loss_alignment_w": 0.22979371160523504, + "eval_loss_sft": 0.5201878637457505, + "eval_loss_total_v6": 0.8844190703816394, + "eval_loss_wm_recon": 0.5377499833066239, + "eval_loss_wm_recon_w": 0.13443749582665598, + "eval_runtime": 229.6524, + "eval_samples_per_second": 8.125, + "eval_steps_per_second": 1.019, + "step": 100 + }, + { + "epoch": 0.037644427879239654, + "grad_norm": 27.91078758239746, + "learning_rate": 5.050000000000001e-06, + "loss": 15.89583683013916, + "loss_alignment": 0.484375, + "loss_alignment_w": 0.2421875, + "loss_sft": 0.7310113906860352, + "loss_total_v6": 1.1108942031860352, + "loss_wm_recon": 0.55078125, + "loss_wm_recon_w": 0.1376953125, + "step": 101 + }, + { + "epoch": 0.0380171449869549, + "grad_norm": 29.22214698791504, + "learning_rate": 5.1e-06, + "loss": 15.193297386169434, + "loss_alignment": 0.4765625, + "loss_alignment_w": 0.23828125, + "loss_sft": 0.3541729152202606, + "loss_total_v6": 0.728196382522583, + "loss_wm_recon": 0.54296875, + "loss_wm_recon_w": 0.1357421875, + "step": 102 + }, + { + "epoch": 0.038389862094670144, + "grad_norm": 29.45073890686035, + "learning_rate": 5.150000000000001e-06, + "loss": 14.998429298400879, + "loss_alignment": 0.46875, + "loss_alignment_w": 0.234375, + "loss_sft": 0.641055703163147, + "loss_total_v6": 1.010196328163147, + "loss_wm_recon": 0.5390625, + "loss_wm_recon_w": 0.134765625, + "step": 103 + }, + { + "epoch": 0.03876257920238539, + "grad_norm": 27.519216537475586, + "learning_rate": 5.2e-06, + "loss": 15.063032150268555, + "loss_alignment": 0.48046875, + "loss_alignment_w": 0.240234375, + "loss_sft": 0.6027105450630188, + "loss_total_v6": 0.9747808575630188, + "loss_wm_recon": 0.52734375, + "loss_wm_recon_w": 0.1318359375, + "step": 104 + }, + { + "epoch": 0.03913529631010063, + "grad_norm": 27.916667938232422, + "learning_rate": 5.2500000000000006e-06, + "loss": 14.963380813598633, + "loss_alignment": 0.45703125, + "loss_alignment_w": 0.228515625, + "loss_sft": 0.6067346334457397, + "loss_total_v6": 0.9651330709457397, + "loss_wm_recon": 0.51953125, + "loss_wm_recon_w": 0.1298828125, + "step": 105 + }, + { + "epoch": 0.03950801341781588, + "grad_norm": 50.800636291503906, + "learning_rate": 5.300000000000001e-06, + "loss": 15.043898582458496, + "loss_alignment": 0.44140625, + "loss_alignment_w": 0.220703125, + "loss_sft": 0.7433100938796997, + "loss_total_v6": 1.0890132188796997, + "loss_wm_recon": 0.5, + "loss_wm_recon_w": 0.125, + "step": 106 + }, + { + "epoch": 0.03988073052553112, + "grad_norm": 106.32879638671875, + "learning_rate": 5.3500000000000004e-06, + "loss": 14.984663963317871, + "loss_alignment": 0.48046875, + "loss_alignment_w": 0.240234375, + "loss_sft": 0.6500458717346191, + "loss_total_v6": 1.0211396217346191, + "loss_wm_recon": 0.5234375, + "loss_wm_recon_w": 0.130859375, + "step": 107 + }, + { + "epoch": 0.04025344763324636, + "grad_norm": 28.743125915527344, + "learning_rate": 5.400000000000001e-06, + "loss": 14.828023910522461, + "loss_alignment": 0.4375, + "loss_alignment_w": 0.21875, + "loss_sft": 0.6223199367523193, + "loss_total_v6": 0.9636285305023193, + "loss_wm_recon": 0.490234375, + "loss_wm_recon_w": 0.12255859375, + "step": 108 + }, + { + "epoch": 0.04062616474096161, + "grad_norm": 69.8284683227539, + "learning_rate": 5.450000000000001e-06, + "loss": 14.812788009643555, + "loss_alignment": 0.48046875, + "loss_alignment_w": 0.240234375, + "loss_sft": 0.716393768787384, + "loss_total_v6": 1.0894405841827393, + "loss_wm_recon": 0.53125, + "loss_wm_recon_w": 0.1328125, + "step": 109 + }, + { + "epoch": 0.04099888184867685, + "grad_norm": 26.56362533569336, + "learning_rate": 5.500000000000001e-06, + "loss": 14.81771469116211, + "loss_alignment": 0.44921875, + "loss_alignment_w": 0.224609375, + "loss_sft": 0.6948627829551697, + "loss_total_v6": 1.0405659675598145, + "loss_wm_recon": 0.484375, + "loss_wm_recon_w": 0.12109375, + "step": 110 + }, + { + "epoch": 0.0413715989563921, + "grad_norm": 28.911123275756836, + "learning_rate": 5.550000000000001e-06, + "loss": 14.819934844970703, + "loss_alignment": 0.4453125, + "loss_alignment_w": 0.22265625, + "loss_sft": 0.42623740434646606, + "loss_total_v6": 0.7690108418464661, + "loss_wm_recon": 0.48046875, + "loss_wm_recon_w": 0.1201171875, + "step": 111 + }, + { + "epoch": 0.04174431606410734, + "grad_norm": 27.69231605529785, + "learning_rate": 5.600000000000001e-06, + "loss": 14.829294204711914, + "loss_alignment": 0.46875, + "loss_alignment_w": 0.234375, + "loss_sft": 0.5102092623710632, + "loss_total_v6": 0.8734905123710632, + "loss_wm_recon": 0.515625, + "loss_wm_recon_w": 0.12890625, + "step": 112 + }, + { + "epoch": 0.04211703317182259, + "grad_norm": 25.921173095703125, + "learning_rate": 5.65e-06, + "loss": 14.930845260620117, + "loss_alignment": 0.4453125, + "loss_alignment_w": 0.22265625, + "loss_sft": 0.6263903379440308, + "loss_total_v6": 0.9701403379440308, + "loss_wm_recon": 0.484375, + "loss_wm_recon_w": 0.12109375, + "step": 113 + }, + { + "epoch": 0.04248975027953783, + "grad_norm": 28.390323638916016, + "learning_rate": 5.7e-06, + "loss": 14.43623161315918, + "loss_alignment": 0.421875, + "loss_alignment_w": 0.2109375, + "loss_sft": 0.3496459722518921, + "loss_total_v6": 0.6748412847518921, + "loss_wm_recon": 0.45703125, + "loss_wm_recon_w": 0.1142578125, + "step": 114 + }, + { + "epoch": 0.04286246738725308, + "grad_norm": 26.2836971282959, + "learning_rate": 5.75e-06, + "loss": 14.518203735351562, + "loss_alignment": 0.42578125, + "loss_alignment_w": 0.212890625, + "loss_sft": 0.8470167517662048, + "loss_total_v6": 1.1795363426208496, + "loss_wm_recon": 0.478515625, + "loss_wm_recon_w": 0.11962890625, + "step": 115 + }, + { + "epoch": 0.04323518449496832, + "grad_norm": 25.656543731689453, + "learning_rate": 5.8e-06, + "loss": 14.636451721191406, + "loss_alignment": 0.4296875, + "loss_alignment_w": 0.21484375, + "loss_sft": 0.5712069272994995, + "loss_total_v6": 0.9022616147994995, + "loss_wm_recon": 0.46484375, + "loss_wm_recon_w": 0.1162109375, + "step": 116 + }, + { + "epoch": 0.04360790160268356, + "grad_norm": 28.09157943725586, + "learning_rate": 5.85e-06, + "loss": 14.450299263000488, + "loss_alignment": 0.4296875, + "loss_alignment_w": 0.21484375, + "loss_sft": 0.4436671733856201, + "loss_total_v6": 0.7713038921356201, + "loss_wm_recon": 0.451171875, + "loss_wm_recon_w": 0.11279296875, + "step": 117 + }, + { + "epoch": 0.04398061871039881, + "grad_norm": 28.25891876220703, + "learning_rate": 5.9e-06, + "loss": 14.673810005187988, + "loss_alignment": 0.43359375, + "loss_alignment_w": 0.216796875, + "loss_sft": 0.5863756537437439, + "loss_total_v6": 0.9184069037437439, + "loss_wm_recon": 0.4609375, + "loss_wm_recon_w": 0.115234375, + "step": 118 + }, + { + "epoch": 0.04435333581811405, + "grad_norm": 27.481279373168945, + "learning_rate": 5.950000000000001e-06, + "loss": 14.560100555419922, + "loss_alignment": 0.38671875, + "loss_alignment_w": 0.193359375, + "loss_sft": 0.610163688659668, + "loss_total_v6": 0.912409782409668, + "loss_wm_recon": 0.435546875, + "loss_wm_recon_w": 0.10888671875, + "step": 119 + }, + { + "epoch": 0.0447260529258293, + "grad_norm": 26.679929733276367, + "learning_rate": 6e-06, + "loss": 13.913738250732422, + "loss_alignment": 0.40234375, + "loss_alignment_w": 0.201171875, + "loss_sft": 0.4415431320667267, + "loss_total_v6": 0.7511134147644043, + "loss_wm_recon": 0.43359375, + "loss_wm_recon_w": 0.1083984375, + "step": 120 + }, + { + "epoch": 0.04509877003354454, + "grad_norm": 27.34861183166504, + "learning_rate": 6.0500000000000005e-06, + "loss": 14.321090698242188, + "loss_alignment": 0.390625, + "loss_alignment_w": 0.1953125, + "loss_sft": 0.5819268226623535, + "loss_total_v6": 0.8827080726623535, + "loss_wm_recon": 0.421875, + "loss_wm_recon_w": 0.10546875, + "step": 121 + }, + { + "epoch": 0.045471487141259787, + "grad_norm": 27.59479331970215, + "learning_rate": 6.1e-06, + "loss": 14.087081909179688, + "loss_alignment": 0.41796875, + "loss_alignment_w": 0.208984375, + "loss_sft": 0.6351821422576904, + "loss_total_v6": 0.9550063610076904, + "loss_wm_recon": 0.443359375, + "loss_wm_recon_w": 0.11083984375, + "step": 122 + }, + { + "epoch": 0.04584420424897503, + "grad_norm": 26.666227340698242, + "learning_rate": 6.15e-06, + "loss": 14.462867736816406, + "loss_alignment": 0.4296875, + "loss_alignment_w": 0.21484375, + "loss_sft": 0.5588797926902771, + "loss_total_v6": 0.8801688551902771, + "loss_wm_recon": 0.42578125, + "loss_wm_recon_w": 0.1064453125, + "step": 123 + }, + { + "epoch": 0.04621692135669027, + "grad_norm": 26.505935668945312, + "learning_rate": 6.200000000000001e-06, + "loss": 14.555377960205078, + "loss_alignment": 0.390625, + "loss_alignment_w": 0.1953125, + "loss_sft": 0.42771103978157043, + "loss_total_v6": 0.731421947479248, + "loss_wm_recon": 0.43359375, + "loss_wm_recon_w": 0.1083984375, + "step": 124 + }, + { + "epoch": 0.04658963846440552, + "grad_norm": 28.565656661987305, + "learning_rate": 6.25e-06, + "loss": 13.765182495117188, + "loss_alignment": 0.43359375, + "loss_alignment_w": 0.216796875, + "loss_sft": 0.7102047204971313, + "loss_total_v6": 1.0393062829971313, + "loss_wm_recon": 0.44921875, + "loss_wm_recon_w": 0.1123046875, + "step": 125 + }, + { + "epoch": 0.04696235557212076, + "grad_norm": 29.96276092529297, + "learning_rate": 6.300000000000001e-06, + "loss": 13.982359886169434, + "loss_alignment": 0.3828125, + "loss_alignment_w": 0.19140625, + "loss_sft": 0.5887995362281799, + "loss_total_v6": 0.8886042237281799, + "loss_wm_recon": 0.43359375, + "loss_wm_recon_w": 0.1083984375, + "step": 126 + }, + { + "epoch": 0.047335072679836006, + "grad_norm": 26.794797897338867, + "learning_rate": 6.35e-06, + "loss": 13.568775177001953, + "loss_alignment": 0.390625, + "loss_alignment_w": 0.1953125, + "loss_sft": 0.5066527724266052, + "loss_total_v6": 0.8084105849266052, + "loss_wm_recon": 0.42578125, + "loss_wm_recon_w": 0.1064453125, + "step": 127 + }, + { + "epoch": 0.04770778978755125, + "grad_norm": 26.03616714477539, + "learning_rate": 6.4000000000000006e-06, + "loss": 13.95271110534668, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.5006424784660339, + "loss_total_v6": 0.7774979472160339, + "loss_wm_recon": 0.419921875, + "loss_wm_recon_w": 0.10498046875, + "step": 128 + }, + { + "epoch": 0.048080506895266495, + "grad_norm": 26.07466697692871, + "learning_rate": 6.450000000000001e-06, + "loss": 13.428468704223633, + "loss_alignment": 0.37890625, + "loss_alignment_w": 0.189453125, + "loss_sft": 0.38079288601875305, + "loss_total_v6": 0.6722967624664307, + "loss_wm_recon": 0.408203125, + "loss_wm_recon_w": 0.10205078125, + "step": 129 + }, + { + "epoch": 0.048453224002981736, + "grad_norm": 26.34423828125, + "learning_rate": 6.5000000000000004e-06, + "loss": 13.515369415283203, + "loss_alignment": 0.375, + "loss_alignment_w": 0.1875, + "loss_sft": 0.4472501575946808, + "loss_total_v6": 0.7368009090423584, + "loss_wm_recon": 0.408203125, + "loss_wm_recon_w": 0.10205078125, + "step": 130 + }, + { + "epoch": 0.048825941110696984, + "grad_norm": 27.20524787902832, + "learning_rate": 6.550000000000001e-06, + "loss": 13.379810333251953, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.5710241198539734, + "loss_total_v6": 0.8532506823539734, + "loss_wm_recon": 0.39453125, + "loss_wm_recon_w": 0.0986328125, + "step": 131 + }, + { + "epoch": 0.049198658218412225, + "grad_norm": 27.477245330810547, + "learning_rate": 6.600000000000001e-06, + "loss": 13.614898681640625, + "loss_alignment": 0.37109375, + "loss_alignment_w": 0.185546875, + "loss_sft": 0.5360734462738037, + "loss_total_v6": 0.8187882900238037, + "loss_wm_recon": 0.388671875, + "loss_wm_recon_w": 0.09716796875, + "step": 132 + }, + { + "epoch": 0.049571375326127466, + "grad_norm": 27.85592269897461, + "learning_rate": 6.650000000000001e-06, + "loss": 13.568267822265625, + "loss_alignment": 0.37890625, + "loss_alignment_w": 0.189453125, + "loss_sft": 0.7417057156562805, + "loss_total_v6": 1.0341861248016357, + "loss_wm_recon": 0.412109375, + "loss_wm_recon_w": 0.10302734375, + "step": 133 + }, + { + "epoch": 0.049944092433842714, + "grad_norm": 25.110519409179688, + "learning_rate": 6.700000000000001e-06, + "loss": 13.42818832397461, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.6766354441642761, + "loss_total_v6": 0.9554440379142761, + "loss_wm_recon": 0.388671875, + "loss_wm_recon_w": 0.09716796875, + "step": 134 + }, + { + "epoch": 0.050316809541557955, + "grad_norm": 27.615671157836914, + "learning_rate": 6.750000000000001e-06, + "loss": 13.443164825439453, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.6159802675247192, + "loss_total_v6": 0.8947888612747192, + "loss_wm_recon": 0.388671875, + "loss_wm_recon_w": 0.09716796875, + "step": 135 + }, + { + "epoch": 0.0506895266492732, + "grad_norm": 26.723936080932617, + "learning_rate": 6.800000000000001e-06, + "loss": 13.594236373901367, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6693643927574158, + "loss_total_v6": 0.9305948615074158, + "loss_wm_recon": 0.373046875, + "loss_wm_recon_w": 0.09326171875, + "step": 136 + }, + { + "epoch": 0.051062243756988444, + "grad_norm": 25.84304428100586, + "learning_rate": 6.850000000000001e-06, + "loss": 13.936576843261719, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.6408809423446655, + "loss_total_v6": 0.9240840673446655, + "loss_wm_recon": 0.3984375, + "loss_wm_recon_w": 0.099609375, + "step": 137 + }, + { + "epoch": 0.05143496086470369, + "grad_norm": 32.684608459472656, + "learning_rate": 6.9e-06, + "loss": 14.152076721191406, + "loss_alignment": 0.35546875, + "loss_alignment_w": 0.177734375, + "loss_sft": 0.5798165798187256, + "loss_total_v6": 0.8527657985687256, + "loss_wm_recon": 0.380859375, + "loss_wm_recon_w": 0.09521484375, + "step": 138 + }, + { + "epoch": 0.05180767797241893, + "grad_norm": 27.01464080810547, + "learning_rate": 6.95e-06, + "loss": 13.352197647094727, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.6516146063804626, + "loss_total_v6": 0.9284700751304626, + "loss_wm_recon": 0.380859375, + "loss_wm_recon_w": 0.09521484375, + "step": 139 + }, + { + "epoch": 0.05218039508013418, + "grad_norm": 26.219526290893555, + "learning_rate": 7e-06, + "loss": 13.3309326171875, + "loss_alignment": 0.37109375, + "loss_alignment_w": 0.185546875, + "loss_sft": 0.42872926592826843, + "loss_total_v6": 0.7099792957305908, + "loss_wm_recon": 0.3828125, + "loss_wm_recon_w": 0.095703125, + "step": 140 + }, + { + "epoch": 0.05255311218784942, + "grad_norm": 26.95772933959961, + "learning_rate": 7.05e-06, + "loss": 13.482292175292969, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.6907346248626709, + "loss_total_v6": 0.9700314998626709, + "loss_wm_recon": 0.390625, + "loss_wm_recon_w": 0.09765625, + "step": 141 + }, + { + "epoch": 0.052925829295564664, + "grad_norm": 25.512826919555664, + "learning_rate": 7.100000000000001e-06, + "loss": 13.552568435668945, + "loss_alignment": 0.35546875, + "loss_alignment_w": 0.177734375, + "loss_sft": 0.6008790135383606, + "loss_total_v6": 0.8713868260383606, + "loss_wm_recon": 0.37109375, + "loss_wm_recon_w": 0.0927734375, + "step": 142 + }, + { + "epoch": 0.05329854640327991, + "grad_norm": 24.996105194091797, + "learning_rate": 7.15e-06, + "loss": 13.565333366394043, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.662481427192688, + "loss_total_v6": 0.921270489692688, + "loss_wm_recon": 0.36328125, + "loss_wm_recon_w": 0.0908203125, + "step": 143 + }, + { + "epoch": 0.05367126351099515, + "grad_norm": 25.75019073486328, + "learning_rate": 7.2000000000000005e-06, + "loss": 13.054479598999023, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.6059414744377136, + "loss_total_v6": 0.8661953806877136, + "loss_wm_recon": 0.361328125, + "loss_wm_recon_w": 0.09033203125, + "step": 144 + }, + { + "epoch": 0.0540439806187104, + "grad_norm": 26.230886459350586, + "learning_rate": 7.25e-06, + "loss": 13.30683708190918, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6850976943969727, + "loss_total_v6": 0.9419336318969727, + "loss_wm_recon": 0.37109375, + "loss_wm_recon_w": 0.0927734375, + "step": 145 + }, + { + "epoch": 0.05441669772642564, + "grad_norm": 26.246198654174805, + "learning_rate": 7.3e-06, + "loss": 13.28740406036377, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.4888063669204712, + "loss_total_v6": 0.7480837106704712, + "loss_wm_recon": 0.357421875, + "loss_wm_recon_w": 0.08935546875, + "step": 146 + }, + { + "epoch": 0.05478941483414089, + "grad_norm": 24.498788833618164, + "learning_rate": 7.350000000000001e-06, + "loss": 13.331707000732422, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.723057746887207, + "loss_total_v6": 0.951085090637207, + "loss_wm_recon": 0.318359375, + "loss_wm_recon_w": 0.07958984375, + "step": 147 + }, + { + "epoch": 0.05516213194185613, + "grad_norm": 24.716854095458984, + "learning_rate": 7.4e-06, + "loss": 13.279703140258789, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.25783422589302063, + "loss_total_v6": 0.5102756023406982, + "loss_wm_recon": 0.345703125, + "loss_wm_recon_w": 0.08642578125, + "step": 148 + }, + { + "epoch": 0.05553484904957137, + "grad_norm": 24.921607971191406, + "learning_rate": 7.450000000000001e-06, + "loss": 13.3544921875, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.6618510484695435, + "loss_total_v6": 0.9274760484695435, + "loss_wm_recon": 0.375, + "loss_wm_recon_w": 0.09375, + "step": 149 + }, + { + "epoch": 0.05590756615728662, + "grad_norm": 25.715179443359375, + "learning_rate": 7.500000000000001e-06, + "loss": 13.080272674560547, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.3711278736591339, + "loss_total_v6": 0.4609716236591339, + "loss_wm_recon": 0.359375, + "loss_wm_recon_w": 0.08984375, + "step": 150 + }, + { + "epoch": 0.05628028326500186, + "grad_norm": 25.551759719848633, + "learning_rate": 7.5500000000000006e-06, + "loss": 13.34568977355957, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.6442106366157532, + "loss_total_v6": 0.9088590741157532, + "loss_wm_recon": 0.36328125, + "loss_wm_recon_w": 0.0908203125, + "step": 151 + }, + { + "epoch": 0.05665300037271711, + "grad_norm": 25.81338119506836, + "learning_rate": 7.600000000000001e-06, + "loss": 13.325825691223145, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.29573705792427063, + "loss_total_v6": 0.5393893718719482, + "loss_wm_recon": 0.333984375, + "loss_wm_recon_w": 0.08349609375, + "step": 152 + }, + { + "epoch": 0.05702571748043235, + "grad_norm": 24.358964920043945, + "learning_rate": 7.650000000000001e-06, + "loss": 12.70477294921875, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.6797473430633545, + "loss_total_v6": 0.9458606243133545, + "loss_wm_recon": 0.369140625, + "loss_wm_recon_w": 0.09228515625, + "step": 153 + }, + { + "epoch": 0.0573984345881476, + "grad_norm": 24.44900894165039, + "learning_rate": 7.7e-06, + "loss": 12.723812103271484, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.252532422542572, + "loss_total_v6": 0.514251172542572, + "loss_wm_recon": 0.359375, + "loss_wm_recon_w": 0.08984375, + "step": 154 + }, + { + "epoch": 0.05777115169586284, + "grad_norm": 25.68391990661621, + "learning_rate": 7.75e-06, + "loss": 13.003815650939941, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.49241504073143005, + "loss_total_v6": 0.7507158517837524, + "loss_wm_recon": 0.353515625, + "loss_wm_recon_w": 0.08837890625, + "step": 155 + }, + { + "epoch": 0.05814386880357809, + "grad_norm": 27.977705001831055, + "learning_rate": 7.800000000000002e-06, + "loss": 13.231951713562012, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.48236986994743347, + "loss_total_v6": 0.7230925559997559, + "loss_wm_recon": 0.314453125, + "loss_wm_recon_w": 0.07861328125, + "step": 156 + }, + { + "epoch": 0.05851658591129333, + "grad_norm": 25.120098114013672, + "learning_rate": 7.850000000000001e-06, + "loss": 12.910268783569336, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.7349309921264648, + "loss_total_v6": 0.9922552108764648, + "loss_wm_recon": 0.357421875, + "loss_wm_recon_w": 0.08935546875, + "step": 157 + }, + { + "epoch": 0.05888930301900857, + "grad_norm": 24.168888092041016, + "learning_rate": 7.9e-06, + "loss": 13.397481918334961, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.5624256730079651, + "loss_total_v6": 0.8168202042579651, + "loss_wm_recon": 0.330078125, + "loss_wm_recon_w": 0.08251953125, + "step": 158 + }, + { + "epoch": 0.05926202012672382, + "grad_norm": 23.820695877075195, + "learning_rate": 7.950000000000002e-06, + "loss": 12.798599243164062, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.36446887254714966, + "loss_total_v6": 0.6154454350471497, + "loss_wm_recon": 0.33984375, + "loss_wm_recon_w": 0.0849609375, + "step": 159 + }, + { + "epoch": 0.05963473723443906, + "grad_norm": 23.655235290527344, + "learning_rate": 8.000000000000001e-06, + "loss": 13.0496187210083, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7144919633865356, + "loss_total_v6": 0.9605857133865356, + "loss_wm_recon": 0.3359375, + "loss_wm_recon_w": 0.083984375, + "step": 160 + }, + { + "epoch": 0.06000745434215431, + "grad_norm": 23.83197021484375, + "learning_rate": 8.050000000000001e-06, + "loss": 13.158914566040039, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.7155365347862244, + "loss_total_v6": 0.9528412222862244, + "loss_wm_recon": 0.31640625, + "loss_wm_recon_w": 0.0791015625, + "step": 161 + }, + { + "epoch": 0.06038017144986955, + "grad_norm": 23.757850646972656, + "learning_rate": 8.1e-06, + "loss": 12.771656036376953, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4863375723361969, + "loss_total_v6": 0.7280367612838745, + "loss_wm_recon": 0.326171875, + "loss_wm_recon_w": 0.08154296875, + "step": 162 + }, + { + "epoch": 0.060752888557584796, + "grad_norm": 25.012893676757812, + "learning_rate": 8.15e-06, + "loss": 12.688763618469238, + "loss_alignment": 0.42578125, + "loss_alignment_w": 0.212890625, + "loss_sft": 0.7367733716964722, + "loss_total_v6": 1.0546444654464722, + "loss_wm_recon": 0.419921875, + "loss_wm_recon_w": 0.10498046875, + "step": 163 + }, + { + "epoch": 0.06112560566530004, + "grad_norm": 24.08907127380371, + "learning_rate": 8.2e-06, + "loss": 12.858004570007324, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.34969988465309143, + "loss_total_v6": 0.584563136100769, + "loss_wm_recon": 0.337890625, + "loss_wm_recon_w": 0.08447265625, + "step": 164 + }, + { + "epoch": 0.061498322773015285, + "grad_norm": 24.385128021240234, + "learning_rate": 8.25e-06, + "loss": 12.83822250366211, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3931788504123688, + "loss_total_v6": 0.6309717893600464, + "loss_wm_recon": 0.318359375, + "loss_wm_recon_w": 0.07958984375, + "step": 165 + }, + { + "epoch": 0.061871039880730526, + "grad_norm": 22.313875198364258, + "learning_rate": 8.3e-06, + "loss": 13.01869010925293, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.525087296962738, + "loss_total_v6": 0.767763078212738, + "loss_wm_recon": 0.322265625, + "loss_wm_recon_w": 0.08056640625, + "step": 166 + }, + { + "epoch": 0.06224375698844577, + "grad_norm": 28.97943878173828, + "learning_rate": 8.35e-06, + "loss": 12.990924835205078, + "loss_alignment": 0.359375, + "loss_alignment_w": 0.1796875, + "loss_sft": 0.5430225133895874, + "loss_total_v6": 0.8110889196395874, + "loss_wm_recon": 0.353515625, + "loss_wm_recon_w": 0.08837890625, + "step": 167 + }, + { + "epoch": 0.06261647409616102, + "grad_norm": 24.97012710571289, + "learning_rate": 8.400000000000001e-06, + "loss": 12.68250560760498, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5789633989334106, + "loss_total_v6": 0.8206626176834106, + "loss_wm_recon": 0.333984375, + "loss_wm_recon_w": 0.08349609375, + "step": 168 + }, + { + "epoch": 0.06298919120387626, + "grad_norm": 25.19610023498535, + "learning_rate": 8.45e-06, + "loss": 12.771345138549805, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6731404662132263, + "loss_total_v6": 0.9133748412132263, + "loss_wm_recon": 0.3203125, + "loss_wm_recon_w": 0.080078125, + "step": 169 + }, + { + "epoch": 0.0633619083115915, + "grad_norm": 23.88608169555664, + "learning_rate": 8.5e-06, + "loss": 12.932181358337402, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5639291405677795, + "loss_total_v6": 0.8163705468177795, + "loss_wm_recon": 0.337890625, + "loss_wm_recon_w": 0.08447265625, + "step": 170 + }, + { + "epoch": 0.06373462541930675, + "grad_norm": 25.365949630737305, + "learning_rate": 8.550000000000001e-06, + "loss": 12.721809387207031, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7040281891822815, + "loss_total_v6": 0.9467039704322815, + "loss_wm_recon": 0.322265625, + "loss_wm_recon_w": 0.08056640625, + "step": 171 + }, + { + "epoch": 0.064107342527022, + "grad_norm": 23.50446891784668, + "learning_rate": 8.6e-06, + "loss": 12.565056800842285, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5428447723388672, + "loss_total_v6": 0.7923564910888672, + "loss_wm_recon": 0.326171875, + "loss_wm_recon_w": 0.08154296875, + "step": 172 + }, + { + "epoch": 0.06448005963473724, + "grad_norm": 24.367725372314453, + "learning_rate": 8.65e-06, + "loss": 13.353194236755371, + "loss_alignment": 0.3515625, + "loss_alignment_w": 0.17578125, + "loss_sft": 0.8114805817604065, + "loss_total_v6": 1.0756406784057617, + "loss_wm_recon": 0.353515625, + "loss_wm_recon_w": 0.08837890625, + "step": 173 + }, + { + "epoch": 0.06485277674245248, + "grad_norm": 23.729921340942383, + "learning_rate": 8.700000000000001e-06, + "loss": 12.272136688232422, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.7221375703811646, + "loss_total_v6": 0.9623719453811646, + "loss_wm_recon": 0.2890625, + "loss_wm_recon_w": 0.072265625, + "step": 174 + }, + { + "epoch": 0.06522549385016772, + "grad_norm": 24.145370483398438, + "learning_rate": 8.750000000000001e-06, + "loss": 12.720224380493164, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.29733985662460327, + "loss_total_v6": 0.5307382941246033, + "loss_wm_recon": 0.31640625, + "loss_wm_recon_w": 0.0791015625, + "step": 175 + }, + { + "epoch": 0.06559821095788297, + "grad_norm": 24.01311683654785, + "learning_rate": 8.8e-06, + "loss": 12.535828590393066, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6707072854042053, + "loss_total_v6": 0.9114299416542053, + "loss_wm_recon": 0.330078125, + "loss_wm_recon_w": 0.08251953125, + "step": 176 + }, + { + "epoch": 0.0659709280655982, + "grad_norm": 27.008981704711914, + "learning_rate": 8.85e-06, + "loss": 12.968091011047363, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5351061820983887, + "loss_total_v6": 0.7821764945983887, + "loss_wm_recon": 0.33203125, + "loss_wm_recon_w": 0.0830078125, + "step": 177 + }, + { + "epoch": 0.06634364517331345, + "grad_norm": 24.805261611938477, + "learning_rate": 8.900000000000001e-06, + "loss": 12.773661613464355, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.32734403014183044, + "loss_total_v6": 0.5880862474441528, + "loss_wm_recon": 0.34765625, + "loss_wm_recon_w": 0.0869140625, + "step": 178 + }, + { + "epoch": 0.0667163622810287, + "grad_norm": 23.415836334228516, + "learning_rate": 8.95e-06, + "loss": 12.665550231933594, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5109899640083313, + "loss_total_v6": 0.7648962140083313, + "loss_wm_recon": 0.34375, + "loss_wm_recon_w": 0.0859375, + "step": 179 + }, + { + "epoch": 0.06708907938874395, + "grad_norm": 24.851898193359375, + "learning_rate": 9e-06, + "loss": 12.456584930419922, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4041411280632019, + "loss_total_v6": 0.6477934718132019, + "loss_wm_recon": 0.326171875, + "loss_wm_recon_w": 0.08154296875, + "step": 180 + }, + { + "epoch": 0.06746179649645918, + "grad_norm": 23.488821029663086, + "learning_rate": 9.050000000000001e-06, + "loss": 12.640741348266602, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.4988459348678589, + "loss_total_v6": 0.7464045286178589, + "loss_wm_recon": 0.333984375, + "loss_wm_recon_w": 0.08349609375, + "step": 181 + }, + { + "epoch": 0.06783451360417443, + "grad_norm": 24.72068977355957, + "learning_rate": 9.100000000000001e-06, + "loss": 12.775335311889648, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6219881176948547, + "loss_total_v6": 0.8734529614448547, + "loss_wm_recon": 0.333984375, + "loss_wm_recon_w": 0.08349609375, + "step": 182 + }, + { + "epoch": 0.06820723071188968, + "grad_norm": 23.89264678955078, + "learning_rate": 9.15e-06, + "loss": 13.080034255981445, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6673559546470642, + "loss_total_v6": 0.9173559546470642, + "loss_wm_recon": 0.328125, + "loss_wm_recon_w": 0.08203125, + "step": 183 + }, + { + "epoch": 0.06857994781960491, + "grad_norm": 24.887868881225586, + "learning_rate": 9.200000000000002e-06, + "loss": 12.917919158935547, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.5923545360565186, + "loss_total_v6": 0.8599326610565186, + "loss_wm_recon": 0.3359375, + "loss_wm_recon_w": 0.083984375, + "step": 184 + }, + { + "epoch": 0.06895266492732016, + "grad_norm": 25.077245712280273, + "learning_rate": 9.250000000000001e-06, + "loss": 12.953877449035645, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7931193709373474, + "loss_total_v6": 1.0401897430419922, + "loss_wm_recon": 0.33984375, + "loss_wm_recon_w": 0.0849609375, + "step": 185 + }, + { + "epoch": 0.06932538203503541, + "grad_norm": 23.1666202545166, + "learning_rate": 9.3e-06, + "loss": 12.800132751464844, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6681950092315674, + "loss_total_v6": 0.9098942279815674, + "loss_wm_recon": 0.326171875, + "loss_wm_recon_w": 0.08154296875, + "step": 186 + }, + { + "epoch": 0.06969809914275066, + "grad_norm": 23.05252456665039, + "learning_rate": 9.350000000000002e-06, + "loss": 12.763373374938965, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5065325498580933, + "loss_total_v6": 0.7521380186080933, + "loss_wm_recon": 0.333984375, + "loss_wm_recon_w": 0.08349609375, + "step": 187 + }, + { + "epoch": 0.07007081625046589, + "grad_norm": 24.225698471069336, + "learning_rate": 9.4e-06, + "loss": 12.569709777832031, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.638306736946106, + "loss_total_v6": 0.880005955696106, + "loss_wm_recon": 0.318359375, + "loss_wm_recon_w": 0.07958984375, + "step": 188 + }, + { + "epoch": 0.07044353335818114, + "grad_norm": 25.450298309326172, + "learning_rate": 9.450000000000001e-06, + "loss": 12.567134857177734, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.8136391043663025, + "loss_total_v6": 1.0489907264709473, + "loss_wm_recon": 0.31640625, + "loss_wm_recon_w": 0.0791015625, + "step": 189 + }, + { + "epoch": 0.07081625046589639, + "grad_norm": 27.21112060546875, + "learning_rate": 9.5e-06, + "loss": 12.666770935058594, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.46556589007377625, + "loss_total_v6": 0.7004292011260986, + "loss_wm_recon": 0.306640625, + "loss_wm_recon_w": 0.07666015625, + "step": 190 + }, + { + "epoch": 0.07118896757361162, + "grad_norm": 24.59716796875, + "learning_rate": 9.55e-06, + "loss": 12.685144424438477, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5237654447555542, + "loss_total_v6": 0.7639998197555542, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 191 + }, + { + "epoch": 0.07156168468132687, + "grad_norm": 27.303152084350586, + "learning_rate": 9.600000000000001e-06, + "loss": 12.976956367492676, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5026719570159912, + "loss_total_v6": 0.7385118007659912, + "loss_wm_recon": 0.310546875, + "loss_wm_recon_w": 0.07763671875, + "step": 192 + }, + { + "epoch": 0.07193440178904212, + "grad_norm": 25.264055252075195, + "learning_rate": 9.65e-06, + "loss": 12.279470443725586, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5458315014839172, + "loss_total_v6": 0.7865541577339172, + "loss_wm_recon": 0.314453125, + "loss_wm_recon_w": 0.07861328125, + "step": 193 + }, + { + "epoch": 0.07230711889675737, + "grad_norm": 24.935300827026367, + "learning_rate": 9.7e-06, + "loss": 13.15893840789795, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.629654049873352, + "loss_total_v6": 0.870864987373352, + "loss_wm_recon": 0.31640625, + "loss_wm_recon_w": 0.0791015625, + "step": 194 + }, + { + "epoch": 0.0726798360044726, + "grad_norm": 24.04203224182129, + "learning_rate": 9.75e-06, + "loss": 13.130416870117188, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6143895983695984, + "loss_total_v6": 0.8604833483695984, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 195 + }, + { + "epoch": 0.07305255311218785, + "grad_norm": 22.2935733795166, + "learning_rate": 9.800000000000001e-06, + "loss": 12.546340942382812, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.539275586605072, + "loss_total_v6": 0.777556836605072, + "loss_wm_recon": 0.3203125, + "loss_wm_recon_w": 0.080078125, + "step": 196 + }, + { + "epoch": 0.0734252702199031, + "grad_norm": 22.307594299316406, + "learning_rate": 9.85e-06, + "loss": 12.917438507080078, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.7040975093841553, + "loss_total_v6": 0.9428670406341553, + "loss_wm_recon": 0.322265625, + "loss_wm_recon_w": 0.08056640625, + "step": 197 + }, + { + "epoch": 0.07379798732761834, + "grad_norm": 21.84006118774414, + "learning_rate": 9.9e-06, + "loss": 12.373948097229004, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.8791847825050354, + "loss_total_v6": 1.1135597229003906, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 198 + }, + { + "epoch": 0.07417070443533358, + "grad_norm": 22.791791915893555, + "learning_rate": 9.950000000000001e-06, + "loss": 12.844953536987305, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6767513751983643, + "loss_total_v6": 0.9130795001983643, + "loss_wm_recon": 0.3203125, + "loss_wm_recon_w": 0.080078125, + "step": 199 + }, + { + "epoch": 0.07454342154304883, + "grad_norm": 22.34391212463379, + "learning_rate": 1e-05, + "loss": 12.410675048828125, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.41038188338279724, + "loss_total_v6": 0.6530576944351196, + "loss_wm_recon": 0.314453125, + "loss_wm_recon_w": 0.07861328125, + "step": 200 + }, + { + "epoch": 0.07454342154304883, + "eval_loss": 0.7421479821205139, + "eval_loss_alignment": 0.2992350260416667, + "eval_loss_alignment_w": 0.14961751302083334, + "eval_loss_sft": 0.5162574474214234, + "eval_loss_total_v6": 0.7421480383819494, + "eval_loss_wm_recon": 0.3050923143696581, + "eval_loss_wm_recon_w": 0.07627307859241453, + "eval_runtime": 227.691, + "eval_samples_per_second": 8.195, + "eval_steps_per_second": 1.028, + "step": 200 + }, + { + "epoch": 0.07491613865076407, + "grad_norm": 23.840904235839844, + "learning_rate": 9.999995997915474e-06, + "loss": 12.787130355834961, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5244817137718201, + "loss_total_v6": 0.7681340575218201, + "loss_wm_recon": 0.318359375, + "loss_wm_recon_w": 0.07958984375, + "step": 201 + }, + { + "epoch": 0.07528885575847931, + "grad_norm": 21.457429885864258, + "learning_rate": 9.9999839916683e-06, + "loss": 12.691133499145508, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.6876522898674011, + "loss_total_v6": 0.9356991648674011, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 202 + }, + { + "epoch": 0.07566157286619456, + "grad_norm": 22.402523040771484, + "learning_rate": 9.999963981277698e-06, + "loss": 12.719423294067383, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6061851382255554, + "loss_total_v6": 0.8513023257255554, + "loss_wm_recon": 0.32421875, + "loss_wm_recon_w": 0.0810546875, + "step": 203 + }, + { + "epoch": 0.0760342899739098, + "grad_norm": 23.109521865844727, + "learning_rate": 9.999935966775703e-06, + "loss": 12.64628791809082, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.3961516320705414, + "loss_total_v6": 0.6271086931228638, + "loss_wm_recon": 0.298828125, + "loss_wm_recon_w": 0.07470703125, + "step": 204 + }, + { + "epoch": 0.07640700708162505, + "grad_norm": 24.797489166259766, + "learning_rate": 9.99989994820716e-06, + "loss": 12.507320404052734, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.7098096013069153, + "loss_total_v6": 0.9427197575569153, + "loss_wm_recon": 0.306640625, + "loss_wm_recon_w": 0.07666015625, + "step": 205 + }, + { + "epoch": 0.07677972418934029, + "grad_norm": 23.02496337890625, + "learning_rate": 9.999855925629728e-06, + "loss": 12.6818265914917, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.613024115562439, + "loss_total_v6": 0.847399115562439, + "loss_wm_recon": 0.296875, + "loss_wm_recon_w": 0.07421875, + "step": 206 + }, + { + "epoch": 0.07715244129705554, + "grad_norm": 22.889219284057617, + "learning_rate": 9.999803899113885e-06, + "loss": 12.617919921875, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.27483952045440674, + "loss_total_v6": 0.5092145204544067, + "loss_wm_recon": 0.3046875, + "loss_wm_recon_w": 0.076171875, + "step": 207 + }, + { + "epoch": 0.07752515840477078, + "grad_norm": 23.183469772338867, + "learning_rate": 9.999743868742911e-06, + "loss": 13.163107872009277, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.989540696144104, + "loss_total_v6": 1.256630539894104, + "loss_wm_recon": 0.341796875, + "loss_wm_recon_w": 0.08544921875, + "step": 208 + }, + { + "epoch": 0.07789787551248602, + "grad_norm": 23.262853622436523, + "learning_rate": 9.999675834612905e-06, + "loss": 12.433422088623047, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.4445054531097412, + "loss_total_v6": 0.6915757656097412, + "loss_wm_recon": 0.32421875, + "loss_wm_recon_w": 0.0810546875, + "step": 209 + }, + { + "epoch": 0.07827059262020127, + "grad_norm": 23.49773597717285, + "learning_rate": 9.999599796832781e-06, + "loss": 12.860153198242188, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6200342774391174, + "loss_total_v6": 0.8602686524391174, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 210 + }, + { + "epoch": 0.07864330972791651, + "grad_norm": 23.617462158203125, + "learning_rate": 9.999515755524262e-06, + "loss": 13.34882926940918, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3503260612487793, + "loss_total_v6": 0.5847010612487793, + "loss_wm_recon": 0.3046875, + "loss_wm_recon_w": 0.076171875, + "step": 211 + }, + { + "epoch": 0.07901602683563176, + "grad_norm": 23.37769889831543, + "learning_rate": 9.999423710821886e-06, + "loss": 12.860208511352539, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.42268243432044983, + "loss_total_v6": 0.6599870920181274, + "loss_wm_recon": 0.30859375, + "loss_wm_recon_w": 0.0771484375, + "step": 212 + }, + { + "epoch": 0.079388743943347, + "grad_norm": 22.267637252807617, + "learning_rate": 9.999323662872998e-06, + "loss": 12.431062698364258, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.394132524728775, + "loss_total_v6": 0.6265543699264526, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 213 + }, + { + "epoch": 0.07976146105106224, + "grad_norm": 21.727460861206055, + "learning_rate": 9.999215611837758e-06, + "loss": 12.534347534179688, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5518964529037476, + "loss_total_v6": 0.7882245779037476, + "loss_wm_recon": 0.3046875, + "loss_wm_recon_w": 0.076171875, + "step": 214 + }, + { + "epoch": 0.08013417815877749, + "grad_norm": 20.766786575317383, + "learning_rate": 9.999099557889141e-06, + "loss": 12.598224639892578, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7107508778572083, + "loss_total_v6": 0.9387782216072083, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 215 + }, + { + "epoch": 0.08050689526649273, + "grad_norm": 22.71192741394043, + "learning_rate": 9.998975501212926e-06, + "loss": 12.507852554321289, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.3604694902896881, + "loss_total_v6": 0.5938678979873657, + "loss_wm_recon": 0.30859375, + "loss_wm_recon_w": 0.0771484375, + "step": 216 + }, + { + "epoch": 0.08087961237420797, + "grad_norm": 22.616283416748047, + "learning_rate": 9.998843442007712e-06, + "loss": 12.790752410888672, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.9668520092964172, + "loss_total_v6": 1.2026917934417725, + "loss_wm_recon": 0.302734375, + "loss_wm_recon_w": 0.07568359375, + "step": 217 + }, + { + "epoch": 0.08125232948192322, + "grad_norm": 22.99342918395996, + "learning_rate": 9.9987033804849e-06, + "loss": 12.461231231689453, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6117072105407715, + "loss_total_v6": 0.8529181480407715, + "loss_wm_recon": 0.31640625, + "loss_wm_recon_w": 0.0791015625, + "step": 218 + }, + { + "epoch": 0.08162504658963847, + "grad_norm": 22.652484893798828, + "learning_rate": 9.998555316868706e-06, + "loss": 12.674074172973633, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.5372273921966553, + "loss_total_v6": 0.7794148921966553, + "loss_wm_recon": 0.3046875, + "loss_wm_recon_w": 0.076171875, + "step": 219 + }, + { + "epoch": 0.0819977636973537, + "grad_norm": 22.935522079467773, + "learning_rate": 9.998399251396156e-06, + "loss": 12.710786819458008, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6966339349746704, + "loss_total_v6": 0.9319854974746704, + "loss_wm_recon": 0.30078125, + "loss_wm_recon_w": 0.0751953125, + "step": 220 + }, + { + "epoch": 0.08237048080506895, + "grad_norm": 23.317041397094727, + "learning_rate": 9.998235184317086e-06, + "loss": 12.916373252868652, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5489178895950317, + "loss_total_v6": 0.7769452333450317, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 221 + }, + { + "epoch": 0.0827431979127842, + "grad_norm": 21.992107391357422, + "learning_rate": 9.998063115894137e-06, + "loss": 12.84865951538086, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.37758317589759827, + "loss_total_v6": 0.5934034585952759, + "loss_wm_recon": 0.28515625, + "loss_wm_recon_w": 0.0712890625, + "step": 222 + }, + { + "epoch": 0.08311591502049943, + "grad_norm": 21.641841888427734, + "learning_rate": 9.997883046402763e-06, + "loss": 12.841352462768555, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5898029804229736, + "loss_total_v6": 0.8227131366729736, + "loss_wm_recon": 0.306640625, + "loss_wm_recon_w": 0.07666015625, + "step": 223 + }, + { + "epoch": 0.08348863212821468, + "grad_norm": 22.39847755432129, + "learning_rate": 9.997694976131229e-06, + "loss": 12.612836837768555, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.43118640780448914, + "loss_total_v6": 0.6719090938568115, + "loss_wm_recon": 0.298828125, + "loss_wm_recon_w": 0.07470703125, + "step": 224 + }, + { + "epoch": 0.08386134923592993, + "grad_norm": 22.493412017822266, + "learning_rate": 9.997498905380598e-06, + "loss": 12.638813018798828, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.7223869562149048, + "loss_total_v6": 0.9572502374649048, + "loss_wm_recon": 0.306640625, + "loss_wm_recon_w": 0.07666015625, + "step": 225 + }, + { + "epoch": 0.08423406634364518, + "grad_norm": 28.34202003479004, + "learning_rate": 9.99729483446475e-06, + "loss": 12.628128051757812, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.7763200402259827, + "loss_total_v6": 1.026808261871338, + "loss_wm_recon": 0.337890625, + "loss_wm_recon_w": 0.08447265625, + "step": 226 + }, + { + "epoch": 0.08460678345136041, + "grad_norm": 23.07426643371582, + "learning_rate": 9.99708276371037e-06, + "loss": 12.231121063232422, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4809027314186096, + "loss_total_v6": 0.6981878876686096, + "loss_wm_recon": 0.291015625, + "loss_wm_recon_w": 0.07275390625, + "step": 227 + }, + { + "epoch": 0.08497950055907566, + "grad_norm": 23.48894500732422, + "learning_rate": 9.996862693456948e-06, + "loss": 12.964494705200195, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6987888813018799, + "loss_total_v6": 0.9277927875518799, + "loss_wm_recon": 0.298828125, + "loss_wm_recon_w": 0.07470703125, + "step": 228 + }, + { + "epoch": 0.08535221766679091, + "grad_norm": 22.425806045532227, + "learning_rate": 9.996634624056775e-06, + "loss": 12.955562591552734, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3331412374973297, + "loss_total_v6": 0.5611685514450073, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 229 + }, + { + "epoch": 0.08572493477450616, + "grad_norm": 22.488933563232422, + "learning_rate": 9.996398555874958e-06, + "loss": 12.473427772521973, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5041531920433044, + "loss_total_v6": 0.7419461607933044, + "loss_wm_recon": 0.302734375, + "loss_wm_recon_w": 0.07568359375, + "step": 230 + }, + { + "epoch": 0.08609765188222139, + "grad_norm": 22.187759399414062, + "learning_rate": 9.9961544892894e-06, + "loss": 12.725934982299805, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.28897547721862793, + "loss_total_v6": 0.5336043834686279, + "loss_wm_recon": 0.306640625, + "loss_wm_recon_w": 0.07666015625, + "step": 231 + }, + { + "epoch": 0.08647036898993664, + "grad_norm": 21.630739212036133, + "learning_rate": 9.995902424690813e-06, + "loss": 12.511314392089844, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5345820188522339, + "loss_total_v6": 0.7694453001022339, + "loss_wm_recon": 0.298828125, + "loss_wm_recon_w": 0.07470703125, + "step": 232 + }, + { + "epoch": 0.08684308609765189, + "grad_norm": 22.445615768432617, + "learning_rate": 9.995642362482708e-06, + "loss": 12.325992584228516, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5915509462356567, + "loss_total_v6": 0.8337384462356567, + "loss_wm_recon": 0.296875, + "loss_wm_recon_w": 0.07421875, + "step": 233 + }, + { + "epoch": 0.08721580320536712, + "grad_norm": 20.931428909301758, + "learning_rate": 9.9953743030814e-06, + "loss": 12.549156188964844, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5700362920761108, + "loss_total_v6": 0.8063644170761108, + "loss_wm_recon": 0.296875, + "loss_wm_recon_w": 0.07421875, + "step": 234 + }, + { + "epoch": 0.08758852031308237, + "grad_norm": 21.407594680786133, + "learning_rate": 9.995098246916013e-06, + "loss": 12.575667381286621, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.2879383862018585, + "loss_total_v6": 0.5359852313995361, + "loss_wm_recon": 0.3203125, + "loss_wm_recon_w": 0.080078125, + "step": 235 + }, + { + "epoch": 0.08796123742079762, + "grad_norm": 22.166051864624023, + "learning_rate": 9.994814194428464e-06, + "loss": 12.763652801513672, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5285205245018005, + "loss_total_v6": 0.7511767745018005, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 236 + }, + { + "epoch": 0.08833395452851286, + "grad_norm": 23.362577438354492, + "learning_rate": 9.994522146073474e-06, + "loss": 12.53693962097168, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.38619017601013184, + "loss_total_v6": 0.6244714260101318, + "loss_wm_recon": 0.3046875, + "loss_wm_recon_w": 0.076171875, + "step": 237 + }, + { + "epoch": 0.0887066716362281, + "grad_norm": 25.006183624267578, + "learning_rate": 9.994222102318563e-06, + "loss": 12.737346649169922, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5608928799629211, + "loss_total_v6": 0.7859905362129211, + "loss_wm_recon": 0.291015625, + "loss_wm_recon_w": 0.07275390625, + "step": 238 + }, + { + "epoch": 0.08907938874394335, + "grad_norm": 22.245027542114258, + "learning_rate": 9.993914063644053e-06, + "loss": 12.432032585144043, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5079066753387451, + "loss_total_v6": 0.7461879253387451, + "loss_wm_recon": 0.3046875, + "loss_wm_recon_w": 0.076171875, + "step": 239 + }, + { + "epoch": 0.0894521058516586, + "grad_norm": 22.150999069213867, + "learning_rate": 9.993598030543061e-06, + "loss": 12.649764060974121, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.33987104892730713, + "loss_total_v6": 0.5713163614273071, + "loss_wm_recon": 0.29296875, + "loss_wm_recon_w": 0.0732421875, + "step": 240 + }, + { + "epoch": 0.08982482295937383, + "grad_norm": 21.68003273010254, + "learning_rate": 9.993274003521504e-06, + "loss": 12.598298072814941, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.797573447227478, + "loss_total_v6": 0.871303915977478, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 241 + }, + { + "epoch": 0.09019754006708908, + "grad_norm": 21.49036407470703, + "learning_rate": 9.992941983098097e-06, + "loss": 12.559039115905762, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.6220930814743042, + "loss_total_v6": 0.8696516752243042, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 242 + }, + { + "epoch": 0.09057025717480433, + "grad_norm": 21.469614028930664, + "learning_rate": 9.992601969804348e-06, + "loss": 12.602605819702148, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.9122849702835083, + "loss_total_v6": 1.1505662202835083, + "loss_wm_recon": 0.3046875, + "loss_wm_recon_w": 0.076171875, + "step": 243 + }, + { + "epoch": 0.09094297428251957, + "grad_norm": 21.618877410888672, + "learning_rate": 9.99225396418456e-06, + "loss": 12.428540229797363, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4956645667552948, + "loss_total_v6": 0.7192974090576172, + "loss_wm_recon": 0.28515625, + "loss_wm_recon_w": 0.0712890625, + "step": 244 + }, + { + "epoch": 0.09131569139023481, + "grad_norm": 22.066186904907227, + "learning_rate": 9.991897966795835e-06, + "loss": 12.236515045166016, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.722461462020874, + "loss_total_v6": 0.961719274520874, + "loss_wm_recon": 0.29296875, + "loss_wm_recon_w": 0.0732421875, + "step": 245 + }, + { + "epoch": 0.09168840849795006, + "grad_norm": 23.04609489440918, + "learning_rate": 9.991533978208065e-06, + "loss": 12.626861572265625, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.3625447154045105, + "loss_total_v6": 0.5676228404045105, + "loss_wm_recon": 0.2890625, + "loss_wm_recon_w": 0.072265625, + "step": 246 + }, + { + "epoch": 0.0920611256056653, + "grad_norm": 22.419076919555664, + "learning_rate": 9.991161999003936e-06, + "loss": 12.802434921264648, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6563497185707092, + "loss_total_v6": 0.7281270623207092, + "loss_wm_recon": 0.287109375, + "loss_wm_recon_w": 0.07177734375, + "step": 247 + }, + { + "epoch": 0.09243384271338054, + "grad_norm": 21.866819381713867, + "learning_rate": 9.990782029778922e-06, + "loss": 12.974115371704102, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5659776329994202, + "loss_total_v6": 0.7832627892494202, + "loss_wm_recon": 0.283203125, + "loss_wm_recon_w": 0.07080078125, + "step": 248 + }, + { + "epoch": 0.09280655982109579, + "grad_norm": 23.673913955688477, + "learning_rate": 9.990394071141295e-06, + "loss": 12.322891235351562, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6706579923629761, + "loss_total_v6": 0.8918493986129761, + "loss_wm_recon": 0.267578125, + "loss_wm_recon_w": 0.06689453125, + "step": 249 + }, + { + "epoch": 0.09317927692881103, + "grad_norm": 21.414785385131836, + "learning_rate": 9.989998123712108e-06, + "loss": 12.144288063049316, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.4495007395744324, + "loss_total_v6": 0.7053601145744324, + "loss_wm_recon": 0.3359375, + "loss_wm_recon_w": 0.083984375, + "step": 250 + }, + { + "epoch": 0.09355199403652628, + "grad_norm": 22.224939346313477, + "learning_rate": 9.98959418812521e-06, + "loss": 12.686963081359863, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.7249839901924133, + "loss_total_v6": 0.9720543026924133, + "loss_wm_recon": 0.30859375, + "loss_wm_recon_w": 0.0771484375, + "step": 251 + }, + { + "epoch": 0.09392471114424152, + "grad_norm": 22.3614444732666, + "learning_rate": 9.989182265027232e-06, + "loss": 12.290075302124023, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.956611156463623, + "loss_total_v6": 1.192451000213623, + "loss_wm_recon": 0.287109375, + "loss_wm_recon_w": 0.07177734375, + "step": 252 + }, + { + "epoch": 0.09429742825195676, + "grad_norm": 22.361797332763672, + "learning_rate": 9.988762355077597e-06, + "loss": 12.603588104248047, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6430270075798035, + "loss_total_v6": 0.8695895075798035, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 253 + }, + { + "epoch": 0.09467014535967201, + "grad_norm": 21.0738468170166, + "learning_rate": 9.98833445894851e-06, + "loss": 13.080114364624023, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.667120099067688, + "loss_total_v6": 0.881475567817688, + "loss_wm_recon": 0.279296875, + "loss_wm_recon_w": 0.06982421875, + "step": 254 + }, + { + "epoch": 0.09504286246738726, + "grad_norm": 22.302005767822266, + "learning_rate": 9.98789857732496e-06, + "loss": 12.471458435058594, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.444446861743927, + "loss_total_v6": 0.679798424243927, + "loss_wm_recon": 0.29296875, + "loss_wm_recon_w": 0.0732421875, + "step": 255 + }, + { + "epoch": 0.0954155795751025, + "grad_norm": 21.809253692626953, + "learning_rate": 9.987454710904725e-06, + "loss": 12.627166748046875, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3442191779613495, + "loss_total_v6": 0.5649223327636719, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 256 + }, + { + "epoch": 0.09578829668281774, + "grad_norm": 21.72726821899414, + "learning_rate": 9.987002860398358e-06, + "loss": 12.403301239013672, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.1950279176235199, + "loss_total_v6": 0.4142661988735199, + "loss_wm_recon": 0.283203125, + "loss_wm_recon_w": 0.07080078125, + "step": 257 + }, + { + "epoch": 0.09616101379053299, + "grad_norm": 21.33702850341797, + "learning_rate": 9.986543026529197e-06, + "loss": 13.07093620300293, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5365093946456909, + "loss_total_v6": 0.7630718946456909, + "loss_wm_recon": 0.2890625, + "loss_wm_recon_w": 0.072265625, + "step": 258 + }, + { + "epoch": 0.09653373089824822, + "grad_norm": 21.17206382751465, + "learning_rate": 9.986075210033362e-06, + "loss": 12.33121109008789, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3949192464351654, + "loss_total_v6": 0.6190403699874878, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 259 + }, + { + "epoch": 0.09690644800596347, + "grad_norm": 21.332971572875977, + "learning_rate": 9.985599411659746e-06, + "loss": 12.577994346618652, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5314092636108398, + "loss_total_v6": 0.7599248886108398, + "loss_wm_recon": 0.2890625, + "loss_wm_recon_w": 0.072265625, + "step": 260 + }, + { + "epoch": 0.09727916511367872, + "grad_norm": 21.226823806762695, + "learning_rate": 9.985115632170024e-06, + "loss": 12.905654907226562, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4050154983997345, + "loss_total_v6": 0.6340193748474121, + "loss_wm_recon": 0.291015625, + "loss_wm_recon_w": 0.07275390625, + "step": 261 + }, + { + "epoch": 0.09765188222139397, + "grad_norm": 22.024843215942383, + "learning_rate": 9.984623872338648e-06, + "loss": 12.026880264282227, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.39445972442626953, + "loss_total_v6": 0.6200456619262695, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 262 + }, + { + "epoch": 0.0980245993291092, + "grad_norm": 21.006839752197266, + "learning_rate": 9.984124132952845e-06, + "loss": 11.816646575927734, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5991682410240173, + "loss_total_v6": 0.8359846472740173, + "loss_wm_recon": 0.291015625, + "loss_wm_recon_w": 0.07275390625, + "step": 263 + }, + { + "epoch": 0.09839731643682445, + "grad_norm": 21.171478271484375, + "learning_rate": 9.983616414812611e-06, + "loss": 12.568696975708008, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.720537543296814, + "loss_total_v6": 0.967119574546814, + "loss_wm_recon": 0.314453125, + "loss_wm_recon_w": 0.07861328125, + "step": 264 + }, + { + "epoch": 0.0987700335445397, + "grad_norm": 22.077844619750977, + "learning_rate": 9.98310071873072e-06, + "loss": 11.996785163879395, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.9048190712928772, + "loss_total_v6": 1.1347994804382324, + "loss_wm_recon": 0.287109375, + "loss_wm_recon_w": 0.07177734375, + "step": 265 + }, + { + "epoch": 0.09914275065225493, + "grad_norm": 20.68411636352539, + "learning_rate": 9.982577045532717e-06, + "loss": 12.444887161254883, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6562925577163696, + "loss_total_v6": 0.9004331827163696, + "loss_wm_recon": 0.3203125, + "loss_wm_recon_w": 0.080078125, + "step": 266 + }, + { + "epoch": 0.09951546775997018, + "grad_norm": 21.153366088867188, + "learning_rate": 9.982045396056916e-06, + "loss": 12.374397277832031, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.8644108772277832, + "loss_total_v6": 0.9386296272277832, + "loss_wm_recon": 0.296875, + "loss_wm_recon_w": 0.07421875, + "step": 267 + }, + { + "epoch": 0.09988818486768543, + "grad_norm": 21.44272232055664, + "learning_rate": 9.981505771154395e-06, + "loss": 12.440423965454102, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.6869066953659058, + "loss_total_v6": 0.9486254453659058, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 268 + }, + { + "epoch": 0.10026090197540068, + "grad_norm": 21.38837242126465, + "learning_rate": 9.980958171689008e-06, + "loss": 12.792752265930176, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6318381428718567, + "loss_total_v6": 0.8574240803718567, + "loss_wm_recon": 0.28515625, + "loss_wm_recon_w": 0.0712890625, + "step": 269 + }, + { + "epoch": 0.10063361908311591, + "grad_norm": 22.50592613220215, + "learning_rate": 9.980402598537371e-06, + "loss": 12.367626190185547, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.9005463719367981, + "loss_total_v6": 1.1266205310821533, + "loss_wm_recon": 0.279296875, + "loss_wm_recon_w": 0.06982421875, + "step": 270 + }, + { + "epoch": 0.10100633619083116, + "grad_norm": 21.26771354675293, + "learning_rate": 9.979839052588863e-06, + "loss": 12.132875442504883, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7107276916503906, + "loss_total_v6": 0.9416847229003906, + "loss_wm_recon": 0.283203125, + "loss_wm_recon_w": 0.07080078125, + "step": 271 + }, + { + "epoch": 0.1013790532985464, + "grad_norm": 21.603654861450195, + "learning_rate": 9.979267534745624e-06, + "loss": 12.62224006652832, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5593388080596924, + "loss_total_v6": 0.7780888080596924, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 272 + }, + { + "epoch": 0.10175177040626164, + "grad_norm": 22.887916564941406, + "learning_rate": 9.978688045922566e-06, + "loss": 12.080484390258789, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5555355548858643, + "loss_total_v6": 0.7659847736358643, + "loss_wm_recon": 0.263671875, + "loss_wm_recon_w": 0.06591796875, + "step": 273 + }, + { + "epoch": 0.10212448751397689, + "grad_norm": 23.146591186523438, + "learning_rate": 9.97810058704735e-06, + "loss": 12.66623306274414, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.540515124797821, + "loss_total_v6": 0.606433093547821, + "loss_wm_recon": 0.263671875, + "loss_wm_recon_w": 0.06591796875, + "step": 274 + }, + { + "epoch": 0.10249720462169214, + "grad_norm": 20.49324607849121, + "learning_rate": 9.977505159060398e-06, + "loss": 12.297586441040039, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5831445455551147, + "loss_total_v6": 0.8228906393051147, + "loss_wm_recon": 0.287109375, + "loss_wm_recon_w": 0.07177734375, + "step": 275 + }, + { + "epoch": 0.10286992172940738, + "grad_norm": 20.566736221313477, + "learning_rate": 9.976901762914898e-06, + "loss": 12.487774848937988, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.7234305143356323, + "loss_total_v6": 0.9446219205856323, + "loss_wm_recon": 0.275390625, + "loss_wm_recon_w": 0.06884765625, + "step": 276 + }, + { + "epoch": 0.10324263883712262, + "grad_norm": 20.78065299987793, + "learning_rate": 9.97629039957678e-06, + "loss": 12.208181381225586, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.433227002620697, + "loss_total_v6": 0.657348096370697, + "loss_wm_recon": 0.279296875, + "loss_wm_recon_w": 0.06982421875, + "step": 277 + }, + { + "epoch": 0.10361535594483787, + "grad_norm": 20.90450096130371, + "learning_rate": 9.975671070024741e-06, + "loss": 12.16139030456543, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5153608918190002, + "loss_total_v6": 0.7492476105690002, + "loss_wm_recon": 0.279296875, + "loss_wm_recon_w": 0.06982421875, + "step": 278 + }, + { + "epoch": 0.10398807305255311, + "grad_norm": 20.74785804748535, + "learning_rate": 9.97504377525022e-06, + "loss": 12.262571334838867, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.7375285029411316, + "loss_total_v6": 0.9636027216911316, + "loss_wm_recon": 0.279296875, + "loss_wm_recon_w": 0.06982421875, + "step": 279 + }, + { + "epoch": 0.10436079016026836, + "grad_norm": 23.0623722076416, + "learning_rate": 9.974408516257414e-06, + "loss": 11.919322967529297, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5344169735908508, + "loss_total_v6": 0.7595146298408508, + "loss_wm_recon": 0.275390625, + "loss_wm_recon_w": 0.06884765625, + "step": 280 + }, + { + "epoch": 0.1047335072679836, + "grad_norm": 21.41135597229004, + "learning_rate": 9.973765294063265e-06, + "loss": 12.1989107131958, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.35466909408569336, + "loss_total_v6": 0.5783019065856934, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 281 + }, + { + "epoch": 0.10510622437569885, + "grad_norm": 20.596590042114258, + "learning_rate": 9.973114109697468e-06, + "loss": 11.946381568908691, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5732504725456238, + "loss_total_v6": 0.8081137537956238, + "loss_wm_recon": 0.283203125, + "loss_wm_recon_w": 0.07080078125, + "step": 282 + }, + { + "epoch": 0.1054789414834141, + "grad_norm": 21.69835090637207, + "learning_rate": 9.972454964202458e-06, + "loss": 12.506143569946289, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6380481123924255, + "loss_total_v6": 0.7078723311424255, + "loss_wm_recon": 0.279296875, + "loss_wm_recon_w": 0.06982421875, + "step": 283 + }, + { + "epoch": 0.10585165859112933, + "grad_norm": 21.213014602661133, + "learning_rate": 9.971787858633419e-06, + "loss": 12.058849334716797, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.24754497408866882, + "loss_total_v6": 0.4594590365886688, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 284 + }, + { + "epoch": 0.10622437569884458, + "grad_norm": 22.404529571533203, + "learning_rate": 9.971112794058278e-06, + "loss": 12.113470077514648, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3160185217857361, + "loss_total_v6": 0.5435575842857361, + "loss_wm_recon": 0.27734375, + "loss_wm_recon_w": 0.0693359375, + "step": 285 + }, + { + "epoch": 0.10659709280655982, + "grad_norm": 21.404861450195312, + "learning_rate": 9.970429771557697e-06, + "loss": 12.115665435791016, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.36168190836906433, + "loss_total_v6": 0.43492409586906433, + "loss_wm_recon": 0.29296875, + "loss_wm_recon_w": 0.0732421875, + "step": 286 + }, + { + "epoch": 0.10696980991427507, + "grad_norm": 21.448774337768555, + "learning_rate": 9.969738792225083e-06, + "loss": 12.146100997924805, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.11060626059770584, + "loss_total_v6": 0.33814531564712524, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 287 + }, + { + "epoch": 0.1073425270219903, + "grad_norm": 22.138656616210938, + "learning_rate": 9.96903985716658e-06, + "loss": 12.250268936157227, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5453094244003296, + "loss_total_v6": 0.7791961431503296, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 288 + }, + { + "epoch": 0.10771524412970555, + "grad_norm": 22.20417022705078, + "learning_rate": 9.968332967501068e-06, + "loss": 12.751657485961914, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.5965635180473328, + "loss_total_v6": 0.8387510180473328, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 289 + }, + { + "epoch": 0.1080879612374208, + "grad_norm": 21.903461456298828, + "learning_rate": 9.967618124360156e-06, + "loss": 12.342233657836914, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5155751705169678, + "loss_total_v6": 0.7499501705169678, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 290 + }, + { + "epoch": 0.10846067834513604, + "grad_norm": 22.334007263183594, + "learning_rate": 9.966895328888195e-06, + "loss": 12.13066291809082, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5323822498321533, + "loss_total_v6": 0.7501556873321533, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 291 + }, + { + "epoch": 0.10883339545285128, + "grad_norm": 21.83871841430664, + "learning_rate": 9.966164582242255e-06, + "loss": 12.289609909057617, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5046919584274292, + "loss_total_v6": 0.7214888334274292, + "loss_wm_recon": 0.265625, + "loss_wm_recon_w": 0.06640625, + "step": 292 + }, + { + "epoch": 0.10920611256056653, + "grad_norm": 21.91131591796875, + "learning_rate": 9.96542588559214e-06, + "loss": 12.04883098602295, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.5494590401649475, + "loss_total_v6": 0.7535606026649475, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 293 + }, + { + "epoch": 0.10957882966828178, + "grad_norm": 22.931089401245117, + "learning_rate": 9.964679240120385e-06, + "loss": 12.33586311340332, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.9100271463394165, + "loss_total_v6": 1.1321951150894165, + "loss_wm_recon": 0.263671875, + "loss_wm_recon_w": 0.06591796875, + "step": 294 + }, + { + "epoch": 0.10995154677599701, + "grad_norm": 19.143795013427734, + "learning_rate": 9.963924647022241e-06, + "loss": 12.559059143066406, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.8562279939651489, + "loss_total_v6": 1.112087368965149, + "loss_wm_recon": 0.328125, + "loss_wm_recon_w": 0.08203125, + "step": 295 + }, + { + "epoch": 0.11032426388371226, + "grad_norm": 21.649147033691406, + "learning_rate": 9.963162107505688e-06, + "loss": 12.238077163696289, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.41811510920524597, + "loss_total_v6": 0.6402831077575684, + "loss_wm_recon": 0.271484375, + "loss_wm_recon_w": 0.06787109375, + "step": 296 + }, + { + "epoch": 0.11069698099142751, + "grad_norm": 22.226049423217773, + "learning_rate": 9.962391622791427e-06, + "loss": 12.379091262817383, + "loss_alignment": 0.35546875, + "loss_alignment_w": 0.177734375, + "loss_sft": 0.6402246356010437, + "loss_total_v6": 0.8960840106010437, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 297 + }, + { + "epoch": 0.11106969809914274, + "grad_norm": 20.858909606933594, + "learning_rate": 9.961613194112872e-06, + "loss": 12.289190292358398, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.9326569437980652, + "loss_total_v6": 1.16996169090271, + "loss_wm_recon": 0.28515625, + "loss_wm_recon_w": 0.0712890625, + "step": 298 + }, + { + "epoch": 0.11144241520685799, + "grad_norm": 23.03713035583496, + "learning_rate": 9.96082682271616e-06, + "loss": 12.25462532043457, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.41840577125549316, + "loss_total_v6": 0.6395971775054932, + "loss_wm_recon": 0.267578125, + "loss_wm_recon_w": 0.06689453125, + "step": 299 + }, + { + "epoch": 0.11181513231457324, + "grad_norm": 21.077634811401367, + "learning_rate": 9.96003250986014e-06, + "loss": 12.336955070495605, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.613166868686676, + "loss_total_v6": 0.832893431186676, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 300 + }, + { + "epoch": 0.11181513231457324, + "eval_loss": 0.7240904569625854, + "eval_loss_alignment": 0.2928539830395299, + "eval_loss_alignment_w": 0.14642699151976496, + "eval_loss_sft": 0.5112465515446204, + "eval_loss_total_v6": 0.7240904873698695, + "eval_loss_wm_recon": 0.2656677767761752, + "eval_loss_wm_recon_w": 0.0664169441940438, + "eval_runtime": 227.9079, + "eval_samples_per_second": 8.188, + "eval_steps_per_second": 1.027, + "step": 300 + }, + { + "epoch": 0.11218784942228849, + "grad_norm": 21.422094345092773, + "learning_rate": 9.959230256816378e-06, + "loss": 12.991676330566406, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6276043057441711, + "loss_total_v6": 0.8468425869941711, + "loss_wm_recon": 0.267578125, + "loss_wm_recon_w": 0.06689453125, + "step": 301 + }, + { + "epoch": 0.11256056653000372, + "grad_norm": 20.93741226196289, + "learning_rate": 9.958420064869143e-06, + "loss": 12.909332275390625, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5944615006446838, + "loss_total_v6": 0.8210240006446838, + "loss_wm_recon": 0.2578125, + "loss_wm_recon_w": 0.064453125, + "step": 302 + }, + { + "epoch": 0.11293328363771897, + "grad_norm": 20.815046310424805, + "learning_rate": 9.957601935315422e-06, + "loss": 12.49583911895752, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3242965042591095, + "loss_total_v6": 0.5513472557067871, + "loss_wm_recon": 0.275390625, + "loss_wm_recon_w": 0.06884765625, + "step": 303 + }, + { + "epoch": 0.11330600074543422, + "grad_norm": 21.63507080078125, + "learning_rate": 9.956775869464901e-06, + "loss": 12.461250305175781, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.7359318137168884, + "loss_total_v6": 0.9532169699668884, + "loss_wm_recon": 0.259765625, + "loss_wm_recon_w": 0.06494140625, + "step": 304 + }, + { + "epoch": 0.11367871785314947, + "grad_norm": 21.413124084472656, + "learning_rate": 9.955941868639976e-06, + "loss": 12.051750183105469, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6972523331642151, + "loss_total_v6": 0.9213734269142151, + "loss_wm_recon": 0.271484375, + "loss_wm_recon_w": 0.06787109375, + "step": 305 + }, + { + "epoch": 0.1140514349608647, + "grad_norm": 21.251787185668945, + "learning_rate": 9.955099934175746e-06, + "loss": 12.31682300567627, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4654812812805176, + "loss_total_v6": 0.6842312812805176, + "loss_wm_recon": 0.265625, + "loss_wm_recon_w": 0.06640625, + "step": 306 + }, + { + "epoch": 0.11442415206857995, + "grad_norm": 20.960660934448242, + "learning_rate": 9.954250067420001e-06, + "loss": 12.427453994750977, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4744947552680969, + "loss_total_v6": 0.6966627240180969, + "loss_wm_recon": 0.263671875, + "loss_wm_recon_w": 0.06591796875, + "step": 307 + }, + { + "epoch": 0.1147968691762952, + "grad_norm": 21.25204086303711, + "learning_rate": 9.953392269733244e-06, + "loss": 12.634344100952148, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5709512829780579, + "loss_total_v6": 0.7945840954780579, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 308 + }, + { + "epoch": 0.11516958628401043, + "grad_norm": 21.066747665405273, + "learning_rate": 9.952526542488663e-06, + "loss": 12.198442459106445, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.3778933882713318, + "loss_total_v6": 0.4413699507713318, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 309 + }, + { + "epoch": 0.11554230339172568, + "grad_norm": 20.630470275878906, + "learning_rate": 9.951652887072142e-06, + "loss": 12.245556831359863, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7141732573509216, + "loss_total_v6": 0.9451302886009216, + "loss_wm_recon": 0.275390625, + "loss_wm_recon_w": 0.06884765625, + "step": 310 + }, + { + "epoch": 0.11591502049944093, + "grad_norm": 20.893924713134766, + "learning_rate": 9.950771304882262e-06, + "loss": 12.152935028076172, + "loss_alignment": 0.375, + "loss_alignment_w": 0.1875, + "loss_sft": 0.39906740188598633, + "loss_total_v6": 0.6617627143859863, + "loss_wm_recon": 0.30078125, + "loss_wm_recon_w": 0.0751953125, + "step": 311 + }, + { + "epoch": 0.11628773760715617, + "grad_norm": 20.71193504333496, + "learning_rate": 9.94988179733029e-06, + "loss": 12.560317993164062, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5657915472984314, + "loss_total_v6": 0.7860063910484314, + "loss_wm_recon": 0.263671875, + "loss_wm_recon_w": 0.06591796875, + "step": 312 + }, + { + "epoch": 0.11666045471487141, + "grad_norm": 20.707637786865234, + "learning_rate": 9.948984365840174e-06, + "loss": 11.987068176269531, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6398442983627319, + "loss_total_v6": 0.8698247671127319, + "loss_wm_recon": 0.279296875, + "loss_wm_recon_w": 0.06982421875, + "step": 313 + }, + { + "epoch": 0.11703317182258666, + "grad_norm": 20.86467933654785, + "learning_rate": 9.948079011848557e-06, + "loss": 11.787416458129883, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6255500316619873, + "loss_total_v6": 0.8535773754119873, + "loss_wm_recon": 0.271484375, + "loss_wm_recon_w": 0.06787109375, + "step": 314 + }, + { + "epoch": 0.1174058889303019, + "grad_norm": 21.74762725830078, + "learning_rate": 9.94716573680476e-06, + "loss": 11.972803115844727, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5662618279457092, + "loss_total_v6": 0.8123555779457092, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 315 + }, + { + "epoch": 0.11777860603801714, + "grad_norm": 21.919052124023438, + "learning_rate": 9.946244542170784e-06, + "loss": 12.0615816116333, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.4694826006889343, + "loss_total_v6": 0.6950685381889343, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 316 + }, + { + "epoch": 0.11815132314573239, + "grad_norm": 20.477672576904297, + "learning_rate": 9.945315429421307e-06, + "loss": 12.565006256103516, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6485776901245117, + "loss_total_v6": 0.8751401901245117, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 317 + }, + { + "epoch": 0.11852404025344764, + "grad_norm": 20.043563842773438, + "learning_rate": 9.944378400043685e-06, + "loss": 12.202713966369629, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6457622647285461, + "loss_total_v6": 0.8693950772285461, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 318 + }, + { + "epoch": 0.11889675736116288, + "grad_norm": 20.6641845703125, + "learning_rate": 9.94343345553795e-06, + "loss": 11.578556060791016, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5840935707092285, + "loss_total_v6": 0.8116326332092285, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 319 + }, + { + "epoch": 0.11926947446887812, + "grad_norm": 20.743371963500977, + "learning_rate": 9.942480597416797e-06, + "loss": 12.102119445800781, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.32354652881622314, + "loss_total_v6": 0.39385902881622314, + "loss_wm_recon": 0.28125, + "loss_wm_recon_w": 0.0703125, + "step": 320 + }, + { + "epoch": 0.11964219157659337, + "grad_norm": 20.15297508239746, + "learning_rate": 9.941519827205595e-06, + "loss": 11.748422622680664, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.4959845542907715, + "loss_total_v6": 0.7220587730407715, + "loss_wm_recon": 0.271484375, + "loss_wm_recon_w": 0.06787109375, + "step": 321 + }, + { + "epoch": 0.12001490868430861, + "grad_norm": 20.712860107421875, + "learning_rate": 9.940551146442375e-06, + "loss": 12.449319839477539, + "loss_alignment": 0.37109375, + "loss_alignment_w": 0.185546875, + "loss_sft": 0.6844797730445862, + "loss_total_v6": 0.9481516480445862, + "loss_wm_recon": 0.3125, + "loss_wm_recon_w": 0.078125, + "step": 322 + }, + { + "epoch": 0.12038762579202385, + "grad_norm": 20.703060150146484, + "learning_rate": 9.939574556677837e-06, + "loss": 11.986990928649902, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.7118476629257202, + "loss_total_v6": 0.9174140691757202, + "loss_wm_recon": 0.267578125, + "loss_wm_recon_w": 0.06689453125, + "step": 323 + }, + { + "epoch": 0.1207603428997391, + "grad_norm": 20.463157653808594, + "learning_rate": 9.938590059475338e-06, + "loss": 11.939722061157227, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5246220231056213, + "loss_total_v6": 0.7304325699806213, + "loss_wm_recon": 0.2451171875, + "loss_wm_recon_w": 0.061279296875, + "step": 324 + }, + { + "epoch": 0.12113306000745434, + "grad_norm": 20.501157760620117, + "learning_rate": 9.937597656410893e-06, + "loss": 12.803718566894531, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.4929143190383911, + "loss_total_v6": 0.7175236940383911, + "loss_wm_recon": 0.265625, + "loss_wm_recon_w": 0.06640625, + "step": 325 + }, + { + "epoch": 0.12150577711516959, + "grad_norm": 20.101604461669922, + "learning_rate": 9.936597349073176e-06, + "loss": 12.350597381591797, + "loss_alignment": 0.35546875, + "loss_alignment_w": 0.177734375, + "loss_sft": 0.4521731734275818, + "loss_total_v6": 0.6928958296775818, + "loss_wm_recon": 0.251953125, + "loss_wm_recon_w": 0.06298828125, + "step": 326 + }, + { + "epoch": 0.12187849422288483, + "grad_norm": 19.733184814453125, + "learning_rate": 9.935589139063513e-06, + "loss": 13.017520904541016, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4355553984642029, + "loss_total_v6": 0.6582116484642029, + "loss_wm_recon": 0.265625, + "loss_wm_recon_w": 0.06640625, + "step": 327 + }, + { + "epoch": 0.12225121133060007, + "grad_norm": 20.318613052368164, + "learning_rate": 9.93457302799588e-06, + "loss": 11.797765731811523, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5645061135292053, + "loss_total_v6": 0.7847209572792053, + "loss_wm_recon": 0.263671875, + "loss_wm_recon_w": 0.06591796875, + "step": 328 + }, + { + "epoch": 0.12262392843831532, + "grad_norm": 20.49799156188965, + "learning_rate": 9.933549017496902e-06, + "loss": 12.0006685256958, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.3406004011631012, + "loss_total_v6": 0.5705808401107788, + "loss_wm_recon": 0.271484375, + "loss_wm_recon_w": 0.06787109375, + "step": 329 + }, + { + "epoch": 0.12299664554603057, + "grad_norm": 20.314592361450195, + "learning_rate": 9.932517109205849e-06, + "loss": 12.272552490234375, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.8431504964828491, + "loss_total_v6": 1.0594590902328491, + "loss_wm_recon": 0.255859375, + "loss_wm_recon_w": 0.06396484375, + "step": 330 + }, + { + "epoch": 0.1233693626537458, + "grad_norm": 21.07034683227539, + "learning_rate": 9.931477304774637e-06, + "loss": 12.205658912658691, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4902617037296295, + "loss_total_v6": 0.7178007364273071, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 331 + }, + { + "epoch": 0.12374207976146105, + "grad_norm": 21.63747215270996, + "learning_rate": 9.930429605867817e-06, + "loss": 11.787710189819336, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6309202909469604, + "loss_total_v6": 0.8604124784469604, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 332 + }, + { + "epoch": 0.1241147968691763, + "grad_norm": 20.163867950439453, + "learning_rate": 9.929374014162581e-06, + "loss": 11.985116958618164, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5765488743782043, + "loss_total_v6": 0.7957871556282043, + "loss_wm_recon": 0.275390625, + "loss_wm_recon_w": 0.06884765625, + "step": 333 + }, + { + "epoch": 0.12448751397689153, + "grad_norm": 20.310298919677734, + "learning_rate": 9.928310531348759e-06, + "loss": 12.412653923034668, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5286705493927002, + "loss_total_v6": 0.7518150806427002, + "loss_wm_recon": 0.259765625, + "loss_wm_recon_w": 0.06494140625, + "step": 334 + }, + { + "epoch": 0.12486023108460678, + "grad_norm": 19.936565399169922, + "learning_rate": 9.927239159128807e-06, + "loss": 12.50228500366211, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.20314006507396698, + "loss_total_v6": 0.4143217206001282, + "loss_wm_recon": 0.2431640625, + "loss_wm_recon_w": 0.060791015625, + "step": 335 + }, + { + "epoch": 0.12523294819232203, + "grad_norm": 20.087875366210938, + "learning_rate": 9.926159899217818e-06, + "loss": 12.238807678222656, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.7800917625427246, + "loss_total_v6": 1.0105605125427246, + "loss_wm_recon": 0.265625, + "loss_wm_recon_w": 0.06640625, + "step": 336 + }, + { + "epoch": 0.12560566530003728, + "grad_norm": 20.20836067199707, + "learning_rate": 9.925072753343501e-06, + "loss": 11.485105514526367, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.40336504578590393, + "loss_total_v6": 0.6201618909835815, + "loss_wm_recon": 0.2578125, + "loss_wm_recon_w": 0.064453125, + "step": 337 + }, + { + "epoch": 0.12597838240775253, + "grad_norm": 20.998056411743164, + "learning_rate": 9.923977723246203e-06, + "loss": 11.740907669067383, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.7333002686500549, + "loss_total_v6": 0.9647455811500549, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 338 + }, + { + "epoch": 0.12635109951546775, + "grad_norm": 21.360801696777344, + "learning_rate": 9.922874810678881e-06, + "loss": 12.207876205444336, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.38291412591934204, + "loss_total_v6": 0.589945375919342, + "loss_wm_recon": 0.25, + "loss_wm_recon_w": 0.0625, + "step": 339 + }, + { + "epoch": 0.126723816623183, + "grad_norm": 20.662273406982422, + "learning_rate": 9.921764017407117e-06, + "loss": 12.065452575683594, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.4327288866043091, + "loss_total_v6": 0.6636859178543091, + "loss_wm_recon": 0.267578125, + "loss_wm_recon_w": 0.06689453125, + "step": 340 + }, + { + "epoch": 0.12709653373089824, + "grad_norm": 24.190149307250977, + "learning_rate": 9.920645345209104e-06, + "loss": 11.960832595825195, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6138558983802795, + "loss_total_v6": 0.8335824608802795, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 341 + }, + { + "epoch": 0.1274692508386135, + "grad_norm": 20.04133415222168, + "learning_rate": 9.919518795875652e-06, + "loss": 12.604628562927246, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5255618095397949, + "loss_total_v6": 0.7413821220397949, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 342 + }, + { + "epoch": 0.12784196794632874, + "grad_norm": 21.60888671875, + "learning_rate": 9.918384371210178e-06, + "loss": 12.229970932006836, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.7310980558395386, + "loss_total_v6": 0.9410589933395386, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 343 + }, + { + "epoch": 0.128214685054044, + "grad_norm": 19.360187530517578, + "learning_rate": 9.917242073028708e-06, + "loss": 12.158677101135254, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4960704743862152, + "loss_total_v6": 0.7280040979385376, + "loss_wm_recon": 0.287109375, + "loss_wm_recon_w": 0.07177734375, + "step": 344 + }, + { + "epoch": 0.12858740216175923, + "grad_norm": 20.8583927154541, + "learning_rate": 9.916091903159873e-06, + "loss": 12.446507453918457, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4200582504272461, + "loss_total_v6": 0.6266012191772461, + "loss_wm_recon": 0.240234375, + "loss_wm_recon_w": 0.06005859375, + "step": 345 + }, + { + "epoch": 0.12896011926947448, + "grad_norm": 19.43974494934082, + "learning_rate": 9.9149338634449e-06, + "loss": 11.895498275756836, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6556217670440674, + "loss_total_v6": 0.8753483295440674, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 346 + }, + { + "epoch": 0.1293328363771897, + "grad_norm": 20.461822509765625, + "learning_rate": 9.913767955737623e-06, + "loss": 11.7334566116333, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5405284762382507, + "loss_total_v6": 0.7622081637382507, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 347 + }, + { + "epoch": 0.12970555348490495, + "grad_norm": 20.917156219482422, + "learning_rate": 9.912594181904463e-06, + "loss": 12.372530937194824, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5533335208892822, + "loss_total_v6": 0.7554819583892822, + "loss_wm_recon": 0.23046875, + "loss_wm_recon_w": 0.0576171875, + "step": 348 + }, + { + "epoch": 0.1300782705926202, + "grad_norm": 21.15930938720703, + "learning_rate": 9.911412543824438e-06, + "loss": 12.267857551574707, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5928500294685364, + "loss_total_v6": 0.8125765919685364, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 349 + }, + { + "epoch": 0.13045098770033545, + "grad_norm": 20.64656639099121, + "learning_rate": 9.910223043389156e-06, + "loss": 12.281024932861328, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.7088293433189392, + "loss_total_v6": 0.9400305151939392, + "loss_wm_recon": 0.2373046875, + "loss_wm_recon_w": 0.059326171875, + "step": 350 + }, + { + "epoch": 0.1308237048080507, + "grad_norm": 21.0394287109375, + "learning_rate": 9.909025682502806e-06, + "loss": 12.331860542297363, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.4979783296585083, + "loss_total_v6": 0.7401658296585083, + "loss_wm_recon": 0.2734375, + "loss_wm_recon_w": 0.068359375, + "step": 351 + }, + { + "epoch": 0.13119642191576594, + "grad_norm": 20.61329460144043, + "learning_rate": 9.907820463082165e-06, + "loss": 11.975004196166992, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6123764514923096, + "loss_total_v6": 0.8228256702423096, + "loss_wm_recon": 0.255859375, + "loss_wm_recon_w": 0.06396484375, + "step": 352 + }, + { + "epoch": 0.1315691390234812, + "grad_norm": 19.868330001831055, + "learning_rate": 9.906607387056591e-06, + "loss": 12.458093643188477, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.32624495029449463, + "loss_total_v6": 0.5508543252944946, + "loss_wm_recon": 0.265625, + "loss_wm_recon_w": 0.06640625, + "step": 353 + }, + { + "epoch": 0.1319418561311964, + "grad_norm": 19.372455596923828, + "learning_rate": 9.905386456368013e-06, + "loss": 12.276512145996094, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.8648515939712524, + "loss_total_v6": 1.0914140939712524, + "loss_wm_recon": 0.2578125, + "loss_wm_recon_w": 0.064453125, + "step": 354 + }, + { + "epoch": 0.13231457323891166, + "grad_norm": 19.88456916809082, + "learning_rate": 9.904157672970943e-06, + "loss": 12.340343475341797, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.30468031764030457, + "loss_total_v6": 0.536125659942627, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 355 + }, + { + "epoch": 0.1326872903466269, + "grad_norm": 20.985422134399414, + "learning_rate": 9.902921038832456e-06, + "loss": 12.336387634277344, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.636760413646698, + "loss_total_v6": 0.865764319896698, + "loss_wm_recon": 0.267578125, + "loss_wm_recon_w": 0.06689453125, + "step": 356 + }, + { + "epoch": 0.13306000745434216, + "grad_norm": 19.725553512573242, + "learning_rate": 9.901676555932199e-06, + "loss": 11.96697998046875, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.6716976761817932, + "loss_total_v6": 0.9046078324317932, + "loss_wm_recon": 0.267578125, + "loss_wm_recon_w": 0.06689453125, + "step": 357 + }, + { + "epoch": 0.1334327245620574, + "grad_norm": 19.211585998535156, + "learning_rate": 9.90042422626238e-06, + "loss": 12.029427528381348, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6090608239173889, + "loss_total_v6": 0.8351350426673889, + "loss_wm_recon": 0.263671875, + "loss_wm_recon_w": 0.06591796875, + "step": 358 + }, + { + "epoch": 0.13380544166977265, + "grad_norm": 18.95944595336914, + "learning_rate": 9.899164051827774e-06, + "loss": 11.79068374633789, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.26174047589302063, + "loss_total_v6": 0.48927953839302063, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 359 + }, + { + "epoch": 0.1341781587774879, + "grad_norm": 19.157054901123047, + "learning_rate": 9.897896034645709e-06, + "loss": 11.787732124328613, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4500651955604553, + "loss_total_v6": 0.6678386330604553, + "loss_wm_recon": 0.24609375, + "loss_wm_recon_w": 0.0615234375, + "step": 360 + }, + { + "epoch": 0.13455087588520312, + "grad_norm": 20.51832389831543, + "learning_rate": 9.89662017674607e-06, + "loss": 12.151185989379883, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6958073377609253, + "loss_total_v6": 0.9169987440109253, + "loss_wm_recon": 0.244140625, + "loss_wm_recon_w": 0.06103515625, + "step": 361 + }, + { + "epoch": 0.13492359299291837, + "grad_norm": 20.33315658569336, + "learning_rate": 9.895336480171294e-06, + "loss": 12.789889335632324, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6346845030784607, + "loss_total_v6": 0.8585614562034607, + "loss_wm_recon": 0.2470703125, + "loss_wm_recon_w": 0.061767578125, + "step": 362 + }, + { + "epoch": 0.13529631010063362, + "grad_norm": 20.2263240814209, + "learning_rate": 9.894044946976365e-06, + "loss": 12.23166275024414, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5110164284706116, + "loss_total_v6": 0.7273250222206116, + "loss_wm_recon": 0.255859375, + "loss_wm_recon_w": 0.06396484375, + "step": 363 + }, + { + "epoch": 0.13566902720834886, + "grad_norm": 20.874488830566406, + "learning_rate": 9.892745579228811e-06, + "loss": 11.821637153625488, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.37401705980300903, + "loss_total_v6": 0.588616669178009, + "loss_wm_recon": 0.2490234375, + "loss_wm_recon_w": 0.062255859375, + "step": 364 + }, + { + "epoch": 0.1360417443160641, + "grad_norm": 20.058595657348633, + "learning_rate": 9.89143837900871e-06, + "loss": 12.051586151123047, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5493885278701782, + "loss_total_v6": 0.7656971216201782, + "loss_wm_recon": 0.248046875, + "loss_wm_recon_w": 0.06201171875, + "step": 365 + }, + { + "epoch": 0.13641446142377936, + "grad_norm": 20.292173385620117, + "learning_rate": 9.890123348408666e-06, + "loss": 12.308347702026367, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.7712271213531494, + "loss_total_v6": 0.9909536838531494, + "loss_wm_recon": 0.25390625, + "loss_wm_recon_w": 0.0634765625, + "step": 366 + }, + { + "epoch": 0.1367871785314946, + "grad_norm": 20.306995391845703, + "learning_rate": 9.888800489533828e-06, + "loss": 11.908998489379883, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.43627288937568665, + "loss_total_v6": 0.6535580158233643, + "loss_wm_recon": 0.244140625, + "loss_wm_recon_w": 0.06103515625, + "step": 367 + }, + { + "epoch": 0.13715989563920983, + "grad_norm": 18.403589248657227, + "learning_rate": 9.887469804501873e-06, + "loss": 11.59951400756836, + "loss_alignment": 0.359375, + "loss_alignment_w": 0.1796875, + "loss_sft": 0.5809271931648254, + "loss_total_v6": 0.8362982869148254, + "loss_wm_recon": 0.302734375, + "loss_wm_recon_w": 0.07568359375, + "step": 368 + }, + { + "epoch": 0.13753261274692508, + "grad_norm": 20.2894287109375, + "learning_rate": 9.886131295443003e-06, + "loss": 12.272658348083496, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6343202590942383, + "loss_total_v6": 0.8423280715942383, + "loss_wm_recon": 0.23828125, + "loss_wm_recon_w": 0.0595703125, + "step": 369 + }, + { + "epoch": 0.13790532985464032, + "grad_norm": 20.991592407226562, + "learning_rate": 9.884784964499954e-06, + "loss": 12.3206787109375, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5430059432983398, + "loss_total_v6": 0.7532110214233398, + "loss_wm_recon": 0.2314453125, + "loss_wm_recon_w": 0.057861328125, + "step": 370 + }, + { + "epoch": 0.13827804696235557, + "grad_norm": 20.66351318359375, + "learning_rate": 9.883430813827975e-06, + "loss": 12.339655876159668, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5223362445831299, + "loss_total_v6": 0.7335178852081299, + "loss_wm_recon": 0.2431640625, + "loss_wm_recon_w": 0.060791015625, + "step": 371 + }, + { + "epoch": 0.13865076407007082, + "grad_norm": 21.642026901245117, + "learning_rate": 9.882068845594835e-06, + "loss": 11.998247146606445, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.7617210745811462, + "loss_total_v6": 0.9990257620811462, + "loss_wm_recon": 0.27734375, + "loss_wm_recon_w": 0.0693359375, + "step": 372 + }, + { + "epoch": 0.13902348117778607, + "grad_norm": 20.49407196044922, + "learning_rate": 9.880699061980822e-06, + "loss": 12.049837112426758, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.8492391109466553, + "loss_total_v6": 1.0714070796966553, + "loss_wm_recon": 0.240234375, + "loss_wm_recon_w": 0.06005859375, + "step": 373 + }, + { + "epoch": 0.13939619828550132, + "grad_norm": 21.54166030883789, + "learning_rate": 9.879321465178729e-06, + "loss": 11.597555160522461, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.40363916754722595, + "loss_total_v6": 0.6280044317245483, + "loss_wm_recon": 0.2412109375, + "loss_wm_recon_w": 0.060302734375, + "step": 374 + }, + { + "epoch": 0.13976891539321654, + "grad_norm": 19.49249267578125, + "learning_rate": 9.877936057393862e-06, + "loss": 12.454745292663574, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.6079155802726746, + "loss_total_v6": 0.8110405802726746, + "loss_wm_recon": 0.234375, + "loss_wm_recon_w": 0.05859375, + "step": 375 + }, + { + "epoch": 0.14014163250093178, + "grad_norm": 21.041847229003906, + "learning_rate": 9.876542840844027e-06, + "loss": 12.049724578857422, + "loss_alignment": 0.3828125, + "loss_alignment_w": 0.19140625, + "loss_sft": 0.6714504361152649, + "loss_total_v6": 0.9404934048652649, + "loss_wm_recon": 0.310546875, + "loss_wm_recon_w": 0.07763671875, + "step": 376 + }, + { + "epoch": 0.14051434960864703, + "grad_norm": 21.031211853027344, + "learning_rate": 9.875141817759533e-06, + "loss": 12.063699722290039, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.7531264424324036, + "loss_total_v6": 0.9611342549324036, + "loss_wm_recon": 0.23828125, + "loss_wm_recon_w": 0.0595703125, + "step": 377 + }, + { + "epoch": 0.14088706671636228, + "grad_norm": 20.7611141204834, + "learning_rate": 9.873732990383185e-06, + "loss": 12.624868392944336, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5715785026550293, + "loss_total_v6": 0.7866663932800293, + "loss_wm_recon": 0.2353515625, + "loss_wm_recon_w": 0.058837890625, + "step": 378 + }, + { + "epoch": 0.14125978382407753, + "grad_norm": 19.78968620300293, + "learning_rate": 9.87231636097028e-06, + "loss": 11.984779357910156, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.46161356568336487, + "loss_total_v6": 0.6791428327560425, + "loss_wm_recon": 0.2451171875, + "loss_wm_recon_w": 0.061279296875, + "step": 379 + }, + { + "epoch": 0.14163250093179278, + "grad_norm": 19.57486915588379, + "learning_rate": 9.870891931788608e-06, + "loss": 11.94521713256836, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6782075762748718, + "loss_total_v6": 0.8849946856498718, + "loss_wm_recon": 0.2412109375, + "loss_wm_recon_w": 0.060302734375, + "step": 380 + }, + { + "epoch": 0.14200521803950802, + "grad_norm": 20.943950653076172, + "learning_rate": 9.869459705118444e-06, + "loss": 11.954065322875977, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4474862217903137, + "loss_total_v6": 0.6620858311653137, + "loss_wm_recon": 0.2490234375, + "loss_wm_recon_w": 0.062255859375, + "step": 381 + }, + { + "epoch": 0.14237793514722324, + "grad_norm": 19.883495330810547, + "learning_rate": 9.868019683252543e-06, + "loss": 11.790205001831055, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4381401538848877, + "loss_total_v6": 0.6524956226348877, + "loss_wm_recon": 0.240234375, + "loss_wm_recon_w": 0.06005859375, + "step": 382 + }, + { + "epoch": 0.1427506522549385, + "grad_norm": 19.470247268676758, + "learning_rate": 9.866571868496142e-06, + "loss": 11.615494728088379, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.18100082874298096, + "loss_total_v6": 0.40170395374298096, + "loss_wm_recon": 0.265625, + "loss_wm_recon_w": 0.06640625, + "step": 383 + }, + { + "epoch": 0.14312336936265374, + "grad_norm": 19.73175621032715, + "learning_rate": 9.86511626316695e-06, + "loss": 11.714481353759766, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.7255929708480835, + "loss_total_v6": 0.9560617208480835, + "loss_wm_recon": 0.2578125, + "loss_wm_recon_w": 0.064453125, + "step": 384 + }, + { + "epoch": 0.143496086470369, + "grad_norm": 21.895673751831055, + "learning_rate": 9.863652869595151e-06, + "loss": 12.392358779907227, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4824565351009369, + "loss_total_v6": 0.6924175024032593, + "loss_wm_recon": 0.23828125, + "loss_wm_recon_w": 0.0595703125, + "step": 385 + }, + { + "epoch": 0.14386880357808424, + "grad_norm": 20.339099884033203, + "learning_rate": 9.862181690123395e-06, + "loss": 12.011767387390137, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5328884124755859, + "loss_total_v6": 0.7574977874755859, + "loss_wm_recon": 0.25, + "loss_wm_recon_w": 0.0625, + "step": 386 + }, + { + "epoch": 0.14424152068579948, + "grad_norm": 22.071687698364258, + "learning_rate": 9.860702727106794e-06, + "loss": 12.484781265258789, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.47660204768180847, + "loss_total_v6": 0.6912016868591309, + "loss_wm_recon": 0.2412109375, + "loss_wm_recon_w": 0.060302734375, + "step": 387 + }, + { + "epoch": 0.14461423779351473, + "grad_norm": 21.027408599853516, + "learning_rate": 9.859215982912927e-06, + "loss": 12.180118560791016, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.49645519256591797, + "loss_total_v6": 0.552607536315918, + "loss_wm_recon": 0.224609375, + "loss_wm_recon_w": 0.05615234375, + "step": 388 + }, + { + "epoch": 0.14498695490122995, + "grad_norm": 20.59291648864746, + "learning_rate": 9.857721459921815e-06, + "loss": 11.605972290039062, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6010158061981201, + "loss_total_v6": 0.8131740093231201, + "loss_wm_recon": 0.2314453125, + "loss_wm_recon_w": 0.057861328125, + "step": 389 + }, + { + "epoch": 0.1453596720089452, + "grad_norm": 22.275493621826172, + "learning_rate": 9.856219160525951e-06, + "loss": 11.66985034942627, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3403271734714508, + "loss_total_v6": 0.5595654249191284, + "loss_wm_recon": 0.244140625, + "loss_wm_recon_w": 0.06103515625, + "step": 390 + }, + { + "epoch": 0.14573238911666045, + "grad_norm": 20.558101654052734, + "learning_rate": 9.854709087130261e-06, + "loss": 12.083114624023438, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.3768450915813446, + "loss_total_v6": 0.598524808883667, + "loss_wm_recon": 0.24609375, + "loss_wm_recon_w": 0.0615234375, + "step": 391 + }, + { + "epoch": 0.1461051062243757, + "grad_norm": 22.120309829711914, + "learning_rate": 9.853191242152123e-06, + "loss": 11.85753059387207, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5663501024246216, + "loss_total_v6": 0.7770434617996216, + "loss_wm_recon": 0.2412109375, + "loss_wm_recon_w": 0.060302734375, + "step": 392 + }, + { + "epoch": 0.14647782333209095, + "grad_norm": 20.814334869384766, + "learning_rate": 9.851665628021354e-06, + "loss": 12.105802536010742, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4885300099849701, + "loss_total_v6": 0.6992233991622925, + "loss_wm_recon": 0.2412109375, + "loss_wm_recon_w": 0.060302734375, + "step": 393 + }, + { + "epoch": 0.1468505404398062, + "grad_norm": 21.33030891418457, + "learning_rate": 9.85013224718021e-06, + "loss": 12.075030326843262, + "loss_alignment": 0.38671875, + "loss_alignment_w": 0.193359375, + "loss_sft": 0.5515455603599548, + "loss_total_v6": 0.8186354041099548, + "loss_wm_recon": 0.294921875, + "loss_wm_recon_w": 0.07373046875, + "step": 394 + }, + { + "epoch": 0.14722325754752144, + "grad_norm": 20.90728187561035, + "learning_rate": 9.848591102083375e-06, + "loss": 11.681729316711426, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5401529669761658, + "loss_total_v6": 0.7498697638511658, + "loss_wm_recon": 0.2216796875, + "loss_wm_recon_w": 0.055419921875, + "step": 395 + }, + { + "epoch": 0.1475959746552367, + "grad_norm": 21.885393142700195, + "learning_rate": 9.847042195197971e-06, + "loss": 11.850236892700195, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.44853639602661133, + "loss_total_v6": 0.5100598335266113, + "loss_wm_recon": 0.24609375, + "loss_wm_recon_w": 0.0615234375, + "step": 396 + }, + { + "epoch": 0.1479686917629519, + "grad_norm": 20.40799331665039, + "learning_rate": 9.84548552900354e-06, + "loss": 11.893010139465332, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7683262228965759, + "loss_total_v6": 0.9860996603965759, + "loss_wm_recon": 0.23046875, + "loss_wm_recon_w": 0.0576171875, + "step": 397 + }, + { + "epoch": 0.14834140887066716, + "grad_norm": 19.268957138061523, + "learning_rate": 9.843921105992043e-06, + "loss": 12.240680694580078, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.42361995577812195, + "loss_total_v6": 0.6294305324554443, + "loss_wm_recon": 0.2294921875, + "loss_wm_recon_w": 0.057373046875, + "step": 398 + }, + { + "epoch": 0.1487141259783824, + "grad_norm": 18.553058624267578, + "learning_rate": 9.842348928667862e-06, + "loss": 12.198005676269531, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.2577238380908966, + "loss_total_v6": 0.4828214943408966, + "loss_wm_recon": 0.251953125, + "loss_wm_recon_w": 0.06298828125, + "step": 399 + }, + { + "epoch": 0.14908684308609765, + "grad_norm": 19.825523376464844, + "learning_rate": 9.840768999547794e-06, + "loss": 11.941835403442383, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.8187446594238281, + "loss_total_v6": 1.0531196594238281, + "loss_wm_recon": 0.2734375, + "loss_wm_recon_w": 0.068359375, + "step": 400 + }, + { + "epoch": 0.14908684308609765, + "eval_loss": 0.7057892680168152, + "eval_loss_alignment": 0.29037919003739315, + "eval_loss_alignment_w": 0.14518959501869658, + "eval_loss_sft": 0.5027955314781294, + "eval_loss_total_v6": 0.7057893326521939, + "eval_loss_wm_recon": 0.2312168219150641, + "eval_loss_wm_recon_w": 0.057804205478766024, + "eval_runtime": 227.4072, + "eval_samples_per_second": 8.206, + "eval_steps_per_second": 1.029, + "step": 400 + }, + { + "epoch": 0.1494595601938129, + "grad_norm": 19.990432739257812, + "learning_rate": 9.83918132116104e-06, + "loss": 11.883703231811523, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.7122136354446411, + "loss_total_v6": 0.9514714479446411, + "loss_wm_recon": 0.26953125, + "loss_wm_recon_w": 0.0673828125, + "step": 401 + }, + { + "epoch": 0.14983227730152815, + "grad_norm": 20.31435775756836, + "learning_rate": 9.83758589604921e-06, + "loss": 11.828006744384766, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.08524571359157562, + "loss_total_v6": 0.2837320566177368, + "loss_wm_recon": 0.2236328125, + "loss_wm_recon_w": 0.055908203125, + "step": 402 + }, + { + "epoch": 0.1502049944092434, + "grad_norm": 20.537567138671875, + "learning_rate": 9.835982726766318e-06, + "loss": 12.22003173828125, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.9936155080795288, + "loss_total_v6": 1.2135862112045288, + "loss_wm_recon": 0.2392578125, + "loss_wm_recon_w": 0.059814453125, + "step": 403 + }, + { + "epoch": 0.15057771151695862, + "grad_norm": 20.356863021850586, + "learning_rate": 9.834371815878766e-06, + "loss": 11.891596794128418, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.13031648099422455, + "loss_total_v6": 0.34491610527038574, + "loss_wm_recon": 0.2333984375, + "loss_wm_recon_w": 0.058349609375, + "step": 404 + }, + { + "epoch": 0.15095042862467387, + "grad_norm": 19.426856994628906, + "learning_rate": 9.832753165965359e-06, + "loss": 11.493783950805664, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.5601513981819153, + "loss_total_v6": 0.7891553044319153, + "loss_wm_recon": 0.251953125, + "loss_wm_recon_w": 0.06298828125, + "step": 405 + }, + { + "epoch": 0.1513231457323891, + "grad_norm": 20.67858123779297, + "learning_rate": 9.831126779617284e-06, + "loss": 12.181806564331055, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5200468301773071, + "loss_total_v6": 0.7370878458023071, + "loss_wm_recon": 0.2431640625, + "loss_wm_recon_w": 0.060791015625, + "step": 406 + }, + { + "epoch": 0.15169586284010436, + "grad_norm": 19.888959884643555, + "learning_rate": 9.829492659438115e-06, + "loss": 12.29250431060791, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3220692574977875, + "loss_total_v6": 0.5264149904251099, + "loss_wm_recon": 0.2236328125, + "loss_wm_recon_w": 0.055908203125, + "step": 407 + }, + { + "epoch": 0.1520685799478196, + "grad_norm": 19.768455505371094, + "learning_rate": 9.82785080804381e-06, + "loss": 11.811956405639648, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6695406436920166, + "loss_total_v6": 0.9004976749420166, + "loss_wm_recon": 0.251953125, + "loss_wm_recon_w": 0.06298828125, + "step": 408 + }, + { + "epoch": 0.15244129705553486, + "grad_norm": 20.680557250976562, + "learning_rate": 9.826201228062696e-06, + "loss": 12.12572956085205, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6160823106765747, + "loss_total_v6": 0.8218928575515747, + "loss_wm_recon": 0.2216796875, + "loss_wm_recon_w": 0.055419921875, + "step": 409 + }, + { + "epoch": 0.1528140141632501, + "grad_norm": 19.533348083496094, + "learning_rate": 9.824543922135479e-06, + "loss": 12.05363655090332, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6354524493217468, + "loss_total_v6": 0.8620149493217468, + "loss_wm_recon": 0.2890625, + "loss_wm_recon_w": 0.072265625, + "step": 410 + }, + { + "epoch": 0.15318673127096533, + "grad_norm": 19.88416862487793, + "learning_rate": 9.822878892915231e-06, + "loss": 11.837359428405762, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.8506869673728943, + "loss_total_v6": 1.063821792602539, + "loss_wm_recon": 0.2275390625, + "loss_wm_recon_w": 0.056884765625, + "step": 411 + }, + { + "epoch": 0.15355944837868057, + "grad_norm": 19.785457611083984, + "learning_rate": 9.821206143067385e-06, + "loss": 11.967639923095703, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.4434342682361603, + "loss_total_v6": 0.6677994728088379, + "loss_wm_recon": 0.2333984375, + "loss_wm_recon_w": 0.058349609375, + "step": 412 + }, + { + "epoch": 0.15393216548639582, + "grad_norm": 19.687837600708008, + "learning_rate": 9.819525675269736e-06, + "loss": 12.29753303527832, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6100414991378784, + "loss_total_v6": 0.8204907178878784, + "loss_wm_recon": 0.224609375, + "loss_wm_recon_w": 0.05615234375, + "step": 413 + }, + { + "epoch": 0.15430488259411107, + "grad_norm": 19.333763122558594, + "learning_rate": 9.817837492212434e-06, + "loss": 11.89186954498291, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.536536455154419, + "loss_total_v6": 0.745032548904419, + "loss_wm_recon": 0.224609375, + "loss_wm_recon_w": 0.05615234375, + "step": 414 + }, + { + "epoch": 0.15467759970182632, + "grad_norm": 20.187280654907227, + "learning_rate": 9.816141596597981e-06, + "loss": 12.000531196594238, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7202457189559937, + "loss_total_v6": 0.9404605627059937, + "loss_wm_recon": 0.240234375, + "loss_wm_recon_w": 0.06005859375, + "step": 415 + }, + { + "epoch": 0.15505031680954157, + "grad_norm": 18.682968139648438, + "learning_rate": 9.81443799114122e-06, + "loss": 11.817131996154785, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6207065582275391, + "loss_total_v6": 0.8235874176025391, + "loss_wm_recon": 0.2177734375, + "loss_wm_recon_w": 0.054443359375, + "step": 416 + }, + { + "epoch": 0.15542303391725681, + "grad_norm": 18.995237350463867, + "learning_rate": 9.812726678569347e-06, + "loss": 11.743927001953125, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.44138389825820923, + "loss_total_v6": 0.6381612420082092, + "loss_wm_recon": 0.216796875, + "loss_wm_recon_w": 0.05419921875, + "step": 417 + }, + { + "epoch": 0.15579575102497203, + "grad_norm": 18.99240493774414, + "learning_rate": 9.811007661621881e-06, + "loss": 11.689666748046875, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.42573270201683044, + "loss_total_v6": 0.6364260911941528, + "loss_wm_recon": 0.2255859375, + "loss_wm_recon_w": 0.056396484375, + "step": 418 + }, + { + "epoch": 0.15616846813268728, + "grad_norm": 19.697479248046875, + "learning_rate": 9.809280943050688e-06, + "loss": 12.034270286560059, + "loss_alignment": 0.34375, + "loss_alignment_w": 0.171875, + "loss_sft": 0.6407233476638794, + "loss_total_v6": 0.8775397539138794, + "loss_wm_recon": 0.259765625, + "loss_wm_recon_w": 0.06494140625, + "step": 419 + }, + { + "epoch": 0.15654118524040253, + "grad_norm": 19.675060272216797, + "learning_rate": 9.807546525619956e-06, + "loss": 11.73412036895752, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5380399227142334, + "loss_total_v6": 0.7611844539642334, + "loss_wm_recon": 0.251953125, + "loss_wm_recon_w": 0.06298828125, + "step": 420 + }, + { + "epoch": 0.15691390234811778, + "grad_norm": 34.0985107421875, + "learning_rate": 9.805804412106197e-06, + "loss": 11.97987174987793, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5843976736068726, + "loss_total_v6": 0.7919172048568726, + "loss_wm_recon": 0.228515625, + "loss_wm_recon_w": 0.05712890625, + "step": 421 + }, + { + "epoch": 0.15728661945583303, + "grad_norm": 19.993911743164062, + "learning_rate": 9.80405460529825e-06, + "loss": 11.709811210632324, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.589796245098114, + "loss_total_v6": 0.805860698223114, + "loss_wm_recon": 0.2392578125, + "loss_wm_recon_w": 0.059814453125, + "step": 422 + }, + { + "epoch": 0.15765933656354827, + "grad_norm": 18.787986755371094, + "learning_rate": 9.802297107997261e-06, + "loss": 11.800540924072266, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5300827622413635, + "loss_total_v6": 0.7468796372413635, + "loss_wm_recon": 0.234375, + "loss_wm_recon_w": 0.05859375, + "step": 423 + }, + { + "epoch": 0.15803205367126352, + "grad_norm": 21.723949432373047, + "learning_rate": 9.80053192301669e-06, + "loss": 12.19625473022461, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5883498191833496, + "loss_total_v6": 0.7956252098083496, + "loss_wm_recon": 0.2197265625, + "loss_wm_recon_w": 0.054931640625, + "step": 424 + }, + { + "epoch": 0.15840477077897874, + "grad_norm": 19.46007537841797, + "learning_rate": 9.79875905318231e-06, + "loss": 11.63441276550293, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6041014194488525, + "loss_total_v6": 0.8128416538238525, + "loss_wm_recon": 0.2333984375, + "loss_wm_recon_w": 0.058349609375, + "step": 425 + }, + { + "epoch": 0.158777487886694, + "grad_norm": 18.777938842773438, + "learning_rate": 9.796978501332185e-06, + "loss": 11.477660179138184, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4344562292098999, + "loss_total_v6": 0.6449054479598999, + "loss_wm_recon": 0.232421875, + "loss_wm_recon_w": 0.05810546875, + "step": 426 + }, + { + "epoch": 0.15915020499440924, + "grad_norm": 20.735225677490234, + "learning_rate": 9.795190270316687e-06, + "loss": 12.097992897033691, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4050775468349457, + "loss_total_v6": 0.6086908578872681, + "loss_wm_recon": 0.220703125, + "loss_wm_recon_w": 0.05517578125, + "step": 427 + }, + { + "epoch": 0.1595229221021245, + "grad_norm": 20.08676528930664, + "learning_rate": 9.793394362998475e-06, + "loss": 11.42597770690918, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.39431437849998474, + "loss_total_v6": 0.6252714395523071, + "loss_wm_recon": 0.259765625, + "loss_wm_recon_w": 0.06494140625, + "step": 428 + }, + { + "epoch": 0.15989563920983973, + "grad_norm": 18.861248016357422, + "learning_rate": 9.791590782252498e-06, + "loss": 11.94564437866211, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.4168059527873993, + "loss_total_v6": 0.6296966075897217, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 429 + }, + { + "epoch": 0.16026835631755498, + "grad_norm": 20.675350189208984, + "learning_rate": 9.789779530965991e-06, + "loss": 12.024707794189453, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.39551445841789246, + "loss_total_v6": 0.6076726913452148, + "loss_wm_recon": 0.2314453125, + "loss_wm_recon_w": 0.057861328125, + "step": 430 + }, + { + "epoch": 0.16064107342527023, + "grad_norm": 20.182022094726562, + "learning_rate": 9.787960612038463e-06, + "loss": 12.264522552490234, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6720423102378845, + "loss_total_v6": 0.8873743414878845, + "loss_wm_recon": 0.228515625, + "loss_wm_recon_w": 0.05712890625, + "step": 431 + }, + { + "epoch": 0.16101379053298545, + "grad_norm": 19.25800323486328, + "learning_rate": 9.786134028381701e-06, + "loss": 11.771211624145508, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.6942839026451111, + "loss_total_v6": 0.8888639807701111, + "loss_wm_recon": 0.2080078125, + "loss_wm_recon_w": 0.052001953125, + "step": 432 + }, + { + "epoch": 0.1613865076407007, + "grad_norm": 19.336082458496094, + "learning_rate": 9.784299782919766e-06, + "loss": 11.776906967163086, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5133835673332214, + "loss_total_v6": 0.7231003642082214, + "loss_wm_recon": 0.2373046875, + "loss_wm_recon_w": 0.059326171875, + "step": 433 + }, + { + "epoch": 0.16175922474841595, + "grad_norm": 19.61280632019043, + "learning_rate": 9.782457878588977e-06, + "loss": 11.953789710998535, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.20807379484176636, + "loss_total_v6": 0.41241949796676636, + "loss_wm_recon": 0.2236328125, + "loss_wm_recon_w": 0.055908203125, + "step": 434 + }, + { + "epoch": 0.1621319418561312, + "grad_norm": 19.086145401000977, + "learning_rate": 9.780608318337917e-06, + "loss": 11.527427673339844, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5291027426719666, + "loss_total_v6": 0.7336925864219666, + "loss_wm_recon": 0.216796875, + "loss_wm_recon_w": 0.05419921875, + "step": 435 + }, + { + "epoch": 0.16250465896384644, + "grad_norm": 19.602062225341797, + "learning_rate": 9.778751105127425e-06, + "loss": 11.658164024353027, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.2791937589645386, + "loss_total_v6": 0.4998968839645386, + "loss_wm_recon": 0.2265625, + "loss_wm_recon_w": 0.056640625, + "step": 436 + }, + { + "epoch": 0.1628773760715617, + "grad_norm": 20.21976089477539, + "learning_rate": 9.77688624193059e-06, + "loss": 11.80183219909668, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5364471077919006, + "loss_total_v6": 0.7522674202919006, + "loss_wm_recon": 0.23046875, + "loss_wm_recon_w": 0.0576171875, + "step": 437 + }, + { + "epoch": 0.16325009317927694, + "grad_norm": 19.480440139770508, + "learning_rate": 9.77501373173275e-06, + "loss": 11.679176330566406, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3843803107738495, + "loss_total_v6": 0.5848197937011719, + "loss_wm_recon": 0.2080078125, + "loss_wm_recon_w": 0.052001953125, + "step": 438 + }, + { + "epoch": 0.16362281028699216, + "grad_norm": 19.689085006713867, + "learning_rate": 9.77313357753148e-06, + "loss": 12.04626178741455, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5409274697303772, + "loss_total_v6": 0.7389255166053772, + "loss_wm_recon": 0.2060546875, + "loss_wm_recon_w": 0.051513671875, + "step": 439 + }, + { + "epoch": 0.1639955273947074, + "grad_norm": 19.042348861694336, + "learning_rate": 9.771245782336596e-06, + "loss": 11.675708770751953, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5866205096244812, + "loss_total_v6": 0.6422845721244812, + "loss_wm_recon": 0.22265625, + "loss_wm_recon_w": 0.0556640625, + "step": 440 + }, + { + "epoch": 0.16436824450242266, + "grad_norm": 20.014902114868164, + "learning_rate": 9.769350349170145e-06, + "loss": 11.995512962341309, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6951375007629395, + "loss_total_v6": 0.8994832038879395, + "loss_wm_recon": 0.2158203125, + "loss_wm_recon_w": 0.053955078125, + "step": 441 + }, + { + "epoch": 0.1647409616101379, + "grad_norm": 20.162372589111328, + "learning_rate": 9.767447281066399e-06, + "loss": 12.009989738464355, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6507883071899414, + "loss_total_v6": 0.8673410415649414, + "loss_wm_recon": 0.2333984375, + "loss_wm_recon_w": 0.058349609375, + "step": 442 + }, + { + "epoch": 0.16511367871785315, + "grad_norm": 20.407678604125977, + "learning_rate": 9.765536581071854e-06, + "loss": 12.087608337402344, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.40275004506111145, + "loss_total_v6": 0.6156406402587891, + "loss_wm_recon": 0.234375, + "loss_wm_recon_w": 0.05859375, + "step": 443 + }, + { + "epoch": 0.1654863958255684, + "grad_norm": 19.127347946166992, + "learning_rate": 9.763618252245224e-06, + "loss": 11.459896087646484, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5605115294456482, + "loss_total_v6": 0.7760877013206482, + "loss_wm_recon": 0.2294921875, + "loss_wm_recon_w": 0.057373046875, + "step": 444 + }, + { + "epoch": 0.16585911293328365, + "grad_norm": 19.12306022644043, + "learning_rate": 9.761692297657432e-06, + "loss": 11.669916152954102, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3972524106502533, + "loss_total_v6": 0.6064809560775757, + "loss_wm_recon": 0.2275390625, + "loss_wm_recon_w": 0.056884765625, + "step": 445 + }, + { + "epoch": 0.16623183004099887, + "grad_norm": 19.69778823852539, + "learning_rate": 9.759758720391616e-06, + "loss": 11.864311218261719, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.25032827258110046, + "loss_total_v6": 0.46858999133110046, + "loss_wm_recon": 0.240234375, + "loss_wm_recon_w": 0.06005859375, + "step": 446 + }, + { + "epoch": 0.16660454714871412, + "grad_norm": 19.533655166625977, + "learning_rate": 9.75781752354311e-06, + "loss": 11.80079460144043, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.7100116014480591, + "loss_total_v6": 0.9045916795730591, + "loss_wm_recon": 0.2158203125, + "loss_wm_recon_w": 0.053955078125, + "step": 447 + }, + { + "epoch": 0.16697726425642936, + "grad_norm": 19.510173797607422, + "learning_rate": 9.755868710219444e-06, + "loss": 11.764150619506836, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.48368576169013977, + "loss_total_v6": 0.6826603412628174, + "loss_wm_recon": 0.2099609375, + "loss_wm_recon_w": 0.052490234375, + "step": 448 + }, + { + "epoch": 0.1673499813641446, + "grad_norm": 19.69495391845703, + "learning_rate": 9.753912283540348e-06, + "loss": 11.311052322387695, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5980384945869446, + "loss_total_v6": 0.8206947445869446, + "loss_wm_recon": 0.2421875, + "loss_wm_recon_w": 0.060546875, + "step": 449 + }, + { + "epoch": 0.16772269847185986, + "grad_norm": 19.85298728942871, + "learning_rate": 9.751948246637734e-06, + "loss": 11.953128814697266, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.20614826679229736, + "loss_total_v6": 0.41391193866729736, + "loss_wm_recon": 0.2216796875, + "loss_wm_recon_w": 0.055419921875, + "step": 450 + }, + { + "epoch": 0.1680954155795751, + "grad_norm": 20.08881187438965, + "learning_rate": 9.749976602655701e-06, + "loss": 11.914461135864258, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7412532567977905, + "loss_total_v6": 0.7949641942977905, + "loss_wm_recon": 0.21484375, + "loss_wm_recon_w": 0.0537109375, + "step": 451 + }, + { + "epoch": 0.16846813268729036, + "grad_norm": 19.188663482666016, + "learning_rate": 9.747997354750523e-06, + "loss": 11.342674255371094, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.612049400806427, + "loss_total_v6": 0.835438072681427, + "loss_wm_recon": 0.2294921875, + "loss_wm_recon_w": 0.057373046875, + "step": 452 + }, + { + "epoch": 0.1688408497950056, + "grad_norm": 18.868650436401367, + "learning_rate": 9.746010506090645e-06, + "loss": 11.654888153076172, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7667372822761536, + "loss_total_v6": 0.8236220479011536, + "loss_wm_recon": 0.2275390625, + "loss_wm_recon_w": 0.056884765625, + "step": 453 + }, + { + "epoch": 0.16921356690272082, + "grad_norm": 19.534236907958984, + "learning_rate": 9.744016059856682e-06, + "loss": 11.873512268066406, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.2711341381072998, + "loss_total_v6": 0.4994056224822998, + "loss_wm_recon": 0.2490234375, + "loss_wm_recon_w": 0.062255859375, + "step": 454 + }, + { + "epoch": 0.16958628401043607, + "grad_norm": 18.71087646484375, + "learning_rate": 9.742014019241413e-06, + "loss": 11.857776641845703, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.49757513403892517, + "loss_total_v6": 0.7104657888412476, + "loss_wm_recon": 0.2265625, + "loss_wm_recon_w": 0.056640625, + "step": 455 + }, + { + "epoch": 0.16995900111815132, + "grad_norm": 19.04155921936035, + "learning_rate": 9.74000438744977e-06, + "loss": 11.261411666870117, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4599761366844177, + "loss_total_v6": 0.5254058241844177, + "loss_wm_recon": 0.26171875, + "loss_wm_recon_w": 0.0654296875, + "step": 456 + }, + { + "epoch": 0.17033171822586657, + "grad_norm": 23.358675003051758, + "learning_rate": 9.737987167698839e-06, + "loss": 12.227781295776367, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6422681212425232, + "loss_total_v6": 0.8615064024925232, + "loss_wm_recon": 0.236328125, + "loss_wm_recon_w": 0.05908203125, + "step": 457 + }, + { + "epoch": 0.17070443533358182, + "grad_norm": 19.47095489501953, + "learning_rate": 9.735962363217854e-06, + "loss": 12.075713157653809, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.15966345369815826, + "loss_total_v6": 0.36986851692199707, + "loss_wm_recon": 0.2236328125, + "loss_wm_recon_w": 0.055908203125, + "step": 458 + }, + { + "epoch": 0.17107715244129706, + "grad_norm": 18.844947814941406, + "learning_rate": 9.733929977248194e-06, + "loss": 12.21061897277832, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.34226658940315247, + "loss_total_v6": 0.5546689033508301, + "loss_wm_recon": 0.232421875, + "loss_wm_recon_w": 0.05810546875, + "step": 459 + }, + { + "epoch": 0.1714498695490123, + "grad_norm": 18.562013626098633, + "learning_rate": 9.731890013043367e-06, + "loss": 12.229217529296875, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3431204557418823, + "loss_total_v6": 0.5574759244918823, + "loss_wm_recon": 0.224609375, + "loss_wm_recon_w": 0.05615234375, + "step": 460 + }, + { + "epoch": 0.17182258665672753, + "grad_norm": 19.487934112548828, + "learning_rate": 9.72984247386902e-06, + "loss": 11.637758255004883, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6683838963508606, + "loss_total_v6": 0.8710206151008606, + "loss_wm_recon": 0.208984375, + "loss_wm_recon_w": 0.05224609375, + "step": 461 + }, + { + "epoch": 0.17219530376444278, + "grad_norm": 19.305227279663086, + "learning_rate": 9.72778736300292e-06, + "loss": 11.506416320800781, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5688260793685913, + "loss_total_v6": 0.7856229543685913, + "loss_wm_recon": 0.234375, + "loss_wm_recon_w": 0.05859375, + "step": 462 + }, + { + "epoch": 0.17256802087215803, + "grad_norm": 19.42057991027832, + "learning_rate": 9.72572468373496e-06, + "loss": 11.831808090209961, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4664902687072754, + "loss_total_v6": 0.5241074562072754, + "loss_wm_recon": 0.23046875, + "loss_wm_recon_w": 0.0576171875, + "step": 463 + }, + { + "epoch": 0.17294073797987328, + "grad_norm": 18.985576629638672, + "learning_rate": 9.723654439367146e-06, + "loss": 11.802421569824219, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.9965344667434692, + "loss_total_v6": 1.2475110292434692, + "loss_wm_recon": 0.27734375, + "loss_wm_recon_w": 0.0693359375, + "step": 464 + }, + { + "epoch": 0.17331345508758852, + "grad_norm": 19.898210525512695, + "learning_rate": 9.721576633213593e-06, + "loss": 11.431806564331055, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.517459511756897, + "loss_total_v6": 0.726443886756897, + "loss_wm_recon": 0.203125, + "loss_wm_recon_w": 0.05078125, + "step": 465 + }, + { + "epoch": 0.17368617219530377, + "grad_norm": 19.686241149902344, + "learning_rate": 9.719491268600527e-06, + "loss": 11.485490798950195, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.460569828748703, + "loss_total_v6": 0.6600327491760254, + "loss_wm_recon": 0.2119140625, + "loss_wm_recon_w": 0.052978515625, + "step": 466 + }, + { + "epoch": 0.17405888930301902, + "grad_norm": 19.624380111694336, + "learning_rate": 9.717398348866266e-06, + "loss": 11.762588500976562, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5211321711540222, + "loss_total_v6": 0.7313372492790222, + "loss_wm_recon": 0.2158203125, + "loss_wm_recon_w": 0.053955078125, + "step": 467 + }, + { + "epoch": 0.17443160641073424, + "grad_norm": 19.654882431030273, + "learning_rate": 9.715297877361231e-06, + "loss": 11.560450553894043, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6747027039527893, + "loss_total_v6": 0.8817339539527893, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 468 + }, + { + "epoch": 0.1748043235184495, + "grad_norm": 19.72140121459961, + "learning_rate": 9.713189857447926e-06, + "loss": 11.759737014770508, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.7908428311347961, + "loss_total_v6": 0.9964092373847961, + "loss_wm_recon": 0.212890625, + "loss_wm_recon_w": 0.05322265625, + "step": 469 + }, + { + "epoch": 0.17517704062616474, + "grad_norm": 18.931882858276367, + "learning_rate": 9.71107429250094e-06, + "loss": 11.724149703979492, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.2859348654747009, + "loss_total_v6": 0.4968723654747009, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 470 + }, + { + "epoch": 0.17554975773387999, + "grad_norm": 19.59602165222168, + "learning_rate": 9.708951185906943e-06, + "loss": 12.06873893737793, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.667386531829834, + "loss_total_v6": 0.877835750579834, + "loss_wm_recon": 0.224609375, + "loss_wm_recon_w": 0.05615234375, + "step": 471 + }, + { + "epoch": 0.17592247484159523, + "grad_norm": 19.57518196105957, + "learning_rate": 9.706820541064675e-06, + "loss": 11.70506477355957, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.502555251121521, + "loss_total_v6": 0.732535719871521, + "loss_wm_recon": 0.240234375, + "loss_wm_recon_w": 0.06005859375, + "step": 472 + }, + { + "epoch": 0.17629519194931048, + "grad_norm": 19.461700439453125, + "learning_rate": 9.704682361384941e-06, + "loss": 11.433874130249023, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.23614370822906494, + "loss_total_v6": 0.44219839572906494, + "loss_wm_recon": 0.21484375, + "loss_wm_recon_w": 0.0537109375, + "step": 473 + }, + { + "epoch": 0.17666790905702573, + "grad_norm": 21.832277297973633, + "learning_rate": 9.702536650290615e-06, + "loss": 11.933969497680664, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.565200924873352, + "loss_total_v6": 0.780288815498352, + "loss_wm_recon": 0.2197265625, + "loss_wm_recon_w": 0.054931640625, + "step": 474 + }, + { + "epoch": 0.17704062616474095, + "grad_norm": 19.76290512084961, + "learning_rate": 9.700383411216623e-06, + "loss": 11.837130546569824, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6710668206214905, + "loss_total_v6": 0.8912816643714905, + "loss_wm_recon": 0.224609375, + "loss_wm_recon_w": 0.05615234375, + "step": 475 + }, + { + "epoch": 0.1774133432724562, + "grad_norm": 18.977962493896484, + "learning_rate": 9.698222647609941e-06, + "loss": 11.147414207458496, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5664660334587097, + "loss_total_v6": 0.7859484553337097, + "loss_wm_recon": 0.2294921875, + "loss_wm_recon_w": 0.057373046875, + "step": 476 + }, + { + "epoch": 0.17778606038017145, + "grad_norm": 20.28502082824707, + "learning_rate": 9.696054362929597e-06, + "loss": 11.691041946411133, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.7464438080787659, + "loss_total_v6": 0.9722738862037659, + "loss_wm_recon": 0.2314453125, + "loss_wm_recon_w": 0.057861328125, + "step": 477 + }, + { + "epoch": 0.1781587774878867, + "grad_norm": 19.81757164001465, + "learning_rate": 9.69387856064665e-06, + "loss": 11.772737503051758, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.44192957878112793, + "loss_total_v6": 0.6504256725311279, + "loss_wm_recon": 0.216796875, + "loss_wm_recon_w": 0.05419921875, + "step": 478 + }, + { + "epoch": 0.17853149459560194, + "grad_norm": 19.687891006469727, + "learning_rate": 9.691695244244199e-06, + "loss": 11.436951637268066, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5762566328048706, + "loss_total_v6": 0.8025749921798706, + "loss_wm_recon": 0.2333984375, + "loss_wm_recon_w": 0.058349609375, + "step": 479 + }, + { + "epoch": 0.1789042117033172, + "grad_norm": 19.653772354125977, + "learning_rate": 9.689504417217371e-06, + "loss": 11.655304908752441, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6150402426719666, + "loss_total_v6": 0.8306164145469666, + "loss_wm_recon": 0.2294921875, + "loss_wm_recon_w": 0.057373046875, + "step": 480 + }, + { + "epoch": 0.17927692881103244, + "grad_norm": 19.211061477661133, + "learning_rate": 9.687306083073316e-06, + "loss": 11.103906631469727, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.781032383441925, + "loss_total_v6": 0.996852695941925, + "loss_wm_recon": 0.22265625, + "loss_wm_recon_w": 0.0556640625, + "step": 481 + }, + { + "epoch": 0.17964964591874766, + "grad_norm": 18.659908294677734, + "learning_rate": 9.685100245331202e-06, + "loss": 11.584522247314453, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6056365370750427, + "loss_total_v6": 0.8058318495750427, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 482 + }, + { + "epoch": 0.1800223630264629, + "grad_norm": 19.19989776611328, + "learning_rate": 9.68288690752221e-06, + "loss": 11.424530029296875, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5182608962059021, + "loss_total_v6": 0.7272452712059021, + "loss_wm_recon": 0.2109375, + "loss_wm_recon_w": 0.052734375, + "step": 483 + }, + { + "epoch": 0.18039508013417815, + "grad_norm": 20.592382431030273, + "learning_rate": 9.680666073189522e-06, + "loss": 11.541372299194336, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6588438749313354, + "loss_total_v6": 0.8739317655563354, + "loss_wm_recon": 0.2275390625, + "loss_wm_recon_w": 0.056884765625, + "step": 484 + }, + { + "epoch": 0.1807677972418934, + "grad_norm": 19.025381088256836, + "learning_rate": 9.678437745888328e-06, + "loss": 11.910209655761719, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6066125631332397, + "loss_total_v6": 0.8155969381332397, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 485 + }, + { + "epoch": 0.18114051434960865, + "grad_norm": 19.407564163208008, + "learning_rate": 9.676201929185809e-06, + "loss": 11.652524948120117, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5703195929527283, + "loss_total_v6": 0.7822336554527283, + "loss_wm_recon": 0.22265625, + "loss_wm_recon_w": 0.0556640625, + "step": 486 + }, + { + "epoch": 0.1815132314573239, + "grad_norm": 18.6895751953125, + "learning_rate": 9.673958626661135e-06, + "loss": 11.247902870178223, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.40166395902633667, + "loss_total_v6": 0.5999061465263367, + "loss_wm_recon": 0.20703125, + "loss_wm_recon_w": 0.0517578125, + "step": 487 + }, + { + "epoch": 0.18188594856503915, + "grad_norm": 18.539997100830078, + "learning_rate": 9.671707841905462e-06, + "loss": 11.512791633605957, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.14422239363193512, + "loss_total_v6": 0.3405114412307739, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 488 + }, + { + "epoch": 0.18225866567275437, + "grad_norm": 18.81951904296875, + "learning_rate": 9.66944957852192e-06, + "loss": 11.718061447143555, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.45370447635650635, + "loss_total_v6": 0.6665951013565063, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 489 + }, + { + "epoch": 0.18263138278046961, + "grad_norm": 19.05141830444336, + "learning_rate": 9.667183840125619e-06, + "loss": 11.81680679321289, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.505286693572998, + "loss_total_v6": 0.707435131072998, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 490 + }, + { + "epoch": 0.18300409988818486, + "grad_norm": 18.36578369140625, + "learning_rate": 9.664910630343622e-06, + "loss": 11.765203475952148, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6033306121826172, + "loss_total_v6": 0.6575298309326172, + "loss_wm_recon": 0.216796875, + "loss_wm_recon_w": 0.05419921875, + "step": 491 + }, + { + "epoch": 0.1833768169959001, + "grad_norm": 18.33992576599121, + "learning_rate": 9.662629952814964e-06, + "loss": 11.77823257446289, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7290310859680176, + "loss_total_v6": 0.7837185859680176, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 492 + }, + { + "epoch": 0.18374953410361536, + "grad_norm": 19.54945182800293, + "learning_rate": 9.66034181119063e-06, + "loss": 12.020421981811523, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6090166568756104, + "loss_total_v6": 0.8258135318756104, + "loss_wm_recon": 0.2109375, + "loss_wm_recon_w": 0.052734375, + "step": 493 + }, + { + "epoch": 0.1841222512113306, + "grad_norm": 18.68202781677246, + "learning_rate": 9.658046209133557e-06, + "loss": 11.924671173095703, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.5980469584465027, + "loss_total_v6": 0.8243653178215027, + "loss_wm_recon": 0.2255859375, + "loss_wm_recon_w": 0.056396484375, + "step": 494 + }, + { + "epoch": 0.18449496831904585, + "grad_norm": 19.22593879699707, + "learning_rate": 9.655743150318618e-06, + "loss": 11.61439323425293, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4073656499385834, + "loss_total_v6": 0.6192797422409058, + "loss_wm_recon": 0.24609375, + "loss_wm_recon_w": 0.0615234375, + "step": 495 + }, + { + "epoch": 0.18486768542676107, + "grad_norm": 19.311481475830078, + "learning_rate": 9.653432638432631e-06, + "loss": 12.267818450927734, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.28382283449172974, + "loss_total_v6": 0.49720174074172974, + "loss_wm_recon": 0.236328125, + "loss_wm_recon_w": 0.05908203125, + "step": 496 + }, + { + "epoch": 0.18524040253447632, + "grad_norm": 18.578821182250977, + "learning_rate": 9.651114677174338e-06, + "loss": 11.682144165039062, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.8209962248802185, + "loss_total_v6": 1.0241212844848633, + "loss_wm_recon": 0.203125, + "loss_wm_recon_w": 0.05078125, + "step": 497 + }, + { + "epoch": 0.18561311964219157, + "grad_norm": 18.780717849731445, + "learning_rate": 9.648789270254415e-06, + "loss": 11.542821884155273, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.812674880027771, + "loss_total_v6": 1.018241286277771, + "loss_wm_recon": 0.197265625, + "loss_wm_recon_w": 0.04931640625, + "step": 498 + }, + { + "epoch": 0.18598583674990682, + "grad_norm": 18.43904685974121, + "learning_rate": 9.646456421395447e-06, + "loss": 11.70185661315918, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.42959633469581604, + "loss_total_v6": 0.6288150548934937, + "loss_wm_recon": 0.1953125, + "loss_wm_recon_w": 0.048828125, + "step": 499 + }, + { + "epoch": 0.18635855385762207, + "grad_norm": 20.73028564453125, + "learning_rate": 9.64411613433194e-06, + "loss": 11.968389511108398, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4602718949317932, + "loss_total_v6": 0.6724300980567932, + "loss_wm_recon": 0.2001953125, + "loss_wm_recon_w": 0.050048828125, + "step": 500 + }, + { + "epoch": 0.18635855385762207, + "eval_loss": 0.6932654976844788, + "eval_loss_alignment": 0.28892269297542733, + "eval_loss_alignment_w": 0.14446134648771367, + "eval_loss_sft": 0.4980294540620003, + "eval_loss_total_v6": 0.6932655297952075, + "eval_loss_wm_recon": 0.20309891659989315, + "eval_loss_wm_recon_w": 0.05077472914997329, + "eval_runtime": 229.2263, + "eval_samples_per_second": 8.14, + "eval_steps_per_second": 1.021, + "step": 500 + }, + { + "epoch": 0.18673127096533731, + "grad_norm": 19.017534255981445, + "learning_rate": 9.641768412810305e-06, + "loss": 11.687788963317871, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4469568133354187, + "loss_total_v6": 0.6676599383354187, + "loss_wm_recon": 0.234375, + "loss_wm_recon_w": 0.05859375, + "step": 501 + }, + { + "epoch": 0.18710398807305256, + "grad_norm": 18.664464950561523, + "learning_rate": 9.63941326058885e-06, + "loss": 11.754846572875977, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5524594187736511, + "loss_total_v6": 0.7673031687736511, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 502 + }, + { + "epoch": 0.1874767051807678, + "grad_norm": 18.255966186523438, + "learning_rate": 9.637050681437787e-06, + "loss": 11.779143333435059, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.8089299201965332, + "loss_total_v6": 1.0354924201965332, + "loss_wm_recon": 0.2421875, + "loss_wm_recon_w": 0.060546875, + "step": 503 + }, + { + "epoch": 0.18784942228848303, + "grad_norm": 19.158920288085938, + "learning_rate": 9.634680679139213e-06, + "loss": 11.581520080566406, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4863094091415405, + "loss_total_v6": 0.6860164403915405, + "loss_wm_recon": 0.197265625, + "loss_wm_recon_w": 0.04931640625, + "step": 504 + }, + { + "epoch": 0.18822213939619828, + "grad_norm": 18.165939331054688, + "learning_rate": 9.632303257487102e-06, + "loss": 11.049430847167969, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4878880977630615, + "loss_total_v6": 0.6897923946380615, + "loss_wm_recon": 0.1904296875, + "loss_wm_recon_w": 0.047607421875, + "step": 505 + }, + { + "epoch": 0.18859485650391353, + "grad_norm": 18.63922691345215, + "learning_rate": 9.629918420287316e-06, + "loss": 11.841766357421875, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.2651588022708893, + "loss_total_v6": 0.4651099741458893, + "loss_wm_recon": 0.2060546875, + "loss_wm_recon_w": 0.051513671875, + "step": 506 + }, + { + "epoch": 0.18896757361162878, + "grad_norm": 20.361156463623047, + "learning_rate": 9.627526171357582e-06, + "loss": 11.581687927246094, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.7784247398376465, + "loss_total_v6": 0.9881415367126465, + "loss_wm_recon": 0.2138671875, + "loss_wm_recon_w": 0.053466796875, + "step": 507 + }, + { + "epoch": 0.18934029071934402, + "grad_norm": 18.655349731445312, + "learning_rate": 9.625126514527492e-06, + "loss": 11.617101669311523, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4483095705509186, + "loss_total_v6": 0.6521669626235962, + "loss_wm_recon": 0.2138671875, + "loss_wm_recon_w": 0.053466796875, + "step": 508 + }, + { + "epoch": 0.18971300782705927, + "grad_norm": 19.94216537475586, + "learning_rate": 9.622719453638499e-06, + "loss": 11.622702598571777, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4357874393463135, + "loss_total_v6": 0.6386682987213135, + "loss_wm_recon": 0.2099609375, + "loss_wm_recon_w": 0.052490234375, + "step": 509 + }, + { + "epoch": 0.19008572493477452, + "grad_norm": 19.144916534423828, + "learning_rate": 9.620304992543908e-06, + "loss": 11.293861389160156, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.41795098781585693, + "loss_total_v6": 0.47459161281585693, + "loss_wm_recon": 0.2265625, + "loss_wm_recon_w": 0.056640625, + "step": 510 + }, + { + "epoch": 0.19045844204248974, + "grad_norm": 19.130369186401367, + "learning_rate": 9.617883135108867e-06, + "loss": 11.74193286895752, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6714141368865967, + "loss_total_v6": 0.8808867931365967, + "loss_wm_recon": 0.197265625, + "loss_wm_recon_w": 0.04931640625, + "step": 511 + }, + { + "epoch": 0.190831159150205, + "grad_norm": 18.97854232788086, + "learning_rate": 9.615453885210368e-06, + "loss": 11.038616180419922, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.13947737216949463, + "loss_total_v6": 0.19440901279449463, + "loss_wm_recon": 0.2197265625, + "loss_wm_recon_w": 0.054931640625, + "step": 512 + }, + { + "epoch": 0.19120387625792024, + "grad_norm": 20.286670684814453, + "learning_rate": 9.613017246737239e-06, + "loss": 11.849242210388184, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.53857421875, + "loss_total_v6": 0.751220703125, + "loss_wm_recon": 0.2099609375, + "loss_wm_recon_w": 0.052490234375, + "step": 513 + }, + { + "epoch": 0.19157659336563548, + "grad_norm": 19.814668655395508, + "learning_rate": 9.610573223590132e-06, + "loss": 11.528711318969727, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5595579147338867, + "loss_total_v6": 0.7734251022338867, + "loss_wm_recon": 0.21484375, + "loss_wm_recon_w": 0.0537109375, + "step": 514 + }, + { + "epoch": 0.19194931047335073, + "grad_norm": 20.422012329101562, + "learning_rate": 9.608121819681521e-06, + "loss": 11.545616149902344, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.7039156556129456, + "loss_total_v6": 0.9072847962379456, + "loss_wm_recon": 0.2119140625, + "loss_wm_recon_w": 0.052978515625, + "step": 515 + }, + { + "epoch": 0.19232202758106598, + "grad_norm": 18.7853946685791, + "learning_rate": 9.605663038935698e-06, + "loss": 11.333206176757812, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.20565757155418396, + "loss_total_v6": 0.41659507155418396, + "loss_wm_recon": 0.2109375, + "loss_wm_recon_w": 0.052734375, + "step": 516 + }, + { + "epoch": 0.19269474468878123, + "grad_norm": 18.718128204345703, + "learning_rate": 9.603196885288759e-06, + "loss": 11.143489837646484, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.5951959490776062, + "loss_total_v6": 0.7902643084526062, + "loss_wm_recon": 0.2099609375, + "loss_wm_recon_w": 0.052490234375, + "step": 517 + }, + { + "epoch": 0.19306746179649645, + "grad_norm": 18.591079711914062, + "learning_rate": 9.60072336268861e-06, + "loss": 11.738654136657715, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4475322663784027, + "loss_total_v6": 0.6540752649307251, + "loss_wm_recon": 0.201171875, + "loss_wm_recon_w": 0.05029296875, + "step": 518 + }, + { + "epoch": 0.1934401789042117, + "grad_norm": 19.399946212768555, + "learning_rate": 9.598242475094948e-06, + "loss": 11.794709205627441, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5698187351226807, + "loss_total_v6": 0.7702581882476807, + "loss_wm_recon": 0.2001953125, + "loss_wm_recon_w": 0.050048828125, + "step": 519 + }, + { + "epoch": 0.19381289601192694, + "grad_norm": 20.08139419555664, + "learning_rate": 9.59575422647926e-06, + "loss": 12.123802185058594, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6595576405525208, + "loss_total_v6": 0.8604853749275208, + "loss_wm_recon": 0.2021484375, + "loss_wm_recon_w": 0.050537109375, + "step": 520 + }, + { + "epoch": 0.1941856131196422, + "grad_norm": 18.721776962280273, + "learning_rate": 9.593258620824822e-06, + "loss": 11.403581619262695, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.37775951623916626, + "loss_total_v6": 0.5916267037391663, + "loss_wm_recon": 0.20703125, + "loss_wm_recon_w": 0.0517578125, + "step": 521 + }, + { + "epoch": 0.19455833022735744, + "grad_norm": 19.647327423095703, + "learning_rate": 9.590755662126683e-06, + "loss": 11.232388496398926, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5236197113990784, + "loss_total_v6": 0.7335806488990784, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 522 + }, + { + "epoch": 0.1949310473350727, + "grad_norm": 18.183862686157227, + "learning_rate": 9.58824535439166e-06, + "loss": 11.723703384399414, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7272627949714661, + "loss_total_v6": 0.9396651387214661, + "loss_wm_recon": 0.208984375, + "loss_wm_recon_w": 0.05224609375, + "step": 523 + }, + { + "epoch": 0.19530376444278794, + "grad_norm": 18.34073829650879, + "learning_rate": 9.585727701638343e-06, + "loss": 11.842060089111328, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7259827256202698, + "loss_total_v6": 0.7806702256202698, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 524 + }, + { + "epoch": 0.19567648155050316, + "grad_norm": 19.004491806030273, + "learning_rate": 9.583202707897075e-06, + "loss": 11.336633682250977, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.460889995098114, + "loss_total_v6": 0.685743510723114, + "loss_wm_recon": 0.2197265625, + "loss_wm_recon_w": 0.054931640625, + "step": 525 + }, + { + "epoch": 0.1960491986582184, + "grad_norm": 18.89388084411621, + "learning_rate": 9.580670377209949e-06, + "loss": 11.70839786529541, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5880444645881653, + "loss_total_v6": 0.7850659489631653, + "loss_wm_recon": 0.1943359375, + "loss_wm_recon_w": 0.048583984375, + "step": 526 + }, + { + "epoch": 0.19642191576593365, + "grad_norm": 20.569551467895508, + "learning_rate": 9.578130713630809e-06, + "loss": 11.671653747558594, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4704154133796692, + "loss_total_v6": 0.6750052571296692, + "loss_wm_recon": 0.201171875, + "loss_wm_recon_w": 0.05029296875, + "step": 527 + }, + { + "epoch": 0.1967946328736489, + "grad_norm": 19.02135467529297, + "learning_rate": 9.57558372122523e-06, + "loss": 11.39577865600586, + "loss_alignment": 0.35546875, + "loss_alignment_w": 0.177734375, + "loss_sft": 0.5709334015846252, + "loss_total_v6": 0.8165388703346252, + "loss_wm_recon": 0.271484375, + "loss_wm_recon_w": 0.06787109375, + "step": 528 + }, + { + "epoch": 0.19716734998136415, + "grad_norm": 19.667367935180664, + "learning_rate": 9.573029404070529e-06, + "loss": 11.516653060913086, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4719126224517822, + "loss_total_v6": 0.6931040287017822, + "loss_wm_recon": 0.236328125, + "loss_wm_recon_w": 0.05908203125, + "step": 529 + }, + { + "epoch": 0.1975400670890794, + "grad_norm": 19.633161544799805, + "learning_rate": 9.570467766255737e-06, + "loss": 11.499807357788086, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.21876393258571625, + "loss_total_v6": 0.41041433811187744, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 530 + }, + { + "epoch": 0.19791278419679464, + "grad_norm": 19.36609649658203, + "learning_rate": 9.567898811881615e-06, + "loss": 11.908544540405273, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6337742209434509, + "loss_total_v6": 0.8415378928184509, + "loss_wm_recon": 0.2216796875, + "loss_wm_recon_w": 0.055419921875, + "step": 531 + }, + { + "epoch": 0.19828550130450986, + "grad_norm": 20.03304672241211, + "learning_rate": 9.565322545060631e-06, + "loss": 11.908838272094727, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4920947849750519, + "loss_total_v6": 0.6976611614227295, + "loss_wm_recon": 0.205078125, + "loss_wm_recon_w": 0.05126953125, + "step": 532 + }, + { + "epoch": 0.1986582184122251, + "grad_norm": 18.517559051513672, + "learning_rate": 9.56273896991696e-06, + "loss": 11.714945793151855, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5630724430084229, + "loss_total_v6": 0.7766954898834229, + "loss_wm_recon": 0.2216796875, + "loss_wm_recon_w": 0.055419921875, + "step": 533 + }, + { + "epoch": 0.19903093551994036, + "grad_norm": 18.306901931762695, + "learning_rate": 9.560148090586475e-06, + "loss": 11.555675506591797, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.542066216468811, + "loss_total_v6": 0.730542778968811, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 534 + }, + { + "epoch": 0.1994036526276556, + "grad_norm": 18.409420013427734, + "learning_rate": 9.557549911216745e-06, + "loss": 11.407207489013672, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.23237201571464539, + "loss_total_v6": 0.4315907657146454, + "loss_wm_recon": 0.203125, + "loss_wm_recon_w": 0.05078125, + "step": 535 + }, + { + "epoch": 0.19977636973537086, + "grad_norm": 18.975313186645508, + "learning_rate": 9.554944435967024e-06, + "loss": 11.485136032104492, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.3943263292312622, + "loss_total_v6": 0.6084376573562622, + "loss_wm_recon": 0.2158203125, + "loss_wm_recon_w": 0.053955078125, + "step": 536 + }, + { + "epoch": 0.2001490868430861, + "grad_norm": 18.380300521850586, + "learning_rate": 9.552331669008243e-06, + "loss": 11.889177322387695, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.29075920581817627, + "loss_total_v6": 0.49290764331817627, + "loss_wm_recon": 0.20703125, + "loss_wm_recon_w": 0.0517578125, + "step": 537 + }, + { + "epoch": 0.20052180395080135, + "grad_norm": 18.62433433532715, + "learning_rate": 9.549711614523007e-06, + "loss": 11.695327758789062, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.736987292766571, + "loss_total_v6": 0.925463855266571, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 538 + }, + { + "epoch": 0.20089452105851657, + "grad_norm": 18.71712303161621, + "learning_rate": 9.547084276705591e-06, + "loss": 11.38623046875, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.3351944386959076, + "loss_total_v6": 0.55003821849823, + "loss_wm_recon": 0.2109375, + "loss_wm_recon_w": 0.052734375, + "step": 539 + }, + { + "epoch": 0.20126723816623182, + "grad_norm": 19.312692642211914, + "learning_rate": 9.544449659761923e-06, + "loss": 11.36465072631836, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6819218397140503, + "loss_total_v6": 0.8965214490890503, + "loss_wm_recon": 0.2177734375, + "loss_wm_recon_w": 0.054443359375, + "step": 540 + }, + { + "epoch": 0.20163995527394707, + "grad_norm": 19.128278732299805, + "learning_rate": 9.541807767909589e-06, + "loss": 11.631983757019043, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4130162000656128, + "loss_total_v6": 0.6075962781906128, + "loss_wm_recon": 0.1845703125, + "loss_wm_recon_w": 0.046142578125, + "step": 541 + }, + { + "epoch": 0.20201267238166232, + "grad_norm": 18.019197463989258, + "learning_rate": 9.539158605377817e-06, + "loss": 11.516382217407227, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3935355246067047, + "loss_total_v6": 0.5986136198043823, + "loss_wm_recon": 0.203125, + "loss_wm_recon_w": 0.05078125, + "step": 542 + }, + { + "epoch": 0.20238538948937757, + "grad_norm": 18.464155197143555, + "learning_rate": 9.536502176407479e-06, + "loss": 11.434341430664062, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.35652703046798706, + "loss_total_v6": 0.5760094523429871, + "loss_wm_recon": 0.2138671875, + "loss_wm_recon_w": 0.053466796875, + "step": 543 + }, + { + "epoch": 0.2027581065970928, + "grad_norm": 19.828750610351562, + "learning_rate": 9.533838485251073e-06, + "loss": 11.14025592803955, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6131283640861511, + "loss_total_v6": 0.8309018015861511, + "loss_wm_recon": 0.21484375, + "loss_wm_recon_w": 0.0537109375, + "step": 544 + }, + { + "epoch": 0.20313082370480806, + "grad_norm": 19.049007415771484, + "learning_rate": 9.531167536172726e-06, + "loss": 11.421564102172852, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.7744016051292419, + "loss_total_v6": 0.9953488707542419, + "loss_wm_recon": 0.2197265625, + "loss_wm_recon_w": 0.054931640625, + "step": 545 + }, + { + "epoch": 0.20350354081252328, + "grad_norm": 18.978931427001953, + "learning_rate": 9.528489333448186e-06, + "loss": 11.646986961364746, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5324510931968689, + "loss_total_v6": 0.7372850775718689, + "loss_wm_recon": 0.1943359375, + "loss_wm_recon_w": 0.048583984375, + "step": 546 + }, + { + "epoch": 0.20387625792023853, + "grad_norm": 18.290206909179688, + "learning_rate": 9.52580388136481e-06, + "loss": 11.576864242553711, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.5151085257530212, + "loss_total_v6": 0.6957725882530212, + "loss_wm_recon": 0.19140625, + "loss_wm_recon_w": 0.0478515625, + "step": 547 + }, + { + "epoch": 0.20424897502795378, + "grad_norm": 18.433320999145508, + "learning_rate": 9.523111184221559e-06, + "loss": 11.20995807647705, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.8160192966461182, + "loss_total_v6": 1.0291540622711182, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 548 + }, + { + "epoch": 0.20462169213566903, + "grad_norm": 20.055150985717773, + "learning_rate": 9.520411246328995e-06, + "loss": 11.472105026245117, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5935553312301636, + "loss_total_v6": 0.7993658781051636, + "loss_wm_recon": 0.2060546875, + "loss_wm_recon_w": 0.051513671875, + "step": 549 + }, + { + "epoch": 0.20499440924338427, + "grad_norm": 19.445661544799805, + "learning_rate": 9.517704072009269e-06, + "loss": 11.957584381103516, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.1011255532503128, + "loss_total_v6": 0.3120630383491516, + "loss_wm_recon": 0.2109375, + "loss_wm_recon_w": 0.052734375, + "step": 550 + }, + { + "epoch": 0.20536712635109952, + "grad_norm": 19.131135940551758, + "learning_rate": 9.514989665596114e-06, + "loss": 11.181144714355469, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.49308648705482483, + "loss_total_v6": 0.7064653635025024, + "loss_wm_recon": 0.212890625, + "loss_wm_recon_w": 0.05322265625, + "step": 551 + }, + { + "epoch": 0.20573984345881477, + "grad_norm": 19.03874397277832, + "learning_rate": 9.512268031434852e-06, + "loss": 11.72032356262207, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.6799632906913757, + "loss_total_v6": 0.9065257906913757, + "loss_wm_recon": 0.2265625, + "loss_wm_recon_w": 0.056640625, + "step": 552 + }, + { + "epoch": 0.20611256056653002, + "grad_norm": 19.159059524536133, + "learning_rate": 9.50953917388236e-06, + "loss": 12.014796257019043, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.7495363354682922, + "loss_total_v6": 0.9385011792182922, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 553 + }, + { + "epoch": 0.20648527767424524, + "grad_norm": 19.274654388427734, + "learning_rate": 9.506803097307087e-06, + "loss": 11.646657943725586, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.8037415146827698, + "loss_total_v6": 1.021026611328125, + "loss_wm_recon": 0.205078125, + "loss_wm_recon_w": 0.05126953125, + "step": 554 + }, + { + "epoch": 0.20685799478196049, + "grad_norm": 18.499727249145508, + "learning_rate": 9.504059806089037e-06, + "loss": 11.254490852355957, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.3605678975582123, + "loss_total_v6": 0.5602749586105347, + "loss_wm_recon": 0.228515625, + "loss_wm_recon_w": 0.05712890625, + "step": 555 + }, + { + "epoch": 0.20723071188967573, + "grad_norm": 19.509754180908203, + "learning_rate": 9.501309304619765e-06, + "loss": 11.404006958007812, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.19700877368450165, + "loss_total_v6": 0.39744824171066284, + "loss_wm_recon": 0.2001953125, + "loss_wm_recon_w": 0.050048828125, + "step": 556 + }, + { + "epoch": 0.20760342899739098, + "grad_norm": 18.62493896484375, + "learning_rate": 9.498551597302364e-06, + "loss": 11.288345336914062, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5540919899940491, + "loss_total_v6": 0.7652736306190491, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 557 + }, + { + "epoch": 0.20797614610510623, + "grad_norm": 19.460655212402344, + "learning_rate": 9.495786688551467e-06, + "loss": 11.585187911987305, + "loss_alignment": 0.359375, + "loss_alignment_w": 0.1796875, + "loss_sft": 0.8984366059303284, + "loss_total_v6": 1.1291494369506836, + "loss_wm_recon": 0.2041015625, + "loss_wm_recon_w": 0.051025390625, + "step": 558 + }, + { + "epoch": 0.20834886321282148, + "grad_norm": 19.332012176513672, + "learning_rate": 9.493014582793234e-06, + "loss": 11.365863800048828, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.13273267447948456, + "loss_total_v6": 0.33024245500564575, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 559 + }, + { + "epoch": 0.20872158032053673, + "grad_norm": 19.574291229248047, + "learning_rate": 9.490235284465343e-06, + "loss": 11.196717262268066, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3479432761669159, + "loss_total_v6": 0.5478944778442383, + "loss_wm_recon": 0.1982421875, + "loss_wm_recon_w": 0.049560546875, + "step": 560 + }, + { + "epoch": 0.20909429742825195, + "grad_norm": 19.219881057739258, + "learning_rate": 9.48744879801699e-06, + "loss": 11.776153564453125, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.44890668988227844, + "loss_total_v6": 0.641045331954956, + "loss_wm_recon": 0.1826171875, + "loss_wm_recon_w": 0.045654296875, + "step": 561 + }, + { + "epoch": 0.2094670145359672, + "grad_norm": 18.905088424682617, + "learning_rate": 9.484655127908878e-06, + "loss": 11.749645233154297, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6743651628494263, + "loss_total_v6": 0.8850585222244263, + "loss_wm_recon": 0.1943359375, + "loss_wm_recon_w": 0.048583984375, + "step": 562 + }, + { + "epoch": 0.20983973164368244, + "grad_norm": 19.501283645629883, + "learning_rate": 9.481854278613208e-06, + "loss": 11.688055038452148, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.36332836747169495, + "loss_total_v6": 0.5701154470443726, + "loss_wm_recon": 0.1787109375, + "loss_wm_recon_w": 0.044677734375, + "step": 563 + }, + { + "epoch": 0.2102124487513977, + "grad_norm": 18.349464416503906, + "learning_rate": 9.479046254613673e-06, + "loss": 11.490503311157227, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3790295124053955, + "loss_total_v6": 0.5809338092803955, + "loss_wm_recon": 0.1904296875, + "loss_wm_recon_w": 0.047607421875, + "step": 564 + }, + { + "epoch": 0.21058516585911294, + "grad_norm": 17.56845474243164, + "learning_rate": 9.476231060405452e-06, + "loss": 11.761730194091797, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3962058126926422, + "loss_total_v6": 0.5951803922653198, + "loss_wm_recon": 0.1865234375, + "loss_wm_recon_w": 0.046630859375, + "step": 565 + }, + { + "epoch": 0.2109578829668282, + "grad_norm": 19.48756217956543, + "learning_rate": 9.473408700495207e-06, + "loss": 11.133480072021484, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5541075468063354, + "loss_total_v6": 0.7545469999313354, + "loss_wm_recon": 0.1923828125, + "loss_wm_recon_w": 0.048095703125, + "step": 566 + }, + { + "epoch": 0.21133060007454343, + "grad_norm": 19.382518768310547, + "learning_rate": 9.470579179401065e-06, + "loss": 11.060298919677734, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.7100681066513062, + "loss_total_v6": 0.9556735754013062, + "loss_wm_recon": 0.248046875, + "loss_wm_recon_w": 0.06201171875, + "step": 567 + }, + { + "epoch": 0.21170331718225865, + "grad_norm": 19.477445602416992, + "learning_rate": 9.467742501652618e-06, + "loss": 11.311279296875, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.6174470782279968, + "loss_total_v6": 0.8064119219779968, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 568 + }, + { + "epoch": 0.2120760342899739, + "grad_norm": 18.45405387878418, + "learning_rate": 9.464898671790917e-06, + "loss": 11.413270950317383, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5366863012313843, + "loss_total_v6": 0.7534831762313843, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 569 + }, + { + "epoch": 0.21244875139768915, + "grad_norm": 17.730186462402344, + "learning_rate": 9.462047694368459e-06, + "loss": 11.628411293029785, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.3540576994419098, + "loss_total_v6": 0.5713428258895874, + "loss_wm_recon": 0.220703125, + "loss_wm_recon_w": 0.05517578125, + "step": 570 + }, + { + "epoch": 0.2128214685054044, + "grad_norm": 19.519737243652344, + "learning_rate": 9.459189573949187e-06, + "loss": 12.035402297973633, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.9108052253723145, + "loss_total_v6": 1.1285786628723145, + "loss_wm_recon": 0.21484375, + "loss_wm_recon_w": 0.0537109375, + "step": 571 + }, + { + "epoch": 0.21319418561311965, + "grad_norm": 19.101057052612305, + "learning_rate": 9.456324315108476e-06, + "loss": 11.434345245361328, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3936409056186676, + "loss_total_v6": 0.5909065008163452, + "loss_wm_recon": 0.1875, + "loss_wm_recon_w": 0.046875, + "step": 572 + }, + { + "epoch": 0.2135669027208349, + "grad_norm": 18.811859130859375, + "learning_rate": 9.45345192243313e-06, + "loss": 11.632303237915039, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.41878828406333923, + "loss_total_v6": 0.6123918294906616, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 573 + }, + { + "epoch": 0.21393961982855014, + "grad_norm": 18.638071060180664, + "learning_rate": 9.450572400521371e-06, + "loss": 11.392507553100586, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.46554118394851685, + "loss_total_v6": 0.6701310276985168, + "loss_wm_recon": 0.201171875, + "loss_wm_recon_w": 0.05029296875, + "step": 574 + }, + { + "epoch": 0.21431233693626536, + "grad_norm": 18.662446975708008, + "learning_rate": 9.447685753982837e-06, + "loss": 11.287070274353027, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7861669063568115, + "loss_total_v6": 0.9900243282318115, + "loss_wm_recon": 0.1982421875, + "loss_wm_recon_w": 0.049560546875, + "step": 575 + }, + { + "epoch": 0.2146850540439806, + "grad_norm": 18.71076774597168, + "learning_rate": 9.44479198743857e-06, + "loss": 11.480387687683105, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.306871622800827, + "loss_total_v6": 0.5073111057281494, + "loss_wm_recon": 0.1923828125, + "loss_wm_recon_w": 0.048095703125, + "step": 576 + }, + { + "epoch": 0.21505777115169586, + "grad_norm": 18.89326286315918, + "learning_rate": 9.441891105521005e-06, + "loss": 11.143659591674805, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.29758551716804504, + "loss_total_v6": 0.49948981404304504, + "loss_wm_recon": 0.1904296875, + "loss_wm_recon_w": 0.047607421875, + "step": 577 + }, + { + "epoch": 0.2154304882594111, + "grad_norm": 18.416440963745117, + "learning_rate": 9.438983112873977e-06, + "loss": 11.321608543395996, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5822997689247131, + "loss_total_v6": 0.7790771126747131, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 578 + }, + { + "epoch": 0.21580320536712636, + "grad_norm": 18.501556396484375, + "learning_rate": 9.436068014152692e-06, + "loss": 11.207778930664062, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.29743432998657227, + "loss_total_v6": 0.49323511123657227, + "loss_wm_recon": 0.189453125, + "loss_wm_recon_w": 0.04736328125, + "step": 579 + }, + { + "epoch": 0.2161759224748416, + "grad_norm": 17.904767990112305, + "learning_rate": 9.433145814023747e-06, + "loss": 11.365896224975586, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3024733364582062, + "loss_total_v6": 0.5019361972808838, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 580 + }, + { + "epoch": 0.21654863958255685, + "grad_norm": 18.259260177612305, + "learning_rate": 9.430216517165094e-06, + "loss": 11.391763687133789, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.8699410557746887, + "loss_total_v6": 1.088446855545044, + "loss_wm_recon": 0.2099609375, + "loss_wm_recon_w": 0.052490234375, + "step": 581 + }, + { + "epoch": 0.21692135669027207, + "grad_norm": 19.705955505371094, + "learning_rate": 9.427280128266049e-06, + "loss": 11.527897834777832, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.4030055105686188, + "loss_total_v6": 0.6251734495162964, + "loss_wm_recon": 0.224609375, + "loss_wm_recon_w": 0.05615234375, + "step": 582 + }, + { + "epoch": 0.21729407379798732, + "grad_norm": 18.5332088470459, + "learning_rate": 9.424336652027288e-06, + "loss": 11.58150863647461, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5915199518203735, + "loss_total_v6": 0.7885414361953735, + "loss_wm_recon": 0.1865234375, + "loss_wm_recon_w": 0.046630859375, + "step": 583 + }, + { + "epoch": 0.21766679090570257, + "grad_norm": 19.652429580688477, + "learning_rate": 9.421386093160823e-06, + "loss": 11.749455451965332, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7985249161720276, + "loss_total_v6": 1.0009174346923828, + "loss_wm_recon": 0.1923828125, + "loss_wm_recon_w": 0.048095703125, + "step": 584 + }, + { + "epoch": 0.21803950801341782, + "grad_norm": 18.561006546020508, + "learning_rate": 9.418428456390008e-06, + "loss": 11.678123474121094, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6175951957702637, + "loss_total_v6": 0.8299975395202637, + "loss_wm_recon": 0.201171875, + "loss_wm_recon_w": 0.05029296875, + "step": 585 + }, + { + "epoch": 0.21841222512113306, + "grad_norm": 18.15987205505371, + "learning_rate": 9.415463746449532e-06, + "loss": 11.454496383666992, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.21777129173278809, + "loss_total_v6": 0.4267556667327881, + "loss_wm_recon": 0.2265625, + "loss_wm_recon_w": 0.056640625, + "step": 586 + }, + { + "epoch": 0.2187849422288483, + "grad_norm": 18.699390411376953, + "learning_rate": 9.4124919680854e-06, + "loss": 11.56719970703125, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7006692290306091, + "loss_total_v6": 0.7497414946556091, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 587 + }, + { + "epoch": 0.21915765933656356, + "grad_norm": 18.494022369384766, + "learning_rate": 9.409513126054934e-06, + "loss": 11.237613677978516, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.9584671258926392, + "loss_total_v6": 1.1637893915176392, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 588 + }, + { + "epoch": 0.21953037644427878, + "grad_norm": 19.200366973876953, + "learning_rate": 9.406527225126767e-06, + "loss": 11.372749328613281, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5614315271377563, + "loss_total_v6": 0.7648006677627563, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 589 + }, + { + "epoch": 0.21990309355199403, + "grad_norm": 19.631643295288086, + "learning_rate": 9.40353427008083e-06, + "loss": 11.493562698364258, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.7438088059425354, + "loss_total_v6": 0.9444923996925354, + "loss_wm_recon": 0.193359375, + "loss_wm_recon_w": 0.04833984375, + "step": 590 + }, + { + "epoch": 0.22027581065970928, + "grad_norm": 18.66900634765625, + "learning_rate": 9.400534265708348e-06, + "loss": 11.564759254455566, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5876560807228088, + "loss_total_v6": 0.7917576432228088, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 591 + }, + { + "epoch": 0.22064852776742452, + "grad_norm": 18.9291934967041, + "learning_rate": 9.397527216811827e-06, + "loss": 11.4781494140625, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.42411527037620544, + "loss_total_v6": 0.6157656908035278, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 592 + }, + { + "epoch": 0.22102124487513977, + "grad_norm": 19.30915641784668, + "learning_rate": 9.394513128205052e-06, + "loss": 11.182174682617188, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.40377068519592285, + "loss_total_v6": 0.6149523258209229, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 593 + }, + { + "epoch": 0.22139396198285502, + "grad_norm": 19.17876625061035, + "learning_rate": 9.39149200471308e-06, + "loss": 11.442878723144531, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.2537783682346344, + "loss_total_v6": 0.4498232901096344, + "loss_wm_recon": 0.1904296875, + "loss_wm_recon_w": 0.047607421875, + "step": 594 + }, + { + "epoch": 0.22176667909057027, + "grad_norm": 18.53878402709961, + "learning_rate": 9.38846385117223e-06, + "loss": 10.870019912719727, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6998717188835144, + "loss_total_v6": 0.9054381251335144, + "loss_wm_recon": 0.197265625, + "loss_wm_recon_w": 0.04931640625, + "step": 595 + }, + { + "epoch": 0.2221393961982855, + "grad_norm": 17.791540145874023, + "learning_rate": 9.385428672430066e-06, + "loss": 11.60394287109375, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3494320213794708, + "loss_total_v6": 0.5349788665771484, + "loss_wm_recon": 0.1796875, + "loss_wm_recon_w": 0.044921875, + "step": 596 + }, + { + "epoch": 0.22251211330600074, + "grad_norm": 18.325780868530273, + "learning_rate": 9.382386473345411e-06, + "loss": 11.367005348205566, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6044634580612183, + "loss_total_v6": 0.8134478330612183, + "loss_wm_recon": 0.1953125, + "loss_wm_recon_w": 0.048828125, + "step": 597 + }, + { + "epoch": 0.22288483041371598, + "grad_norm": 18.60590171813965, + "learning_rate": 9.379337258788316e-06, + "loss": 10.919450759887695, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.6808759570121765, + "loss_total_v6": 0.8649579882621765, + "loss_wm_recon": 0.181640625, + "loss_wm_recon_w": 0.04541015625, + "step": 598 + }, + { + "epoch": 0.22325754752143123, + "grad_norm": 19.23051643371582, + "learning_rate": 9.376281033640068e-06, + "loss": 11.554670333862305, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.7345584630966187, + "loss_total_v6": 0.9411014318466187, + "loss_wm_recon": 0.193359375, + "loss_wm_recon_w": 0.04833984375, + "step": 599 + }, + { + "epoch": 0.22363026462914648, + "grad_norm": 18.297542572021484, + "learning_rate": 9.373217802793177e-06, + "loss": 11.501348495483398, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6127470135688782, + "loss_total_v6": 0.6610868573188782, + "loss_wm_recon": 0.193359375, + "loss_wm_recon_w": 0.04833984375, + "step": 600 + }, + { + "epoch": 0.22363026462914648, + "eval_loss": 0.683212399482727, + "eval_loss_alignment": 0.2875475761217949, + "eval_loss_alignment_w": 0.14377378806089744, + "eval_loss_sft": 0.4936057537014031, + "eval_loss_total_v6": 0.683212379566752, + "eval_loss_wm_recon": 0.1833313509949252, + "eval_loss_wm_recon_w": 0.0458328377487313, + "eval_runtime": 227.6166, + "eval_samples_per_second": 8.198, + "eval_steps_per_second": 1.028, + "step": 600 + }, + { + "epoch": 0.22400298173686173, + "grad_norm": 19.43748664855957, + "learning_rate": 9.370147571151364e-06, + "loss": 11.309226989746094, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.2973366677761078, + "loss_total_v6": 0.5058327913284302, + "loss_wm_recon": 0.193359375, + "loss_wm_recon_w": 0.04833984375, + "step": 601 + }, + { + "epoch": 0.22437569884457698, + "grad_norm": 18.72801971435547, + "learning_rate": 9.367070343629562e-06, + "loss": 11.180782318115234, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.5786223411560059, + "loss_total_v6": 0.7971282005310059, + "loss_wm_recon": 0.2099609375, + "loss_wm_recon_w": 0.052490234375, + "step": 602 + }, + { + "epoch": 0.2247484159522922, + "grad_norm": 18.758522033691406, + "learning_rate": 9.3639861251539e-06, + "loss": 11.760799407958984, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.547668993473053, + "loss_total_v6": 0.734680712223053, + "loss_wm_recon": 0.177734375, + "loss_wm_recon_w": 0.04443359375, + "step": 603 + }, + { + "epoch": 0.22512113306000744, + "grad_norm": 18.73381996154785, + "learning_rate": 9.3608949206617e-06, + "loss": 11.423620223999023, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.3845919668674469, + "loss_total_v6": 0.5982149839401245, + "loss_wm_recon": 0.2060546875, + "loss_wm_recon_w": 0.051513671875, + "step": 604 + }, + { + "epoch": 0.2254938501677227, + "grad_norm": 20.163318634033203, + "learning_rate": 9.357796735101463e-06, + "loss": 11.815756797790527, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2953304946422577, + "loss_total_v6": 0.4923519790172577, + "loss_wm_recon": 0.1865234375, + "loss_wm_recon_w": 0.046630859375, + "step": 605 + }, + { + "epoch": 0.22586656727543794, + "grad_norm": 18.367061614990234, + "learning_rate": 9.354691573432875e-06, + "loss": 11.473675727844238, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5996421575546265, + "loss_total_v6": 0.8037437200546265, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 606 + }, + { + "epoch": 0.2262392843831532, + "grad_norm": 18.60525131225586, + "learning_rate": 9.35157944062678e-06, + "loss": 11.434455871582031, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.45681992173194885, + "loss_total_v6": 0.6589683294296265, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 607 + }, + { + "epoch": 0.22661200149086844, + "grad_norm": 18.139556884765625, + "learning_rate": 9.348460341665184e-06, + "loss": 10.80173397064209, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.48072072863578796, + "loss_total_v6": 0.6833574771881104, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 608 + }, + { + "epoch": 0.22698471859858368, + "grad_norm": 18.93004608154297, + "learning_rate": 9.34533428154125e-06, + "loss": 11.249388694763184, + "loss_alignment": 0.375, + "loss_alignment_w": 0.1875, + "loss_sft": 0.6517637372016907, + "loss_total_v6": 0.8998106122016907, + "loss_wm_recon": 0.2421875, + "loss_wm_recon_w": 0.060546875, + "step": 609 + }, + { + "epoch": 0.22735743570629893, + "grad_norm": 19.04774284362793, + "learning_rate": 9.342201265259277e-06, + "loss": 11.063532829284668, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4068918228149414, + "loss_total_v6": 0.6095285415649414, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 610 + }, + { + "epoch": 0.22773015281401415, + "grad_norm": 18.2890567779541, + "learning_rate": 9.339061297834708e-06, + "loss": 11.321370124816895, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.8293566703796387, + "loss_total_v6": 1.0568957328796387, + "loss_wm_recon": 0.21484375, + "loss_wm_recon_w": 0.0537109375, + "step": 611 + }, + { + "epoch": 0.2281028699217294, + "grad_norm": 18.367403030395508, + "learning_rate": 9.335914384294105e-06, + "loss": 11.295036315917969, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4493926167488098, + "loss_total_v6": 0.6466582417488098, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 612 + }, + { + "epoch": 0.22847558702944465, + "grad_norm": 17.461292266845703, + "learning_rate": 9.332760529675154e-06, + "loss": 11.06434440612793, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.387751966714859, + "loss_total_v6": 0.6025956869125366, + "loss_wm_recon": 0.21875, + "loss_wm_recon_w": 0.0546875, + "step": 613 + }, + { + "epoch": 0.2288483041371599, + "grad_norm": 17.847305297851562, + "learning_rate": 9.329599739026655e-06, + "loss": 11.041646003723145, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7226378917694092, + "loss_total_v6": 0.9338195323944092, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 614 + }, + { + "epoch": 0.22922102124487514, + "grad_norm": 18.186323165893555, + "learning_rate": 9.326432017408506e-06, + "loss": 11.089080810546875, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6555778384208679, + "loss_total_v6": 0.8677360415458679, + "loss_wm_recon": 0.2001953125, + "loss_wm_recon_w": 0.050048828125, + "step": 615 + }, + { + "epoch": 0.2295937383525904, + "grad_norm": 18.913055419921875, + "learning_rate": 9.323257369891702e-06, + "loss": 11.627847671508789, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5254105925559998, + "loss_total_v6": 0.7216996550559998, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 616 + }, + { + "epoch": 0.22996645546030564, + "grad_norm": 19.130239486694336, + "learning_rate": 9.32007580155833e-06, + "loss": 11.17888355255127, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4279932677745819, + "loss_total_v6": 0.6338038444519043, + "loss_wm_recon": 0.1826171875, + "loss_wm_recon_w": 0.045654296875, + "step": 617 + }, + { + "epoch": 0.23033917256802086, + "grad_norm": 18.406665802001953, + "learning_rate": 9.316887317501549e-06, + "loss": 11.45753002166748, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6662705540657043, + "loss_total_v6": 0.8630478978157043, + "loss_wm_recon": 0.177734375, + "loss_wm_recon_w": 0.04443359375, + "step": 618 + }, + { + "epoch": 0.2307118896757361, + "grad_norm": 19.47599983215332, + "learning_rate": 9.313691922825592e-06, + "loss": 11.537994384765625, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6378594040870667, + "loss_total_v6": 0.8636894822120667, + "loss_wm_recon": 0.2314453125, + "loss_wm_recon_w": 0.057861328125, + "step": 619 + }, + { + "epoch": 0.23108460678345136, + "grad_norm": 18.97117042541504, + "learning_rate": 9.310489622645756e-06, + "loss": 11.437297821044922, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5458651185035706, + "loss_total_v6": 0.7541170716285706, + "loss_wm_recon": 0.1923828125, + "loss_wm_recon_w": 0.048095703125, + "step": 620 + }, + { + "epoch": 0.2314573238911666, + "grad_norm": 19.434478759765625, + "learning_rate": 9.30728042208839e-06, + "loss": 11.901823997497559, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6163490414619446, + "loss_total_v6": 0.8145912289619446, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 621 + }, + { + "epoch": 0.23183004099888185, + "grad_norm": 17.88884162902832, + "learning_rate": 9.304064326290894e-06, + "loss": 11.023761749267578, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6876418590545654, + "loss_total_v6": 0.9054152965545654, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 622 + }, + { + "epoch": 0.2322027581065971, + "grad_norm": 19.17205047607422, + "learning_rate": 9.300841340401699e-06, + "loss": 11.458474159240723, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.8051641583442688, + "loss_total_v6": 1.007556676864624, + "loss_wm_recon": 0.1767578125, + "loss_wm_recon_w": 0.044189453125, + "step": 623 + }, + { + "epoch": 0.23257547521431235, + "grad_norm": 18.042591094970703, + "learning_rate": 9.297611469580272e-06, + "loss": 11.134865760803223, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.43563154339790344, + "loss_total_v6": 0.628746747970581, + "loss_wm_recon": 0.1787109375, + "loss_wm_recon_w": 0.044677734375, + "step": 624 + }, + { + "epoch": 0.23294819232202757, + "grad_norm": 18.319656372070312, + "learning_rate": 9.2943747189971e-06, + "loss": 11.206107139587402, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6571556329727173, + "loss_total_v6": 0.8661400079727173, + "loss_wm_recon": 0.1953125, + "loss_wm_recon_w": 0.048828125, + "step": 625 + }, + { + "epoch": 0.23332090942974282, + "grad_norm": 18.401552200317383, + "learning_rate": 9.291131093833678e-06, + "loss": 11.557129859924316, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.4690782129764557, + "loss_total_v6": 0.6751328706741333, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 626 + }, + { + "epoch": 0.23369362653745807, + "grad_norm": 18.096349716186523, + "learning_rate": 9.287880599282517e-06, + "loss": 10.739750862121582, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.43571358919143677, + "loss_total_v6": 0.6356647610664368, + "loss_wm_recon": 0.1826171875, + "loss_wm_recon_w": 0.045654296875, + "step": 627 + }, + { + "epoch": 0.2340663436451733, + "grad_norm": 18.128679275512695, + "learning_rate": 9.284623240547117e-06, + "loss": 11.248051643371582, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7785777449607849, + "loss_total_v6": 0.8254527449607849, + "loss_wm_recon": 0.1875, + "loss_wm_recon_w": 0.046875, + "step": 628 + }, + { + "epoch": 0.23443906075288856, + "grad_norm": 18.269075393676758, + "learning_rate": 9.281359022841966e-06, + "loss": 11.294702529907227, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.8617718815803528, + "loss_total_v6": 1.070267915725708, + "loss_wm_recon": 0.201171875, + "loss_wm_recon_w": 0.05029296875, + "step": 629 + }, + { + "epoch": 0.2348117778606038, + "grad_norm": 18.3060245513916, + "learning_rate": 9.278087951392536e-06, + "loss": 11.425514221191406, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5645918250083923, + "loss_total_v6": 0.7574629187583923, + "loss_wm_recon": 0.177734375, + "loss_wm_recon_w": 0.04443359375, + "step": 630 + }, + { + "epoch": 0.23518449496831906, + "grad_norm": 17.54067611694336, + "learning_rate": 9.274810031435267e-06, + "loss": 11.398565292358398, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.52314692735672, + "loss_total_v6": 0.71260005235672, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 631 + }, + { + "epoch": 0.23555721207603428, + "grad_norm": 17.613727569580078, + "learning_rate": 9.271525268217564e-06, + "loss": 11.516745567321777, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.4286928176879883, + "loss_total_v6": 0.6647768020629883, + "loss_wm_recon": 0.2177734375, + "loss_wm_recon_w": 0.054443359375, + "step": 632 + }, + { + "epoch": 0.23592992918374953, + "grad_norm": 19.669525146484375, + "learning_rate": 9.268233666997789e-06, + "loss": 11.070870399475098, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6187655329704285, + "loss_total_v6": 0.8187167048454285, + "loss_wm_recon": 0.1748046875, + "loss_wm_recon_w": 0.043701171875, + "step": 633 + }, + { + "epoch": 0.23630264629146477, + "grad_norm": 18.270343780517578, + "learning_rate": 9.264935233045248e-06, + "loss": 11.147068977355957, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6862545609474182, + "loss_total_v6": 0.8974362015724182, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 634 + }, + { + "epoch": 0.23667536339918002, + "grad_norm": 18.769258499145508, + "learning_rate": 9.261629971640185e-06, + "loss": 11.115280151367188, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4617498219013214, + "loss_total_v6": 0.661700963973999, + "loss_wm_recon": 0.1826171875, + "loss_wm_recon_w": 0.045654296875, + "step": 635 + }, + { + "epoch": 0.23704808050689527, + "grad_norm": 19.443811416625977, + "learning_rate": 9.258317888073773e-06, + "loss": 11.523392677307129, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.47782593965530396, + "loss_total_v6": 0.685589611530304, + "loss_wm_recon": 0.1904296875, + "loss_wm_recon_w": 0.047607421875, + "step": 636 + }, + { + "epoch": 0.23742079761461052, + "grad_norm": 18.566993713378906, + "learning_rate": 9.25499898764811e-06, + "loss": 11.811761856079102, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5695509314537048, + "loss_total_v6": 0.7633985877037048, + "loss_wm_recon": 0.173828125, + "loss_wm_recon_w": 0.04345703125, + "step": 637 + }, + { + "epoch": 0.23779351472232577, + "grad_norm": 19.089534759521484, + "learning_rate": 9.251673275676203e-06, + "loss": 11.405553817749023, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.5133590698242188, + "loss_total_v6": 0.6935348510742188, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 638 + }, + { + "epoch": 0.238166231830041, + "grad_norm": 18.738473892211914, + "learning_rate": 9.248340757481965e-06, + "loss": 11.372932434082031, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.40301603078842163, + "loss_total_v6": 0.5968636870384216, + "loss_wm_recon": 0.173828125, + "loss_wm_recon_w": 0.04345703125, + "step": 639 + }, + { + "epoch": 0.23853894893775623, + "grad_norm": 19.369007110595703, + "learning_rate": 9.2450014384002e-06, + "loss": 10.900022506713867, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.19171220064163208, + "loss_total_v6": 0.4021614193916321, + "loss_wm_recon": 0.216796875, + "loss_wm_recon_w": 0.05419921875, + "step": 640 + }, + { + "epoch": 0.23891166604547148, + "grad_norm": 18.420644760131836, + "learning_rate": 9.241655323776606e-06, + "loss": 11.17026138305664, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6793969869613647, + "loss_total_v6": 0.8793481588363647, + "loss_wm_recon": 0.1826171875, + "loss_wm_recon_w": 0.045654296875, + "step": 641 + }, + { + "epoch": 0.23928438315318673, + "grad_norm": 18.542224884033203, + "learning_rate": 9.238302418967757e-06, + "loss": 11.260557174682617, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.38637298345565796, + "loss_total_v6": 0.583882749080658, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 642 + }, + { + "epoch": 0.23965710026090198, + "grad_norm": 19.442155838012695, + "learning_rate": 9.234942729341096e-06, + "loss": 11.04030990600586, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4252205491065979, + "loss_total_v6": 0.6156502366065979, + "loss_wm_recon": 0.16015625, + "loss_wm_recon_w": 0.0400390625, + "step": 643 + }, + { + "epoch": 0.24002981736861723, + "grad_norm": 18.462038040161133, + "learning_rate": 9.231576260274926e-06, + "loss": 11.449230194091797, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5226579308509827, + "loss_total_v6": 0.7213883996009827, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 644 + }, + { + "epoch": 0.24040253447633247, + "grad_norm": 18.53040313720703, + "learning_rate": 9.228203017158406e-06, + "loss": 11.837604522705078, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5337145328521729, + "loss_total_v6": 0.7229235172271729, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 645 + }, + { + "epoch": 0.2407752515840477, + "grad_norm": 18.734481811523438, + "learning_rate": 9.224823005391537e-06, + "loss": 11.557422637939453, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5904735922813416, + "loss_total_v6": 0.8011669516563416, + "loss_wm_recon": 0.1865234375, + "loss_wm_recon_w": 0.046630859375, + "step": 646 + }, + { + "epoch": 0.24114796869176294, + "grad_norm": 18.310319900512695, + "learning_rate": 9.221436230385157e-06, + "loss": 11.127309799194336, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 1.023357629776001, + "loss_total_v6": 1.222332239151001, + "loss_wm_recon": 0.1865234375, + "loss_wm_recon_w": 0.046630859375, + "step": 647 + }, + { + "epoch": 0.2415206857994782, + "grad_norm": 17.917394638061523, + "learning_rate": 9.218042697560928e-06, + "loss": 11.302465438842773, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.7409822344779968, + "loss_total_v6": 0.9394685626029968, + "loss_wm_recon": 0.1845703125, + "loss_wm_recon_w": 0.046142578125, + "step": 648 + }, + { + "epoch": 0.24189340290719344, + "grad_norm": 18.696563720703125, + "learning_rate": 9.214642412351334e-06, + "loss": 11.432767868041992, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.24638451635837555, + "loss_total_v6": 0.46269309520721436, + "loss_wm_recon": 0.201171875, + "loss_wm_recon_w": 0.05029296875, + "step": 649 + }, + { + "epoch": 0.2422661200149087, + "grad_norm": 18.395906448364258, + "learning_rate": 9.211235380199666e-06, + "loss": 11.167657852172852, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.44806990027427673, + "loss_total_v6": 0.6502183675765991, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 650 + }, + { + "epoch": 0.24263883712262393, + "grad_norm": 18.76120948791504, + "learning_rate": 9.207821606560017e-06, + "loss": 11.192625999450684, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.4920019805431366, + "loss_total_v6": 0.673154354095459, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 651 + }, + { + "epoch": 0.24301155423033918, + "grad_norm": 18.321334838867188, + "learning_rate": 9.204401096897271e-06, + "loss": 11.466813087463379, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6041194796562195, + "loss_total_v6": 0.8065120577812195, + "loss_wm_recon": 0.1923828125, + "loss_wm_recon_w": 0.048095703125, + "step": 652 + }, + { + "epoch": 0.2433842713380544, + "grad_norm": 20.4305362701416, + "learning_rate": 9.200973856687093e-06, + "loss": 10.756723403930664, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.30972346663475037, + "loss_total_v6": 0.5248113870620728, + "loss_wm_recon": 0.2041015625, + "loss_wm_recon_w": 0.051025390625, + "step": 653 + }, + { + "epoch": 0.24375698844576965, + "grad_norm": 19.622840881347656, + "learning_rate": 9.197539891415928e-06, + "loss": 11.53824520111084, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.19278131425380707, + "loss_total_v6": 0.23233209550380707, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 654 + }, + { + "epoch": 0.2441297055534849, + "grad_norm": 18.368314743041992, + "learning_rate": 9.194099206580981e-06, + "loss": 11.129281044006348, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.8034856915473938, + "loss_total_v6": 0.8474310040473938, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 655 + }, + { + "epoch": 0.24450242266120015, + "grad_norm": 17.39061164855957, + "learning_rate": 9.190651807690221e-06, + "loss": 11.03170394897461, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.23698914051055908, + "loss_total_v6": 0.4449969530105591, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 656 + }, + { + "epoch": 0.2448751397689154, + "grad_norm": 19.641910552978516, + "learning_rate": 9.187197700262359e-06, + "loss": 10.489980697631836, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5141644477844238, + "loss_total_v6": 0.7163128852844238, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 657 + }, + { + "epoch": 0.24524785687663064, + "grad_norm": 19.33245086669922, + "learning_rate": 9.183736889826842e-06, + "loss": 11.410552978515625, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5408626794815063, + "loss_total_v6": 0.7286068201065063, + "loss_wm_recon": 0.1650390625, + "loss_wm_recon_w": 0.041259765625, + "step": 658 + }, + { + "epoch": 0.2456205739843459, + "grad_norm": 19.64995002746582, + "learning_rate": 9.180269381923859e-06, + "loss": 11.465497970581055, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5315073728561401, + "loss_total_v6": 0.7370737791061401, + "loss_wm_recon": 0.181640625, + "loss_wm_recon_w": 0.04541015625, + "step": 659 + }, + { + "epoch": 0.24599329109206114, + "grad_norm": 18.078763961791992, + "learning_rate": 9.17679518210431e-06, + "loss": 11.066855430603027, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.8649060130119324, + "loss_total_v6": 1.0724256038665771, + "loss_wm_recon": 0.197265625, + "loss_wm_recon_w": 0.04931640625, + "step": 660 + }, + { + "epoch": 0.24636600819977636, + "grad_norm": 18.005203247070312, + "learning_rate": 9.173314295929814e-06, + "loss": 11.436785697937012, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4143352806568146, + "loss_total_v6": 0.6030559539794922, + "loss_wm_recon": 0.1611328125, + "loss_wm_recon_w": 0.040283203125, + "step": 661 + }, + { + "epoch": 0.2467387253074916, + "grad_norm": 17.72521209716797, + "learning_rate": 9.169826728972688e-06, + "loss": 11.350900650024414, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.532723605632782, + "loss_total_v6": 0.729256808757782, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 662 + }, + { + "epoch": 0.24711144241520686, + "grad_norm": 17.633317947387695, + "learning_rate": 9.166332486815952e-06, + "loss": 11.467840194702148, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6300952434539795, + "loss_total_v6": 0.8261401653289795, + "loss_wm_recon": 0.1748046875, + "loss_wm_recon_w": 0.043701171875, + "step": 663 + }, + { + "epoch": 0.2474841595229221, + "grad_norm": 17.853357315063477, + "learning_rate": 9.162831575053302e-06, + "loss": 11.3939790725708, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.21902437508106232, + "loss_total_v6": 0.4106747508049011, + "loss_wm_recon": 0.1650390625, + "loss_wm_recon_w": 0.041259765625, + "step": 664 + }, + { + "epoch": 0.24785687663063735, + "grad_norm": 18.6029052734375, + "learning_rate": 9.159323999289119e-06, + "loss": 11.298269271850586, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5029868483543396, + "loss_total_v6": 0.7019614577293396, + "loss_wm_recon": 0.1943359375, + "loss_wm_recon_w": 0.048583984375, + "step": 665 + }, + { + "epoch": 0.2482295937383526, + "grad_norm": 18.122295379638672, + "learning_rate": 9.155809765138447e-06, + "loss": 10.537416458129883, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.31439444422721863, + "loss_total_v6": 0.504091739654541, + "loss_wm_recon": 0.1728515625, + "loss_wm_recon_w": 0.043212890625, + "step": 666 + }, + { + "epoch": 0.24860231084606785, + "grad_norm": 18.385643005371094, + "learning_rate": 9.152288878226991e-06, + "loss": 11.477355003356934, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.43985477089881897, + "loss_total_v6": 0.6271106004714966, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 667 + }, + { + "epoch": 0.24897502795378307, + "grad_norm": 19.09526252746582, + "learning_rate": 9.14876134419111e-06, + "loss": 11.16849422454834, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5702037215232849, + "loss_total_v6": 0.7735728621482849, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 668 + }, + { + "epoch": 0.24934774506149832, + "grad_norm": 18.164222717285156, + "learning_rate": 9.145227168677794e-06, + "loss": 11.285287857055664, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6680577397346497, + "loss_total_v6": 0.8702061772346497, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 669 + }, + { + "epoch": 0.24972046216921356, + "grad_norm": 17.998111724853516, + "learning_rate": 9.141686357344672e-06, + "loss": 11.024059295654297, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.23438745737075806, + "loss_total_v6": 0.42774683237075806, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 670 + }, + { + "epoch": 0.2500931792769288, + "grad_norm": 18.761333465576172, + "learning_rate": 9.138138915859997e-06, + "loss": 11.704107284545898, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7403715252876282, + "loss_total_v6": 0.9503324627876282, + "loss_wm_recon": 0.19140625, + "loss_wm_recon_w": 0.0478515625, + "step": 671 + }, + { + "epoch": 0.25046589638464406, + "grad_norm": 18.74468231201172, + "learning_rate": 9.134584849902632e-06, + "loss": 11.557519912719727, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6486681699752808, + "loss_total_v6": 0.8622912168502808, + "loss_wm_recon": 0.1982421875, + "loss_wm_recon_w": 0.049560546875, + "step": 672 + }, + { + "epoch": 0.2508386134923593, + "grad_norm": 17.987239837646484, + "learning_rate": 9.131024165162047e-06, + "loss": 11.484891891479492, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4648893475532532, + "loss_total_v6": 0.6702116131782532, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 673 + }, + { + "epoch": 0.25121133060007456, + "grad_norm": 20.251798629760742, + "learning_rate": 9.127456867338302e-06, + "loss": 11.241579055786133, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.26315024495124817, + "loss_total_v6": 0.46066001057624817, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 674 + }, + { + "epoch": 0.2515840477077898, + "grad_norm": 17.247026443481445, + "learning_rate": 9.123882962142051e-06, + "loss": 11.061201095581055, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6474798917770386, + "loss_total_v6": 0.8391302824020386, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 675 + }, + { + "epoch": 0.25195676481550505, + "grad_norm": 18.901369094848633, + "learning_rate": 9.120302455294526e-06, + "loss": 11.822590827941895, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.6096789240837097, + "loss_total_v6": 0.7974230647087097, + "loss_wm_recon": 0.1728515625, + "loss_wm_recon_w": 0.043212890625, + "step": 676 + }, + { + "epoch": 0.2523294819232203, + "grad_norm": 17.796144485473633, + "learning_rate": 9.11671535252752e-06, + "loss": 11.422616958618164, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6159847378730774, + "loss_total_v6": 0.6623714566230774, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 677 + }, + { + "epoch": 0.2527021990309355, + "grad_norm": 18.848194122314453, + "learning_rate": 9.113121659583385e-06, + "loss": 10.831405639648438, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.7093085050582886, + "loss_total_v6": 0.8946112394332886, + "loss_wm_recon": 0.1708984375, + "loss_wm_recon_w": 0.042724609375, + "step": 678 + }, + { + "epoch": 0.25307491613865074, + "grad_norm": 18.374614715576172, + "learning_rate": 9.10952138221503e-06, + "loss": 11.48575496673584, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5164483189582825, + "loss_total_v6": 0.7071221470832825, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 679 + }, + { + "epoch": 0.253447633246366, + "grad_norm": 18.3971004486084, + "learning_rate": 9.105914526185902e-06, + "loss": 10.93242359161377, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.1838282197713852, + "loss_total_v6": 0.3891505002975464, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 680 + }, + { + "epoch": 0.25382035035408124, + "grad_norm": 17.832801818847656, + "learning_rate": 9.102301097269974e-06, + "loss": 11.564461708068848, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.49542397260665894, + "loss_total_v6": 0.6863419413566589, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 681 + }, + { + "epoch": 0.2541930674617965, + "grad_norm": 17.378711700439453, + "learning_rate": 9.09868110125175e-06, + "loss": 11.294937133789062, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5329266786575317, + "loss_total_v6": 0.7301923036575317, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 682 + }, + { + "epoch": 0.25456578456951173, + "grad_norm": 17.473844528198242, + "learning_rate": 9.095054543926238e-06, + "loss": 11.208340644836426, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4181574881076813, + "loss_total_v6": 0.4596613943576813, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 683 + }, + { + "epoch": 0.254938501677227, + "grad_norm": 21.725582122802734, + "learning_rate": 9.091421431098955e-06, + "loss": 11.220587730407715, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.39171651005744934, + "loss_total_v6": 0.579948902130127, + "loss_wm_recon": 0.1669921875, + "loss_wm_recon_w": 0.041748046875, + "step": 684 + }, + { + "epoch": 0.25531121878494223, + "grad_norm": 17.599010467529297, + "learning_rate": 9.087781768585913e-06, + "loss": 11.298379898071289, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.618870735168457, + "loss_total_v6": 0.814671516418457, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 685 + }, + { + "epoch": 0.2556839358926575, + "grad_norm": 17.918336868286133, + "learning_rate": 9.084135562213603e-06, + "loss": 10.950932502746582, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4792872667312622, + "loss_total_v6": 0.6933985948562622, + "loss_wm_recon": 0.2080078125, + "loss_wm_recon_w": 0.052001953125, + "step": 686 + }, + { + "epoch": 0.2560566530003727, + "grad_norm": 18.892301559448242, + "learning_rate": 9.080482817818996e-06, + "loss": 11.150432586669922, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.35062772035598755, + "loss_total_v6": 0.5596120953559875, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 687 + }, + { + "epoch": 0.256429370108088, + "grad_norm": 17.78862190246582, + "learning_rate": 9.076823541249532e-06, + "loss": 10.841629028320312, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.647559404373169, + "loss_total_v6": 0.691504716873169, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 688 + }, + { + "epoch": 0.2568020872158032, + "grad_norm": 19.272703170776367, + "learning_rate": 9.073157738363101e-06, + "loss": 11.116943359375, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5180073976516724, + "loss_total_v6": 0.7238179445266724, + "loss_wm_recon": 0.1748046875, + "loss_wm_recon_w": 0.043701171875, + "step": 689 + }, + { + "epoch": 0.25717480432351847, + "grad_norm": 18.169754028320312, + "learning_rate": 9.069485415028047e-06, + "loss": 11.45927619934082, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4839436113834381, + "loss_total_v6": 0.6946369409561157, + "loss_wm_recon": 0.1943359375, + "loss_wm_recon_w": 0.048583984375, + "step": 690 + }, + { + "epoch": 0.2575475214312337, + "grad_norm": 19.260536193847656, + "learning_rate": 9.065806577123148e-06, + "loss": 11.584025382995605, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.34526655077934265, + "loss_total_v6": 0.537649393081665, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 691 + }, + { + "epoch": 0.25792023853894896, + "grad_norm": 17.779964447021484, + "learning_rate": 9.062121230537611e-06, + "loss": 11.308935165405273, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.6578603386878967, + "loss_total_v6": 0.8458486199378967, + "loss_wm_recon": 0.173828125, + "loss_wm_recon_w": 0.04345703125, + "step": 692 + }, + { + "epoch": 0.25829295564666416, + "grad_norm": 18.277921676635742, + "learning_rate": 9.058429381171066e-06, + "loss": 10.701242446899414, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.3681928217411041, + "loss_total_v6": 0.5542279481887817, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 693 + }, + { + "epoch": 0.2586656727543794, + "grad_norm": 18.187759399414062, + "learning_rate": 9.05473103493355e-06, + "loss": 11.40457534790039, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6317232847213745, + "loss_total_v6": 0.8397310972213745, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 694 + }, + { + "epoch": 0.25903838986209465, + "grad_norm": 19.112783432006836, + "learning_rate": 9.051026197745498e-06, + "loss": 11.407235145568848, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5911787152290344, + "loss_total_v6": 0.7967451214790344, + "loss_wm_recon": 0.189453125, + "loss_wm_recon_w": 0.04736328125, + "step": 695 + }, + { + "epoch": 0.2594111069698099, + "grad_norm": 17.477916717529297, + "learning_rate": 9.04731487553774e-06, + "loss": 10.966540336608887, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.8685978651046753, + "loss_total_v6": 1.0692814588546753, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 696 + }, + { + "epoch": 0.25978382407752515, + "grad_norm": 18.26980209350586, + "learning_rate": 9.043597074251487e-06, + "loss": 11.400192260742188, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.2228439599275589, + "loss_total_v6": 0.4093673825263977, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 697 + }, + { + "epoch": 0.2601565411852404, + "grad_norm": 18.948450088500977, + "learning_rate": 9.03987279983832e-06, + "loss": 11.219286918640137, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5676681399345398, + "loss_total_v6": 0.7578536868095398, + "loss_wm_recon": 0.1591796875, + "loss_wm_recon_w": 0.039794921875, + "step": 698 + }, + { + "epoch": 0.26052925829295565, + "grad_norm": 18.149646759033203, + "learning_rate": 9.036142058260183e-06, + "loss": 11.286869049072266, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4846668541431427, + "loss_total_v6": 0.6821765899658203, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 699 + }, + { + "epoch": 0.2609019754006709, + "grad_norm": 19.323596954345703, + "learning_rate": 9.032404855489375e-06, + "loss": 11.421483993530273, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5366638898849487, + "loss_total_v6": 0.7483338117599487, + "loss_wm_recon": 0.2294921875, + "loss_wm_recon_w": 0.057373046875, + "step": 700 + }, + { + "epoch": 0.2609019754006709, + "eval_loss": 0.6747047305107117, + "eval_loss_alignment": 0.2861891526442308, + "eval_loss_alignment_w": 0.1430945763221154, + "eval_loss_sft": 0.4900815882361852, + "eval_loss_total_v6": 0.6747047196214015, + "eval_loss_wm_recon": 0.16611422025240385, + "eval_loss_wm_recon_w": 0.041528555063100964, + "eval_runtime": 228.0091, + "eval_samples_per_second": 8.184, + "eval_steps_per_second": 1.026, + "step": 700 + }, + { + "epoch": 0.26127469250838614, + "grad_norm": 18.028766632080078, + "learning_rate": 9.028661197508536e-06, + "loss": 10.91467571258545, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5476023554801941, + "loss_total_v6": 0.7419382929801941, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 701 + }, + { + "epoch": 0.2616474096161014, + "grad_norm": 17.40701675415039, + "learning_rate": 9.024911090310638e-06, + "loss": 11.359811782836914, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.24418184161186218, + "loss_total_v6": 0.4326584041118622, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 702 + }, + { + "epoch": 0.26202012672381664, + "grad_norm": 18.468324661254883, + "learning_rate": 9.021154539898984e-06, + "loss": 10.72498893737793, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6963828802108765, + "loss_total_v6": 0.9090293645858765, + "loss_wm_recon": 0.1943359375, + "loss_wm_recon_w": 0.048583984375, + "step": 703 + }, + { + "epoch": 0.2623928438315319, + "grad_norm": 16.99452781677246, + "learning_rate": 9.017391552287182e-06, + "loss": 11.37222671508789, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7662792205810547, + "loss_total_v6": 0.9676952362060547, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 704 + }, + { + "epoch": 0.26276556093924713, + "grad_norm": 18.835844039916992, + "learning_rate": 9.013622133499153e-06, + "loss": 11.122583389282227, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5272092819213867, + "loss_total_v6": 0.7298460006713867, + "loss_wm_recon": 0.185546875, + "loss_wm_recon_w": 0.04638671875, + "step": 705 + }, + { + "epoch": 0.2631382780469624, + "grad_norm": 17.075796127319336, + "learning_rate": 9.009846289569108e-06, + "loss": 11.143514633178711, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5463293790817261, + "loss_total_v6": 0.7465246915817261, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 706 + }, + { + "epoch": 0.2635109951546776, + "grad_norm": 19.565462112426758, + "learning_rate": 9.006064026541549e-06, + "loss": 11.472612380981445, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4279286563396454, + "loss_total_v6": 0.6188466548919678, + "loss_wm_recon": 0.162109375, + "loss_wm_recon_w": 0.04052734375, + "step": 707 + }, + { + "epoch": 0.2638837122623928, + "grad_norm": 18.072755813598633, + "learning_rate": 9.002275350471244e-06, + "loss": 11.371152877807617, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5047443509101868, + "loss_total_v6": 0.6932209134101868, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 708 + }, + { + "epoch": 0.26425642937010807, + "grad_norm": 18.697420120239258, + "learning_rate": 8.99848026742324e-06, + "loss": 11.306083679199219, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5124074220657349, + "loss_total_v6": 0.7140675783157349, + "loss_wm_recon": 0.181640625, + "loss_wm_recon_w": 0.04541015625, + "step": 709 + }, + { + "epoch": 0.2646291464778233, + "grad_norm": 17.31546974182129, + "learning_rate": 8.994678783472834e-06, + "loss": 11.29623031616211, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.44449612498283386, + "loss_total_v6": 0.6329727172851562, + "loss_wm_recon": 0.16015625, + "loss_wm_recon_w": 0.0400390625, + "step": 710 + }, + { + "epoch": 0.26500186358553857, + "grad_norm": 17.807361602783203, + "learning_rate": 8.990870904705566e-06, + "loss": 11.106110572814941, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.42787274718284607, + "loss_total_v6": 0.6314860582351685, + "loss_wm_recon": 0.181640625, + "loss_wm_recon_w": 0.04541015625, + "step": 711 + }, + { + "epoch": 0.2653745806932538, + "grad_norm": 17.612207412719727, + "learning_rate": 8.98705663721722e-06, + "loss": 10.957186698913574, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.516775369644165, + "loss_total_v6": 0.703787088394165, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 712 + }, + { + "epoch": 0.26574729780096906, + "grad_norm": 19.31757354736328, + "learning_rate": 8.983235987113802e-06, + "loss": 10.901084899902344, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.40764057636260986, + "loss_total_v6": 0.5924550294876099, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 713 + }, + { + "epoch": 0.2661200149086843, + "grad_norm": 18.169437408447266, + "learning_rate": 8.979408960511541e-06, + "loss": 11.304801940917969, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6439356803894043, + "loss_total_v6": 0.8460841178894043, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 714 + }, + { + "epoch": 0.26649273201639956, + "grad_norm": 18.170530319213867, + "learning_rate": 8.97557556353687e-06, + "loss": 11.08355712890625, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.7422190308570862, + "loss_total_v6": 0.9219065308570862, + "loss_wm_recon": 0.15625, + "loss_wm_recon_w": 0.0390625, + "step": 715 + }, + { + "epoch": 0.2668654491241148, + "grad_norm": 17.936542510986328, + "learning_rate": 8.971735802326418e-06, + "loss": 11.10611343383789, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.32898738980293274, + "loss_total_v6": 0.5052568912506104, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 716 + }, + { + "epoch": 0.26723816623183005, + "grad_norm": 19.31932258605957, + "learning_rate": 8.967889683027009e-06, + "loss": 10.949352264404297, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3162432909011841, + "loss_total_v6": 0.5117999315261841, + "loss_wm_recon": 0.1728515625, + "loss_wm_recon_w": 0.043212890625, + "step": 717 + }, + { + "epoch": 0.2676108833395453, + "grad_norm": 19.03207015991211, + "learning_rate": 8.964037211795634e-06, + "loss": 11.632619857788086, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.49485692381858826, + "loss_total_v6": 0.7011557817459106, + "loss_wm_recon": 0.1845703125, + "loss_wm_recon_w": 0.046142578125, + "step": 718 + }, + { + "epoch": 0.26798360044726055, + "grad_norm": 18.16714096069336, + "learning_rate": 8.960178394799464e-06, + "loss": 11.522253036499023, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6334911584854126, + "loss_total_v6": 0.8414989709854126, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 719 + }, + { + "epoch": 0.2683563175549758, + "grad_norm": 17.99528694152832, + "learning_rate": 8.956313238215824e-06, + "loss": 10.911216735839844, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4294573962688446, + "loss_total_v6": 0.620131254196167, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 720 + }, + { + "epoch": 0.268729034662691, + "grad_norm": 20.037479400634766, + "learning_rate": 8.952441748232186e-06, + "loss": 11.17654800415039, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6998178362846375, + "loss_total_v6": 0.8970834612846375, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 721 + }, + { + "epoch": 0.26910175177040624, + "grad_norm": 17.605730056762695, + "learning_rate": 8.948563931046162e-06, + "loss": 11.134626388549805, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5078632831573486, + "loss_total_v6": 0.6919453144073486, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 722 + }, + { + "epoch": 0.2694744688781215, + "grad_norm": 18.473833084106445, + "learning_rate": 8.944679792865494e-06, + "loss": 11.351239204406738, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6162755489349365, + "loss_total_v6": 0.8174474239349365, + "loss_wm_recon": 0.1953125, + "loss_wm_recon_w": 0.048828125, + "step": 723 + }, + { + "epoch": 0.26984718598583673, + "grad_norm": 18.23761749267578, + "learning_rate": 8.940789339908041e-06, + "loss": 10.776121139526367, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.3999849259853363, + "loss_total_v6": 0.5984712839126587, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 724 + }, + { + "epoch": 0.270219903093552, + "grad_norm": 17.81463623046875, + "learning_rate": 8.93689257840177e-06, + "loss": 10.924766540527344, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.26498904824256897, + "loss_total_v6": 0.44955936074256897, + "loss_wm_recon": 0.15234375, + "loss_wm_recon_w": 0.0380859375, + "step": 725 + }, + { + "epoch": 0.27059262020126723, + "grad_norm": 18.22230339050293, + "learning_rate": 8.93298951458475e-06, + "loss": 11.614791870117188, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7469919323921204, + "loss_total_v6": 0.9491403698921204, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 726 + }, + { + "epoch": 0.2709653373089825, + "grad_norm": 17.943748474121094, + "learning_rate": 8.92908015470514e-06, + "loss": 11.492210388183594, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.558927595615387, + "loss_total_v6": 0.773527204990387, + "loss_wm_recon": 0.1943359375, + "loss_wm_recon_w": 0.048583984375, + "step": 727 + }, + { + "epoch": 0.2713380544166977, + "grad_norm": 17.75202751159668, + "learning_rate": 8.92516450502117e-06, + "loss": 11.517831802368164, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.27949920296669006, + "loss_total_v6": 0.46773162484169006, + "loss_wm_recon": 0.1591796875, + "loss_wm_recon_w": 0.039794921875, + "step": 728 + }, + { + "epoch": 0.271710771524413, + "grad_norm": 17.378917694091797, + "learning_rate": 8.92124257180115e-06, + "loss": 10.958306312561035, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.2488236129283905, + "loss_total_v6": 0.4485306441783905, + "loss_wm_recon": 0.173828125, + "loss_wm_recon_w": 0.04345703125, + "step": 729 + }, + { + "epoch": 0.2720834886321282, + "grad_norm": 18.525741577148438, + "learning_rate": 8.917314361323439e-06, + "loss": 11.096778869628906, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.20434774458408356, + "loss_total_v6": 0.40698444843292236, + "loss_wm_recon": 0.177734375, + "loss_wm_recon_w": 0.04443359375, + "step": 730 + }, + { + "epoch": 0.27245620573984347, + "grad_norm": 17.379222869873047, + "learning_rate": 8.913379879876451e-06, + "loss": 11.039932250976562, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5657222270965576, + "loss_total_v6": 0.7559077739715576, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 731 + }, + { + "epoch": 0.2728289228475587, + "grad_norm": 17.513072967529297, + "learning_rate": 8.909439133758637e-06, + "loss": 11.011167526245117, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5200347900390625, + "loss_total_v6": 0.7241363525390625, + "loss_wm_recon": 0.18359375, + "loss_wm_recon_w": 0.0458984375, + "step": 732 + }, + { + "epoch": 0.27320163995527397, + "grad_norm": 17.6724910736084, + "learning_rate": 8.905492129278478e-06, + "loss": 11.135841369628906, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5004266500473022, + "loss_total_v6": 0.7098993062973022, + "loss_wm_recon": 0.189453125, + "loss_wm_recon_w": 0.04736328125, + "step": 733 + }, + { + "epoch": 0.2735743570629892, + "grad_norm": 19.056699752807617, + "learning_rate": 8.90153887275447e-06, + "loss": 11.028495788574219, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.518989622592926, + "loss_total_v6": 0.713813841342926, + "loss_wm_recon": 0.169921875, + "loss_wm_recon_w": 0.04248046875, + "step": 734 + }, + { + "epoch": 0.2739470741707044, + "grad_norm": 18.50543212890625, + "learning_rate": 8.897579370515119e-06, + "loss": 11.011411666870117, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5809550285339355, + "loss_total_v6": 0.7713847160339355, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 735 + }, + { + "epoch": 0.27431979127841966, + "grad_norm": 18.04804039001465, + "learning_rate": 8.893613628898934e-06, + "loss": 10.992257118225098, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.579936683177948, + "loss_total_v6": 0.776225745677948, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 736 + }, + { + "epoch": 0.2746925083861349, + "grad_norm": 17.20016098022461, + "learning_rate": 8.889641654254404e-06, + "loss": 10.733926773071289, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.46376967430114746, + "loss_total_v6": 0.6634767055511475, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 737 + }, + { + "epoch": 0.27506522549385015, + "grad_norm": 18.247817993164062, + "learning_rate": 8.88566345294e-06, + "loss": 10.918272972106934, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6253045797348022, + "loss_total_v6": 0.8267205953598022, + "loss_wm_recon": 0.1884765625, + "loss_wm_recon_w": 0.047119140625, + "step": 738 + }, + { + "epoch": 0.2754379426015654, + "grad_norm": 17.407649993896484, + "learning_rate": 8.881679031324167e-06, + "loss": 11.22254467010498, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.16979429125785828, + "loss_total_v6": 0.3528997600078583, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 739 + }, + { + "epoch": 0.27581065970928065, + "grad_norm": 18.6637020111084, + "learning_rate": 8.877688395785296e-06, + "loss": 10.890960693359375, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6428777575492859, + "loss_total_v6": 0.8479558825492859, + "loss_wm_recon": 0.1953125, + "loss_wm_recon_w": 0.048828125, + "step": 740 + }, + { + "epoch": 0.2761833768169959, + "grad_norm": 17.238672256469727, + "learning_rate": 8.873691552711734e-06, + "loss": 11.426185607910156, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2225385457277298, + "loss_total_v6": 0.4105268120765686, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 741 + }, + { + "epoch": 0.27655609392471114, + "grad_norm": 19.27814292907715, + "learning_rate": 8.869688508501762e-06, + "loss": 11.38132381439209, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5128265619277954, + "loss_total_v6": 0.7100921869277954, + "loss_wm_recon": 0.15625, + "loss_wm_recon_w": 0.0390625, + "step": 742 + }, + { + "epoch": 0.2769288110324264, + "grad_norm": 18.374174118041992, + "learning_rate": 8.865679269563588e-06, + "loss": 11.347251892089844, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.47759509086608887, + "loss_total_v6": 0.6636302471160889, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 743 + }, + { + "epoch": 0.27730152814014164, + "grad_norm": 17.875450134277344, + "learning_rate": 8.861663842315339e-06, + "loss": 11.167024612426758, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.9026228189468384, + "loss_total_v6": 1.1015974283218384, + "loss_wm_recon": 0.1787109375, + "loss_wm_recon_w": 0.044677734375, + "step": 744 + }, + { + "epoch": 0.2776742452478569, + "grad_norm": 17.432846069335938, + "learning_rate": 8.857642233185043e-06, + "loss": 11.02779769897461, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6040837168693542, + "loss_total_v6": 0.7881657481193542, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 745 + }, + { + "epoch": 0.27804696235557214, + "grad_norm": 18.778905868530273, + "learning_rate": 8.85361444861063e-06, + "loss": 11.167202949523926, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3192097842693329, + "loss_total_v6": 0.5091512203216553, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 746 + }, + { + "epoch": 0.2784196794632874, + "grad_norm": 18.24215316772461, + "learning_rate": 8.849580495039916e-06, + "loss": 11.2423095703125, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5880089402198792, + "loss_total_v6": 0.7847862839698792, + "loss_wm_recon": 0.162109375, + "loss_wm_recon_w": 0.04052734375, + "step": 747 + }, + { + "epoch": 0.27879239657100263, + "grad_norm": 19.022266387939453, + "learning_rate": 8.845540378930589e-06, + "loss": 11.454193115234375, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5432534217834473, + "loss_total_v6": 0.7527260780334473, + "loss_wm_recon": 0.181640625, + "loss_wm_recon_w": 0.04541015625, + "step": 748 + }, + { + "epoch": 0.2791651136787179, + "grad_norm": 16.82155418395996, + "learning_rate": 8.8414941067502e-06, + "loss": 10.94529914855957, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.5042112469673157, + "loss_total_v6": 0.6863401532173157, + "loss_wm_recon": 0.173828125, + "loss_wm_recon_w": 0.04345703125, + "step": 749 + }, + { + "epoch": 0.2795378307864331, + "grad_norm": 17.86532974243164, + "learning_rate": 8.837441684976163e-06, + "loss": 11.405227661132812, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.46543654799461365, + "loss_total_v6": 0.669049859046936, + "loss_wm_recon": 0.173828125, + "loss_wm_recon_w": 0.04345703125, + "step": 750 + }, + { + "epoch": 0.2799105478941483, + "grad_norm": 17.836435317993164, + "learning_rate": 8.833383120095729e-06, + "loss": 10.485034942626953, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.36988455057144165, + "loss_total_v6": 0.5734978318214417, + "loss_wm_recon": 0.173828125, + "loss_wm_recon_w": 0.04345703125, + "step": 751 + }, + { + "epoch": 0.28028326500186357, + "grad_norm": 18.854516983032227, + "learning_rate": 8.829318418605986e-06, + "loss": 11.229129791259766, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3905474543571472, + "loss_total_v6": 0.5846392512321472, + "loss_wm_recon": 0.1591796875, + "loss_wm_recon_w": 0.039794921875, + "step": 752 + }, + { + "epoch": 0.2806559821095788, + "grad_norm": 17.993497848510742, + "learning_rate": 8.825247587013846e-06, + "loss": 11.31869888305664, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.755892813205719, + "loss_total_v6": 0.973910391330719, + "loss_wm_recon": 0.1923828125, + "loss_wm_recon_w": 0.048095703125, + "step": 753 + }, + { + "epoch": 0.28102869921729406, + "grad_norm": 18.112396240234375, + "learning_rate": 8.821170631836035e-06, + "loss": 11.133394241333008, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.706168532371521, + "loss_total_v6": 0.904899001121521, + "loss_wm_recon": 0.169921875, + "loss_wm_recon_w": 0.04248046875, + "step": 754 + }, + { + "epoch": 0.2814014163250093, + "grad_norm": 18.001869201660156, + "learning_rate": 8.81708755959908e-06, + "loss": 11.331931114196777, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.4537099599838257, + "loss_total_v6": 0.6675771474838257, + "loss_wm_recon": 0.19140625, + "loss_wm_recon_w": 0.0478515625, + "step": 755 + }, + { + "epoch": 0.28177413343272456, + "grad_norm": 18.061811447143555, + "learning_rate": 8.8129983768393e-06, + "loss": 10.730531692504883, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4945855438709259, + "loss_total_v6": 0.6864800453186035, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 756 + }, + { + "epoch": 0.2821468505404398, + "grad_norm": 18.895282745361328, + "learning_rate": 8.808903090102799e-06, + "loss": 11.240121841430664, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3243245780467987, + "loss_total_v6": 0.5193929672241211, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 757 + }, + { + "epoch": 0.28251956764815506, + "grad_norm": 16.516151428222656, + "learning_rate": 8.804801705945448e-06, + "loss": 11.12568473815918, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.2651827037334442, + "loss_total_v6": 0.4582979381084442, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 758 + }, + { + "epoch": 0.2828922847558703, + "grad_norm": 17.444995880126953, + "learning_rate": 8.800694230932885e-06, + "loss": 11.066231727600098, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6472204327583313, + "loss_total_v6": 0.8422887921333313, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 759 + }, + { + "epoch": 0.28326500186358555, + "grad_norm": 17.511075973510742, + "learning_rate": 8.79658067164049e-06, + "loss": 11.445018768310547, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5165337324142456, + "loss_total_v6": 0.7086724042892456, + "loss_wm_recon": 0.1669921875, + "loss_wm_recon_w": 0.041748046875, + "step": 760 + }, + { + "epoch": 0.2836377189713008, + "grad_norm": 18.320993423461914, + "learning_rate": 8.792461034653393e-06, + "loss": 11.45813274383545, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6553735733032227, + "loss_total_v6": 0.8592309951782227, + "loss_wm_recon": 0.1748046875, + "loss_wm_recon_w": 0.043701171875, + "step": 761 + }, + { + "epoch": 0.28401043607901605, + "grad_norm": 18.064517974853516, + "learning_rate": 8.788335326566444e-06, + "loss": 11.132150650024414, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.46596765518188477, + "loss_total_v6": 0.6639657020568848, + "loss_wm_recon": 0.1748046875, + "loss_wm_recon_w": 0.043701171875, + "step": 762 + }, + { + "epoch": 0.2843831531867313, + "grad_norm": 18.98895263671875, + "learning_rate": 8.784203553984216e-06, + "loss": 11.013301849365234, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.31337833404541016, + "loss_total_v6": 0.5028314590454102, + "loss_wm_recon": 0.15625, + "loss_wm_recon_w": 0.0390625, + "step": 763 + }, + { + "epoch": 0.2847558702944465, + "grad_norm": 18.275121688842773, + "learning_rate": 8.780065723520995e-06, + "loss": 10.983911514282227, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.53657466173172, + "loss_total_v6": 0.72431880235672, + "loss_wm_recon": 0.1572265625, + "loss_wm_recon_w": 0.039306640625, + "step": 764 + }, + { + "epoch": 0.28512858740216174, + "grad_norm": 18.81539535522461, + "learning_rate": 8.775921841800756e-06, + "loss": 11.619482040405273, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4406852126121521, + "loss_total_v6": 0.6294059157371521, + "loss_wm_recon": 0.1611328125, + "loss_wm_recon_w": 0.040283203125, + "step": 765 + }, + { + "epoch": 0.285501304509877, + "grad_norm": 18.1484432220459, + "learning_rate": 8.771771915457165e-06, + "loss": 11.003780364990234, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.7076355218887329, + "loss_total_v6": 0.9056335687637329, + "loss_wm_recon": 0.1669921875, + "loss_wm_recon_w": 0.041748046875, + "step": 766 + }, + { + "epoch": 0.28587402161759223, + "grad_norm": 17.892507553100586, + "learning_rate": 8.767615951133566e-06, + "loss": 11.257993698120117, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6220554113388062, + "loss_total_v6": 0.8251804113388062, + "loss_wm_recon": 0.1796875, + "loss_wm_recon_w": 0.044921875, + "step": 767 + }, + { + "epoch": 0.2862467387253075, + "grad_norm": 18.56181526184082, + "learning_rate": 8.763453955482966e-06, + "loss": 11.43193244934082, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.45241421461105347, + "loss_total_v6": 0.6272189021110535, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 768 + }, + { + "epoch": 0.28661945583302273, + "grad_norm": 17.570053100585938, + "learning_rate": 8.75928593516803e-06, + "loss": 11.463973999023438, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.24542459845542908, + "loss_total_v6": 0.4517234265804291, + "loss_wm_recon": 0.1845703125, + "loss_wm_recon_w": 0.046142578125, + "step": 769 + }, + { + "epoch": 0.286992172940738, + "grad_norm": 18.17591667175293, + "learning_rate": 8.755111896861062e-06, + "loss": 11.189615249633789, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.47069448232650757, + "loss_total_v6": 0.6581944823265076, + "loss_wm_recon": 0.15625, + "loss_wm_recon_w": 0.0390625, + "step": 770 + }, + { + "epoch": 0.2873648900484532, + "grad_norm": 18.374544143676758, + "learning_rate": 8.750931847244007e-06, + "loss": 10.689035415649414, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5022376775741577, + "loss_total_v6": 0.6899818181991577, + "loss_wm_recon": 0.1572265625, + "loss_wm_recon_w": 0.039306640625, + "step": 771 + }, + { + "epoch": 0.2877376071561685, + "grad_norm": 19.08561897277832, + "learning_rate": 8.74674579300843e-06, + "loss": 11.263944625854492, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6788980960845947, + "loss_total_v6": 0.8746988773345947, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 772 + }, + { + "epoch": 0.2881103242638837, + "grad_norm": 18.346879959106445, + "learning_rate": 8.742553740855507e-06, + "loss": 11.37938117980957, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.8165284395217896, + "loss_total_v6": 1.0396729707717896, + "loss_wm_recon": 0.197265625, + "loss_wm_recon_w": 0.04931640625, + "step": 773 + }, + { + "epoch": 0.28848304137159897, + "grad_norm": 18.036943435668945, + "learning_rate": 8.738355697496015e-06, + "loss": 10.999689102172852, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.491664320230484, + "loss_total_v6": 0.6879533529281616, + "loss_wm_recon": 0.16015625, + "loss_wm_recon_w": 0.0400390625, + "step": 774 + }, + { + "epoch": 0.2888557584793142, + "grad_norm": 18.66415786743164, + "learning_rate": 8.734151669650327e-06, + "loss": 11.204508781433105, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.6370150446891785, + "loss_total_v6": 0.8503939509391785, + "loss_wm_recon": 0.181640625, + "loss_wm_recon_w": 0.04541015625, + "step": 775 + }, + { + "epoch": 0.28922847558702947, + "grad_norm": 17.439289093017578, + "learning_rate": 8.729941664048394e-06, + "loss": 11.407195091247559, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.590977132320404, + "loss_total_v6": 0.775059163570404, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 776 + }, + { + "epoch": 0.2896011926947447, + "grad_norm": 16.998212814331055, + "learning_rate": 8.72572568742973e-06, + "loss": 11.44969367980957, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5541064143180847, + "loss_total_v6": 0.7442919611930847, + "loss_wm_recon": 0.1669921875, + "loss_wm_recon_w": 0.041748046875, + "step": 777 + }, + { + "epoch": 0.2899739098024599, + "grad_norm": 17.444137573242188, + "learning_rate": 8.721503746543417e-06, + "loss": 10.881900787353516, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.48890265822410583, + "loss_total_v6": 0.6900745630264282, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 778 + }, + { + "epoch": 0.29034662691017515, + "grad_norm": 18.163785934448242, + "learning_rate": 8.717275848148078e-06, + "loss": 10.818928718566895, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5572531819343567, + "loss_total_v6": 0.7430441975593567, + "loss_wm_recon": 0.1572265625, + "loss_wm_recon_w": 0.039306640625, + "step": 779 + }, + { + "epoch": 0.2907193440178904, + "grad_norm": 17.49033546447754, + "learning_rate": 8.71304199901188e-06, + "loss": 10.76034164428711, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.6015605330467224, + "loss_total_v6": 0.7832011580467224, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 780 + }, + { + "epoch": 0.29109206112560565, + "grad_norm": 18.198156356811523, + "learning_rate": 8.708802205912508e-06, + "loss": 11.151517868041992, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.2570493519306183, + "loss_total_v6": 0.4386899769306183, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 781 + }, + { + "epoch": 0.2914647782333209, + "grad_norm": 17.020700454711914, + "learning_rate": 8.704556475637166e-06, + "loss": 11.335586547851562, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6582809090614319, + "loss_total_v6": 0.8523727059364319, + "loss_wm_recon": 0.1591796875, + "loss_wm_recon_w": 0.039794921875, + "step": 782 + }, + { + "epoch": 0.29183749534103615, + "grad_norm": 17.420116424560547, + "learning_rate": 8.700304814982565e-06, + "loss": 11.35561752319336, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5643653869628906, + "loss_total_v6": 0.7599220275878906, + "loss_wm_recon": 0.1728515625, + "loss_wm_recon_w": 0.043212890625, + "step": 783 + }, + { + "epoch": 0.2922102124487514, + "grad_norm": 17.810178756713867, + "learning_rate": 8.696047230754905e-06, + "loss": 11.12025260925293, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5468472242355347, + "loss_total_v6": 0.7292202711105347, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 784 + }, + { + "epoch": 0.29258292955646664, + "grad_norm": 18.741167068481445, + "learning_rate": 8.691783729769874e-06, + "loss": 10.87432861328125, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7000793814659119, + "loss_total_v6": 0.8951477408409119, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 785 + }, + { + "epoch": 0.2929556466641819, + "grad_norm": 17.96908187866211, + "learning_rate": 8.687514318852623e-06, + "loss": 11.157781600952148, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4291881024837494, + "loss_total_v6": 0.6203502416610718, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 786 + }, + { + "epoch": 0.29332836377189714, + "grad_norm": 18.761444091796875, + "learning_rate": 8.683239004837776e-06, + "loss": 11.042768478393555, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.6708563566207886, + "loss_total_v6": 0.8896063566207886, + "loss_wm_recon": 0.1953125, + "loss_wm_recon_w": 0.048828125, + "step": 787 + }, + { + "epoch": 0.2937010808796124, + "grad_norm": 18.86800193786621, + "learning_rate": 8.678957794569395e-06, + "loss": 10.872343063354492, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.48149481415748596, + "loss_total_v6": 0.6726568937301636, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 788 + }, + { + "epoch": 0.29407379798732763, + "grad_norm": 17.920000076293945, + "learning_rate": 8.674670694900988e-06, + "loss": 10.780218124389648, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.5657777786254883, + "loss_total_v6": 0.7857484817504883, + "loss_wm_recon": 0.1845703125, + "loss_wm_recon_w": 0.046142578125, + "step": 789 + }, + { + "epoch": 0.2944465150950429, + "grad_norm": 17.95012664794922, + "learning_rate": 8.67037771269549e-06, + "loss": 11.339173316955566, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.524253249168396, + "loss_total_v6": 0.715171217918396, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 790 + }, + { + "epoch": 0.29481923220275813, + "grad_norm": 19.181379318237305, + "learning_rate": 8.66607885482525e-06, + "loss": 10.988399505615234, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.499503493309021, + "loss_total_v6": 0.692862868309021, + "loss_wm_recon": 0.15625, + "loss_wm_recon_w": 0.0390625, + "step": 791 + }, + { + "epoch": 0.2951919493104734, + "grad_norm": 17.986711502075195, + "learning_rate": 8.661774128172027e-06, + "loss": 10.887916564941406, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4681573510169983, + "loss_total_v6": 0.6478448510169983, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 792 + }, + { + "epoch": 0.29556466641818857, + "grad_norm": 17.039806365966797, + "learning_rate": 8.657463539626971e-06, + "loss": 10.751676559448242, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.517224133014679, + "loss_total_v6": 0.716198742389679, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 793 + }, + { + "epoch": 0.2959373835259038, + "grad_norm": 17.764480590820312, + "learning_rate": 8.65314709609062e-06, + "loss": 10.930370330810547, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.16713634133338928, + "loss_total_v6": 0.3551246225833893, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 794 + }, + { + "epoch": 0.29631010063361907, + "grad_norm": 17.15452766418457, + "learning_rate": 8.648824804472882e-06, + "loss": 10.741781234741211, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6064562797546387, + "loss_total_v6": 0.8134875297546387, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 795 + }, + { + "epoch": 0.2966828177413343, + "grad_norm": 18.18975257873535, + "learning_rate": 8.644496671693028e-06, + "loss": 10.766834259033203, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.3708752393722534, + "loss_total_v6": 0.5686291456222534, + "loss_wm_recon": 0.166015625, + "loss_wm_recon_w": 0.04150390625, + "step": 796 + }, + { + "epoch": 0.29705553484904956, + "grad_norm": 17.130380630493164, + "learning_rate": 8.640162704679677e-06, + "loss": 11.439313888549805, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.512315034866333, + "loss_total_v6": 0.698350191116333, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 797 + }, + { + "epoch": 0.2974282519567648, + "grad_norm": 18.7088623046875, + "learning_rate": 8.635822910370793e-06, + "loss": 11.410015106201172, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5701771378517151, + "loss_total_v6": 0.7571888566017151, + "loss_wm_recon": 0.169921875, + "loss_wm_recon_w": 0.04248046875, + "step": 798 + }, + { + "epoch": 0.29780096906448006, + "grad_norm": 17.740633010864258, + "learning_rate": 8.631477295713663e-06, + "loss": 10.811182975769043, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.579584002494812, + "loss_total_v6": 0.780023455619812, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 799 + }, + { + "epoch": 0.2981736861721953, + "grad_norm": 17.862728118896484, + "learning_rate": 8.627125867664895e-06, + "loss": 11.061519622802734, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.41315895318984985, + "loss_total_v6": 0.6060300469398499, + "loss_wm_recon": 0.146484375, + "loss_wm_recon_w": 0.03662109375, + "step": 800 + }, + { + "epoch": 0.2981736861721953, + "eval_loss": 0.6659027934074402, + "eval_loss_alignment": 0.28511868990384615, + "eval_loss_alignment_w": 0.14255934495192307, + "eval_loss_sft": 0.4858297789628562, + "eval_loss_total_v6": 0.6659027922293569, + "eval_loss_wm_recon": 0.15005467080662394, + "eval_loss_wm_recon_w": 0.037513667701655984, + "eval_runtime": 227.4044, + "eval_samples_per_second": 8.206, + "eval_steps_per_second": 1.029, + "step": 800 + }, + { + "epoch": 0.29854640327991055, + "grad_norm": 16.79267692565918, + "learning_rate": 8.622768633190404e-06, + "loss": 10.923697471618652, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6370656490325928, + "loss_total_v6": 0.8221242427825928, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 801 + }, + { + "epoch": 0.2989191203876258, + "grad_norm": 17.21137237548828, + "learning_rate": 8.618405599265394e-06, + "loss": 10.810615539550781, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5049671530723572, + "loss_total_v6": 0.6873401999473572, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 802 + }, + { + "epoch": 0.29929183749534105, + "grad_norm": 17.46150779724121, + "learning_rate": 8.61403677287436e-06, + "loss": 11.057168006896973, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6680131554603577, + "loss_total_v6": 0.8567338585853577, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 803 + }, + { + "epoch": 0.2996645546030563, + "grad_norm": 17.459726333618164, + "learning_rate": 8.609662161011068e-06, + "loss": 10.888982772827148, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.7369223237037659, + "loss_total_v6": 0.9380941987037659, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 804 + }, + { + "epoch": 0.30003727171077155, + "grad_norm": 17.38762855529785, + "learning_rate": 8.605281770678541e-06, + "loss": 11.046594619750977, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6165145039558411, + "loss_total_v6": 0.8076766133308411, + "loss_wm_recon": 0.1552734375, + "loss_wm_recon_w": 0.038818359375, + "step": 805 + }, + { + "epoch": 0.3004099888184868, + "grad_norm": 17.935680389404297, + "learning_rate": 8.600895608889059e-06, + "loss": 11.294960021972656, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5085169672966003, + "loss_total_v6": 0.6999232172966003, + "loss_wm_recon": 0.15625, + "loss_wm_recon_w": 0.0390625, + "step": 806 + }, + { + "epoch": 0.300782705926202, + "grad_norm": 18.34597396850586, + "learning_rate": 8.596503682664138e-06, + "loss": 10.995735168457031, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.4320051670074463, + "loss_total_v6": 0.6402571201324463, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 807 + }, + { + "epoch": 0.30115542303391724, + "grad_norm": 18.879138946533203, + "learning_rate": 8.59210599903452e-06, + "loss": 11.183565139770508, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.34411129355430603, + "loss_total_v6": 0.5296581983566284, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 808 + }, + { + "epoch": 0.3015281401416325, + "grad_norm": 17.804588317871094, + "learning_rate": 8.587702565040168e-06, + "loss": 11.1253023147583, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.8702419996261597, + "loss_total_v6": 1.0606716871261597, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 809 + }, + { + "epoch": 0.30190085724934773, + "grad_norm": 16.65618324279785, + "learning_rate": 8.583293387730245e-06, + "loss": 11.073122024536133, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.690631091594696, + "loss_total_v6": 0.887408435344696, + "loss_wm_recon": 0.162109375, + "loss_wm_recon_w": 0.04052734375, + "step": 810 + }, + { + "epoch": 0.302273574357063, + "grad_norm": 17.59060287475586, + "learning_rate": 8.578878474163115e-06, + "loss": 10.945995330810547, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2385185807943344, + "loss_total_v6": 0.4321221113204956, + "loss_wm_recon": 0.1728515625, + "loss_wm_recon_w": 0.043212890625, + "step": 811 + }, + { + "epoch": 0.3026462914647782, + "grad_norm": 18.4986572265625, + "learning_rate": 8.574457831406317e-06, + "loss": 10.954998016357422, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6376509070396423, + "loss_total_v6": 0.8202680945396423, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 812 + }, + { + "epoch": 0.3030190085724935, + "grad_norm": 16.922683715820312, + "learning_rate": 8.57003146653657e-06, + "loss": 10.943249702453613, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.267297625541687, + "loss_total_v6": 0.457971453666687, + "loss_wm_recon": 0.1611328125, + "loss_wm_recon_w": 0.040283203125, + "step": 813 + }, + { + "epoch": 0.3033917256802087, + "grad_norm": 17.98798942565918, + "learning_rate": 8.565599386639744e-06, + "loss": 10.872267723083496, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.6764183640480042, + "loss_total_v6": 0.8861351609230042, + "loss_wm_recon": 0.1748046875, + "loss_wm_recon_w": 0.043701171875, + "step": 814 + }, + { + "epoch": 0.30376444278792397, + "grad_norm": 17.314346313476562, + "learning_rate": 8.561161598810864e-06, + "loss": 10.532357215881348, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4185904562473297, + "loss_total_v6": 0.6146353483200073, + "loss_wm_recon": 0.1591796875, + "loss_wm_recon_w": 0.039794921875, + "step": 815 + }, + { + "epoch": 0.3041371598956392, + "grad_norm": 17.444852828979492, + "learning_rate": 8.55671811015409e-06, + "loss": 11.264108657836914, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.9412499666213989, + "loss_total_v6": 1.132900357246399, + "loss_wm_recon": 0.2041015625, + "loss_wm_recon_w": 0.051025390625, + "step": 816 + }, + { + "epoch": 0.30450987700335447, + "grad_norm": 18.756942749023438, + "learning_rate": 8.552268927782713e-06, + "loss": 10.853053092956543, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5178345441818237, + "loss_total_v6": 0.7292603254318237, + "loss_wm_recon": 0.189453125, + "loss_wm_recon_w": 0.04736328125, + "step": 817 + }, + { + "epoch": 0.3048825941110697, + "grad_norm": 18.73107147216797, + "learning_rate": 8.547814058819128e-06, + "loss": 10.777295112609863, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.811385989189148, + "loss_total_v6": 1.006454348564148, + "loss_wm_recon": 0.1552734375, + "loss_wm_recon_w": 0.038818359375, + "step": 818 + }, + { + "epoch": 0.30525531121878496, + "grad_norm": 18.14306640625, + "learning_rate": 8.543353510394846e-06, + "loss": 11.23138427734375, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.25694283843040466, + "loss_total_v6": 0.43882760405540466, + "loss_wm_recon": 0.1416015625, + "loss_wm_recon_w": 0.035400390625, + "step": 819 + }, + { + "epoch": 0.3056280283265002, + "grad_norm": 18.447202682495117, + "learning_rate": 8.53888728965046e-06, + "loss": 11.472843170166016, + "loss_alignment": 0.26953125, + "loss_alignment_w": 0.134765625, + "loss_sft": 0.605525016784668, + "loss_total_v6": 0.781550407409668, + "loss_wm_recon": 0.1650390625, + "loss_wm_recon_w": 0.041259765625, + "step": 820 + }, + { + "epoch": 0.3060007454342154, + "grad_norm": 17.948583602905273, + "learning_rate": 8.534415403735647e-06, + "loss": 11.148838996887207, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6490423083305359, + "loss_total_v6": 0.8333684802055359, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 821 + }, + { + "epoch": 0.30637346254193065, + "grad_norm": 17.653013229370117, + "learning_rate": 8.529937859809156e-06, + "loss": 11.323156356811523, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6537446975708008, + "loss_total_v6": 0.8412446975708008, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 822 + }, + { + "epoch": 0.3067461796496459, + "grad_norm": 17.122854232788086, + "learning_rate": 8.525454665038788e-06, + "loss": 10.447013854980469, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.623266339302063, + "loss_total_v6": 0.820287823677063, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 823 + }, + { + "epoch": 0.30711889675736115, + "grad_norm": 17.223718643188477, + "learning_rate": 8.520965826601394e-06, + "loss": 11.005620002746582, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5514560341835022, + "loss_total_v6": 0.7499423623085022, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 824 + }, + { + "epoch": 0.3074916138650764, + "grad_norm": 18.056371688842773, + "learning_rate": 8.516471351682858e-06, + "loss": 11.169576644897461, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.7019863128662109, + "loss_total_v6": 0.9021816253662109, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 825 + }, + { + "epoch": 0.30786433097279164, + "grad_norm": 17.055187225341797, + "learning_rate": 8.511971247478089e-06, + "loss": 10.92779541015625, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.7100106477737427, + "loss_total_v6": 0.9170418977737427, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 826 + }, + { + "epoch": 0.3082370480805069, + "grad_norm": 17.8328914642334, + "learning_rate": 8.507465521191003e-06, + "loss": 11.003281593322754, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.4332487881183624, + "loss_total_v6": 0.64394211769104, + "loss_wm_recon": 0.1708984375, + "loss_wm_recon_w": 0.042724609375, + "step": 827 + }, + { + "epoch": 0.30860976518822214, + "grad_norm": 18.424480438232422, + "learning_rate": 8.502954180034519e-06, + "loss": 11.076530456542969, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4341293275356293, + "loss_total_v6": 0.6326156854629517, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 828 + }, + { + "epoch": 0.3089824822959374, + "grad_norm": 18.16493034362793, + "learning_rate": 8.49843723123055e-06, + "loss": 10.897007942199707, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.3660478889942169, + "loss_total_v6": 0.5325517654418945, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 829 + }, + { + "epoch": 0.30935519940365264, + "grad_norm": 17.18381690979004, + "learning_rate": 8.493914682009974e-06, + "loss": 11.00456714630127, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7855401039123535, + "loss_total_v6": 0.9869561195373535, + "loss_wm_recon": 0.1572265625, + "loss_wm_recon_w": 0.039306640625, + "step": 830 + }, + { + "epoch": 0.3097279165113679, + "grad_norm": 17.61945343017578, + "learning_rate": 8.489386539612643e-06, + "loss": 11.045431137084961, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.36327654123306274, + "loss_total_v6": 0.5512648224830627, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 831 + }, + { + "epoch": 0.31010063361908313, + "grad_norm": 17.16276741027832, + "learning_rate": 8.48485281128736e-06, + "loss": 10.787399291992188, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5968448519706726, + "loss_total_v6": 0.7782413363456726, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 832 + }, + { + "epoch": 0.3104733507267984, + "grad_norm": 17.81385040283203, + "learning_rate": 8.480313504291871e-06, + "loss": 11.329097747802734, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.3960227370262146, + "loss_total_v6": 0.5906028151512146, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 833 + }, + { + "epoch": 0.31084606783451363, + "grad_norm": 18.208402633666992, + "learning_rate": 8.475768625892853e-06, + "loss": 11.088558197021484, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.7198526859283447, + "loss_total_v6": 0.9014933109283447, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 834 + }, + { + "epoch": 0.3112187849422288, + "grad_norm": 17.375812530517578, + "learning_rate": 8.471218183365901e-06, + "loss": 10.534139633178711, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.6234918236732483, + "loss_total_v6": 0.7951226830482483, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 835 + }, + { + "epoch": 0.31159150204994407, + "grad_norm": 17.398752212524414, + "learning_rate": 8.466662183995515e-06, + "loss": 10.460285186767578, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.26777270436286926, + "loss_total_v6": 0.46259692311286926, + "loss_wm_recon": 0.162109375, + "loss_wm_recon_w": 0.04052734375, + "step": 836 + }, + { + "epoch": 0.3119642191576593, + "grad_norm": 17.8947696685791, + "learning_rate": 8.462100635075097e-06, + "loss": 10.557136535644531, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.039112288504838943, + "loss_total_v6": 0.23808689415454865, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 837 + }, + { + "epoch": 0.31233693626537457, + "grad_norm": 17.457454681396484, + "learning_rate": 8.457533543906923e-06, + "loss": 10.31386947631836, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5338636636734009, + "loss_total_v6": 0.7299085855484009, + "loss_wm_recon": 0.1591796875, + "loss_wm_recon_w": 0.039794921875, + "step": 838 + }, + { + "epoch": 0.3127096533730898, + "grad_norm": 18.018390655517578, + "learning_rate": 8.452960917802151e-06, + "loss": 11.008853912353516, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.738299548625946, + "loss_total_v6": 0.957537829875946, + "loss_wm_recon": 0.181640625, + "loss_wm_recon_w": 0.04541015625, + "step": 839 + }, + { + "epoch": 0.31308237048080506, + "grad_norm": 17.228593826293945, + "learning_rate": 8.448382764080794e-06, + "loss": 11.176362991333008, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4731014370918274, + "loss_total_v6": 0.6671932339668274, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 840 + }, + { + "epoch": 0.3134550875885203, + "grad_norm": 17.981285095214844, + "learning_rate": 8.443799090071716e-06, + "loss": 10.397979736328125, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5000571012496948, + "loss_total_v6": 0.6875571012496948, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 841 + }, + { + "epoch": 0.31382780469623556, + "grad_norm": 17.1243839263916, + "learning_rate": 8.439209903112617e-06, + "loss": 10.975902557373047, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.42484983801841736, + "loss_total_v6": 0.6074670553207397, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 842 + }, + { + "epoch": 0.3142005218039508, + "grad_norm": 17.599185943603516, + "learning_rate": 8.434615210550024e-06, + "loss": 11.03162670135498, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.149996817111969, + "loss_total_v6": 0.328951895236969, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 843 + }, + { + "epoch": 0.31457323891166605, + "grad_norm": 16.93216896057129, + "learning_rate": 8.430015019739274e-06, + "loss": 10.8333101272583, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.3648439347743988, + "loss_total_v6": 0.5442873239517212, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 844 + }, + { + "epoch": 0.3149459560193813, + "grad_norm": 18.46134376525879, + "learning_rate": 8.425409338044509e-06, + "loss": 11.259218215942383, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.514308750629425, + "loss_total_v6": 0.706691563129425, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 845 + }, + { + "epoch": 0.31531867312709655, + "grad_norm": 18.444486618041992, + "learning_rate": 8.42079817283866e-06, + "loss": 11.211040496826172, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.8022963404655457, + "loss_total_v6": 0.8462416529655457, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 846 + }, + { + "epoch": 0.3156913902348118, + "grad_norm": 17.915128707885742, + "learning_rate": 8.416181531503436e-06, + "loss": 10.644079208374023, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6991205811500549, + "loss_total_v6": 0.8983393311500549, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 847 + }, + { + "epoch": 0.31606410734252705, + "grad_norm": 19.237436294555664, + "learning_rate": 8.411559421429312e-06, + "loss": 11.002278327941895, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5007269382476807, + "loss_total_v6": 0.6953070163726807, + "loss_wm_recon": 0.1611328125, + "loss_wm_recon_w": 0.040283203125, + "step": 848 + }, + { + "epoch": 0.3164368244502423, + "grad_norm": 17.80813980102539, + "learning_rate": 8.406931850015519e-06, + "loss": 10.84538459777832, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.09292908012866974, + "loss_total_v6": 0.26993101835250854, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 849 + }, + { + "epoch": 0.3168095415579575, + "grad_norm": 17.856794357299805, + "learning_rate": 8.40229882467003e-06, + "loss": 11.035343170166016, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4889184832572937, + "loss_total_v6": 0.6864282488822937, + "loss_wm_recon": 0.1494140625, + "loss_wm_recon_w": 0.037353515625, + "step": 850 + }, + { + "epoch": 0.31718225866567273, + "grad_norm": 17.37639617919922, + "learning_rate": 8.397660352809546e-06, + "loss": 10.763286590576172, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7877388596534729, + "loss_total_v6": 0.9789009690284729, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 851 + }, + { + "epoch": 0.317554975773388, + "grad_norm": 17.87166404724121, + "learning_rate": 8.393016441859493e-06, + "loss": 11.173757553100586, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.34182870388031006, + "loss_total_v6": 0.5246900320053101, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 852 + }, + { + "epoch": 0.31792769288110323, + "grad_norm": 18.23757553100586, + "learning_rate": 8.388367099254002e-06, + "loss": 11.1879301071167, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.48110222816467285, + "loss_total_v6": 0.6610338687896729, + "loss_wm_recon": 0.1416015625, + "loss_wm_recon_w": 0.035400390625, + "step": 853 + }, + { + "epoch": 0.3183004099888185, + "grad_norm": 18.054365158081055, + "learning_rate": 8.383712332435892e-06, + "loss": 10.906827926635742, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6839125156402588, + "loss_total_v6": 0.8704359531402588, + "loss_wm_recon": 0.15234375, + "loss_wm_recon_w": 0.0380859375, + "step": 854 + }, + { + "epoch": 0.3186731270965337, + "grad_norm": 17.24694061279297, + "learning_rate": 8.379052148856675e-06, + "loss": 10.752659797668457, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.41665637493133545, + "loss_total_v6": 0.6295469999313354, + "loss_wm_recon": 0.1796875, + "loss_wm_recon_w": 0.044921875, + "step": 855 + }, + { + "epoch": 0.319045844204249, + "grad_norm": 17.528200149536133, + "learning_rate": 8.374386555976528e-06, + "loss": 10.994588851928711, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.8273427486419678, + "loss_total_v6": 1.0250966548919678, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 856 + }, + { + "epoch": 0.3194185613119642, + "grad_norm": 16.813701629638672, + "learning_rate": 8.369715561264295e-06, + "loss": 11.002857208251953, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3495194911956787, + "loss_total_v6": 0.5231034755706787, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 857 + }, + { + "epoch": 0.31979127841967947, + "grad_norm": 16.861730575561523, + "learning_rate": 8.365039172197455e-06, + "loss": 10.717174530029297, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4321209788322449, + "loss_total_v6": 0.6162030100822449, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 858 + }, + { + "epoch": 0.3201639955273947, + "grad_norm": 17.753746032714844, + "learning_rate": 8.360357396262136e-06, + "loss": 11.184101104736328, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6394201517105103, + "loss_total_v6": 0.8308264017105103, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 859 + }, + { + "epoch": 0.32053671263510997, + "grad_norm": 17.578983306884766, + "learning_rate": 8.355670240953082e-06, + "loss": 11.211069107055664, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5826131701469421, + "loss_total_v6": 0.7666952013969421, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 860 + }, + { + "epoch": 0.3209094297428252, + "grad_norm": 18.261280059814453, + "learning_rate": 8.350977713773645e-06, + "loss": 10.884723663330078, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4450872838497162, + "loss_total_v6": 0.6242865324020386, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 861 + }, + { + "epoch": 0.32128214685054046, + "grad_norm": 17.429689407348633, + "learning_rate": 8.346279822235788e-06, + "loss": 11.12757682800293, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.25799238681793213, + "loss_total_v6": 0.43841230869293213, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 862 + }, + { + "epoch": 0.3216548639582557, + "grad_norm": 17.822376251220703, + "learning_rate": 8.341576573860049e-06, + "loss": 10.645953178405762, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.44774895906448364, + "loss_total_v6": 0.6386669278144836, + "loss_wm_recon": 0.146484375, + "loss_wm_recon_w": 0.03662109375, + "step": 863 + }, + { + "epoch": 0.3220275810659709, + "grad_norm": 18.040283203125, + "learning_rate": 8.336867976175549e-06, + "loss": 10.943458557128906, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6023635864257812, + "loss_total_v6": 0.7901077270507812, + "loss_wm_recon": 0.1494140625, + "loss_wm_recon_w": 0.037353515625, + "step": 864 + }, + { + "epoch": 0.32240029817368615, + "grad_norm": 16.76264762878418, + "learning_rate": 8.332154036719971e-06, + "loss": 10.559425354003906, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5175008773803711, + "loss_total_v6": 0.7201375961303711, + "loss_wm_recon": 0.162109375, + "loss_wm_recon_w": 0.04052734375, + "step": 865 + }, + { + "epoch": 0.3227730152814014, + "grad_norm": 17.50139808654785, + "learning_rate": 8.32743476303955e-06, + "loss": 10.521183013916016, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.42599841952323914, + "loss_total_v6": 0.5978734493255615, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 866 + }, + { + "epoch": 0.32314573238911665, + "grad_norm": 18.19720458984375, + "learning_rate": 8.322710162689056e-06, + "loss": 11.004425048828125, + "loss_alignment": 0.37109375, + "loss_alignment_w": 0.185546875, + "loss_sft": 0.47741013765335083, + "loss_total_v6": 0.7120292782783508, + "loss_wm_recon": 0.1962890625, + "loss_wm_recon_w": 0.049072265625, + "step": 867 + }, + { + "epoch": 0.3235184494968319, + "grad_norm": 19.086650848388672, + "learning_rate": 8.317980243231788e-06, + "loss": 11.55465030670166, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.38571083545684814, + "loss_total_v6": 0.5863944292068481, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 868 + }, + { + "epoch": 0.32389116660454714, + "grad_norm": 17.65747833251953, + "learning_rate": 8.313245012239565e-06, + "loss": 10.802458763122559, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4992510974407196, + "loss_total_v6": 0.690413236618042, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 869 + }, + { + "epoch": 0.3242638837122624, + "grad_norm": 18.037818908691406, + "learning_rate": 8.308504477292704e-06, + "loss": 11.127660751342773, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3609935939311981, + "loss_total_v6": 0.5582592487335205, + "loss_wm_recon": 0.15625, + "loss_wm_recon_w": 0.0390625, + "step": 870 + }, + { + "epoch": 0.32463660081997764, + "grad_norm": 18.3907527923584, + "learning_rate": 8.303758645980013e-06, + "loss": 11.201944351196289, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5791814923286438, + "loss_total_v6": 0.6167791485786438, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 871 + }, + { + "epoch": 0.3250093179276929, + "grad_norm": 18.464027404785156, + "learning_rate": 8.299007525898776e-06, + "loss": 10.890579223632812, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.3282531499862671, + "loss_total_v6": 0.5208801031112671, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 872 + }, + { + "epoch": 0.32538203503540813, + "grad_norm": 16.81673240661621, + "learning_rate": 8.294251124654753e-06, + "loss": 10.744051933288574, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6580523252487183, + "loss_total_v6": 0.8553179502487183, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 873 + }, + { + "epoch": 0.3257547521431234, + "grad_norm": 17.98123550415039, + "learning_rate": 8.289489449862145e-06, + "loss": 10.875805854797363, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5825560092926025, + "loss_total_v6": 0.7795774936676025, + "loss_wm_recon": 0.1630859375, + "loss_wm_recon_w": 0.040771484375, + "step": 874 + }, + { + "epoch": 0.32612746925083863, + "grad_norm": 17.928752899169922, + "learning_rate": 8.284722509143608e-06, + "loss": 11.147884368896484, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.667991042137146, + "loss_total_v6": 0.880393385887146, + "loss_wm_recon": 0.177734375, + "loss_wm_recon_w": 0.04443359375, + "step": 875 + }, + { + "epoch": 0.3265001863585539, + "grad_norm": 17.765762329101562, + "learning_rate": 8.279950310130218e-06, + "loss": 10.534235000610352, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3924567997455597, + "loss_total_v6": 0.5784919261932373, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 876 + }, + { + "epoch": 0.3268729034662691, + "grad_norm": 16.92908477783203, + "learning_rate": 8.275172860461474e-06, + "loss": 10.7225980758667, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3197259306907654, + "loss_total_v6": 0.5062493681907654, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 877 + }, + { + "epoch": 0.3272456205739843, + "grad_norm": 18.19706153869629, + "learning_rate": 8.270390167785278e-06, + "loss": 11.442337036132812, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5321493744850159, + "loss_total_v6": 0.7250204682350159, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 878 + }, + { + "epoch": 0.32761833768169957, + "grad_norm": 17.839689254760742, + "learning_rate": 8.265602239757929e-06, + "loss": 10.886903762817383, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5202434659004211, + "loss_total_v6": 0.7199504971504211, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 879 + }, + { + "epoch": 0.3279910547894148, + "grad_norm": 18.500804901123047, + "learning_rate": 8.260809084044102e-06, + "loss": 11.050191879272461, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.42771196365356445, + "loss_total_v6": 0.6066670417785645, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 880 + }, + { + "epoch": 0.32836377189713006, + "grad_norm": 17.047168731689453, + "learning_rate": 8.256010708316839e-06, + "loss": 10.63095474243164, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.44024527072906494, + "loss_total_v6": 0.6221300363540649, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 881 + }, + { + "epoch": 0.3287364890048453, + "grad_norm": 18.507280349731445, + "learning_rate": 8.251207120257549e-06, + "loss": 11.445924758911133, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6350137591362, + "loss_total_v6": 0.8300821185112, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 882 + }, + { + "epoch": 0.32910920611256056, + "grad_norm": 17.502952575683594, + "learning_rate": 8.246398327555976e-06, + "loss": 10.905923843383789, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.45148414373397827, + "loss_total_v6": 0.6438669562339783, + "loss_wm_recon": 0.15234375, + "loss_wm_recon_w": 0.0380859375, + "step": 883 + }, + { + "epoch": 0.3294819232202758, + "grad_norm": 17.759689331054688, + "learning_rate": 8.241584337910195e-06, + "loss": 11.074392318725586, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.30126938223838806, + "loss_total_v6": 0.48486313223838806, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 884 + }, + { + "epoch": 0.32985464032799106, + "grad_norm": 17.594329833984375, + "learning_rate": 8.236765159026605e-06, + "loss": 11.000177383422852, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 1.143117904663086, + "loss_total_v6": 1.355276107788086, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 885 + }, + { + "epoch": 0.3302273574357063, + "grad_norm": 17.71466064453125, + "learning_rate": 8.231940798619913e-06, + "loss": 11.24614143371582, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.2926161587238312, + "loss_total_v6": 0.4976942837238312, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 886 + }, + { + "epoch": 0.33060007454342155, + "grad_norm": 16.951814651489258, + "learning_rate": 8.227111264413115e-06, + "loss": 10.86280345916748, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6672229766845703, + "loss_total_v6": 0.8517932891845703, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 887 + }, + { + "epoch": 0.3309727916511368, + "grad_norm": 17.66973876953125, + "learning_rate": 8.222276564137492e-06, + "loss": 11.0276460647583, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.21878622472286224, + "loss_total_v6": 0.40799522399902344, + "loss_wm_recon": 0.1552734375, + "loss_wm_recon_w": 0.038818359375, + "step": 888 + }, + { + "epoch": 0.33134550875885205, + "grad_norm": 16.778396606445312, + "learning_rate": 8.2174367055326e-06, + "loss": 10.730463027954102, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.43504762649536133, + "loss_total_v6": 0.6157116889953613, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 889 + }, + { + "epoch": 0.3317182258665673, + "grad_norm": 18.655006408691406, + "learning_rate": 8.212591696346244e-06, + "loss": 10.823308944702148, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.4663754403591156, + "loss_total_v6": 0.644109845161438, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 890 + }, + { + "epoch": 0.33209094297428254, + "grad_norm": 18.395082473754883, + "learning_rate": 8.207741544334482e-06, + "loss": 10.996978759765625, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4732736051082611, + "loss_total_v6": 0.6602853536605835, + "loss_wm_recon": 0.146484375, + "loss_wm_recon_w": 0.03662109375, + "step": 891 + }, + { + "epoch": 0.33246366008199774, + "grad_norm": 18.080657958984375, + "learning_rate": 8.202886257261598e-06, + "loss": 11.169897079467773, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.32581827044487, + "loss_total_v6": 0.5147831439971924, + "loss_wm_recon": 0.146484375, + "loss_wm_recon_w": 0.03662109375, + "step": 892 + }, + { + "epoch": 0.332836377189713, + "grad_norm": 17.58184814453125, + "learning_rate": 8.198025842900103e-06, + "loss": 10.769020080566406, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.41519176959991455, + "loss_total_v6": 0.6212464570999146, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 893 + }, + { + "epoch": 0.33320909429742823, + "grad_norm": 18.73101806640625, + "learning_rate": 8.193160309030708e-06, + "loss": 10.923580169677734, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6158775091171265, + "loss_total_v6": 0.8121665716171265, + "loss_wm_recon": 0.16015625, + "loss_wm_recon_w": 0.0400390625, + "step": 894 + }, + { + "epoch": 0.3335818114051435, + "grad_norm": 17.174978256225586, + "learning_rate": 8.18828966344233e-06, + "loss": 11.163101196289062, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.37555617094039917, + "loss_total_v6": 0.5586616396903992, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 895 + }, + { + "epoch": 0.33395452851285873, + "grad_norm": 17.05672264099121, + "learning_rate": 8.18341391393206e-06, + "loss": 10.417567253112793, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.17613300681114197, + "loss_total_v6": 0.35533222556114197, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 896 + }, + { + "epoch": 0.334327245620574, + "grad_norm": 18.16240692138672, + "learning_rate": 8.17853306830516e-06, + "loss": 10.759359359741211, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.37605053186416626, + "loss_total_v6": 0.5669685006141663, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 897 + }, + { + "epoch": 0.3346999627282892, + "grad_norm": 18.204029083251953, + "learning_rate": 8.173647134375058e-06, + "loss": 10.63343620300293, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.7811546325683594, + "loss_total_v6": 0.9591331481933594, + "loss_wm_recon": 0.1416015625, + "loss_wm_recon_w": 0.035400390625, + "step": 898 + }, + { + "epoch": 0.3350726798360045, + "grad_norm": 17.78144073486328, + "learning_rate": 8.16875611996332e-06, + "loss": 11.121180534362793, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6839061975479126, + "loss_total_v6": 0.7210155725479126, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 899 + }, + { + "epoch": 0.3354453969437197, + "grad_norm": 19.053916931152344, + "learning_rate": 8.163860032899646e-06, + "loss": 11.004119873046875, + "loss_alignment": 0.359375, + "loss_alignment_w": 0.1796875, + "loss_sft": 0.5330380797386169, + "loss_total_v6": 0.7625302672386169, + "loss_wm_recon": 0.19921875, + "loss_wm_recon_w": 0.0498046875, + "step": 900 + }, + { + "epoch": 0.3354453969437197, + "eval_loss": 0.6578972935676575, + "eval_loss_alignment": 0.2839438935630342, + "eval_loss_alignment_w": 0.1419719467815171, + "eval_loss_sft": 0.4811150649976399, + "eval_loss_total_v6": 0.6578972651344589, + "eval_loss_wm_recon": 0.13924101479033119, + "eval_loss_wm_recon_w": 0.034810253697582796, + "eval_runtime": 227.2168, + "eval_samples_per_second": 8.212, + "eval_steps_per_second": 1.03, + "step": 900 + }, + { + "epoch": 0.33581811405143497, + "grad_norm": 18.57257652282715, + "learning_rate": 8.15895888102186e-06, + "loss": 11.0081205368042, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5815721750259399, + "loss_total_v6": 0.7700487375259399, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 901 + }, + { + "epoch": 0.3361908311591502, + "grad_norm": 17.259929656982422, + "learning_rate": 8.154052672175888e-06, + "loss": 10.678447723388672, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5567764043807983, + "loss_total_v6": 0.7516006231307983, + "loss_wm_recon": 0.146484375, + "loss_wm_recon_w": 0.03662109375, + "step": 902 + }, + { + "epoch": 0.33656354826686546, + "grad_norm": 17.425735473632812, + "learning_rate": 8.149141414215759e-06, + "loss": 10.690276145935059, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.6890512704849243, + "loss_total_v6": 0.9095102548599243, + "loss_wm_recon": 0.1865234375, + "loss_wm_recon_w": 0.046630859375, + "step": 903 + }, + { + "epoch": 0.3369362653745807, + "grad_norm": 17.744932174682617, + "learning_rate": 8.144225115003578e-06, + "loss": 10.492607116699219, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5570262670516968, + "loss_total_v6": 0.7589305639266968, + "loss_wm_recon": 0.1591796875, + "loss_wm_recon_w": 0.039794921875, + "step": 904 + }, + { + "epoch": 0.33730898248229596, + "grad_norm": 17.858230590820312, + "learning_rate": 8.139303782409527e-06, + "loss": 10.588359832763672, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.44211965799331665, + "loss_total_v6": 0.6354790329933167, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 905 + }, + { + "epoch": 0.3376816995900112, + "grad_norm": 17.542179107666016, + "learning_rate": 8.134377424311838e-06, + "loss": 11.36000919342041, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.408169686794281, + "loss_total_v6": 0.584195077419281, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 906 + }, + { + "epoch": 0.3380544166977264, + "grad_norm": 17.24946403503418, + "learning_rate": 8.12944604859679e-06, + "loss": 11.122291564941406, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.645226001739502, + "loss_total_v6": 0.835655689239502, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 907 + }, + { + "epoch": 0.33842713380544165, + "grad_norm": 18.42997169494629, + "learning_rate": 8.124509663158699e-06, + "loss": 10.928701400756836, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5546146035194397, + "loss_total_v6": 0.7355228066444397, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 908 + }, + { + "epoch": 0.3387998509131569, + "grad_norm": 18.173349380493164, + "learning_rate": 8.119568275899897e-06, + "loss": 10.71731185913086, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.600070595741272, + "loss_total_v6": 0.806613564491272, + "loss_wm_recon": 0.162109375, + "loss_wm_recon_w": 0.04052734375, + "step": 909 + }, + { + "epoch": 0.33917256802087214, + "grad_norm": 16.859298706054688, + "learning_rate": 8.114621894730725e-06, + "loss": 10.665241241455078, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7860665917396545, + "loss_total_v6": 0.9774728417396545, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 910 + }, + { + "epoch": 0.3395452851285874, + "grad_norm": 17.408964157104492, + "learning_rate": 8.109670527569517e-06, + "loss": 10.823158264160156, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4471644461154938, + "loss_total_v6": 0.6358851194381714, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 911 + }, + { + "epoch": 0.33991800223630264, + "grad_norm": 18.18544578552246, + "learning_rate": 8.104714182342586e-06, + "loss": 10.783727645874023, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.35624220967292786, + "loss_total_v6": 0.5256757736206055, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 912 + }, + { + "epoch": 0.3402907193440179, + "grad_norm": 18.54878807067871, + "learning_rate": 8.09975286698422e-06, + "loss": 10.710648536682129, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5467882752418518, + "loss_total_v6": 0.7303820252418518, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 913 + }, + { + "epoch": 0.34066343645173314, + "grad_norm": 17.241331100463867, + "learning_rate": 8.09478658943666e-06, + "loss": 11.032183647155762, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4558236002922058, + "loss_total_v6": 0.6445443034172058, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 914 + }, + { + "epoch": 0.3410361535594484, + "grad_norm": 17.538042068481445, + "learning_rate": 8.08981535765009e-06, + "loss": 11.061264038085938, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5889388918876648, + "loss_total_v6": 0.7769271731376648, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 915 + }, + { + "epoch": 0.34140887066716363, + "grad_norm": 16.5853214263916, + "learning_rate": 8.084839179582626e-06, + "loss": 10.760754585266113, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.8583781123161316, + "loss_total_v6": 1.0534465312957764, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 916 + }, + { + "epoch": 0.3417815877748789, + "grad_norm": 17.994178771972656, + "learning_rate": 8.079858063200303e-06, + "loss": 11.022127151489258, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.46324023604393005, + "loss_total_v6": 0.6592851877212524, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 917 + }, + { + "epoch": 0.34215430488259413, + "grad_norm": 18.960235595703125, + "learning_rate": 8.07487201647706e-06, + "loss": 11.228610038757324, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6388012766838074, + "loss_total_v6": 0.8221508860588074, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 918 + }, + { + "epoch": 0.3425270219903094, + "grad_norm": 17.459848403930664, + "learning_rate": 8.06988104739473e-06, + "loss": 11.156917572021484, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5634446144104004, + "loss_total_v6": 0.7570481300354004, + "loss_wm_recon": 0.1494140625, + "loss_wm_recon_w": 0.037353515625, + "step": 919 + }, + { + "epoch": 0.3428997390980246, + "grad_norm": 17.462121963500977, + "learning_rate": 8.064885163943022e-06, + "loss": 10.786858558654785, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6821396350860596, + "loss_total_v6": 0.8698837757110596, + "loss_wm_recon": 0.1416015625, + "loss_wm_recon_w": 0.035400390625, + "step": 920 + }, + { + "epoch": 0.3432724562057398, + "grad_norm": 17.04437255859375, + "learning_rate": 8.059884374119518e-06, + "loss": 11.14362907409668, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6988535523414612, + "loss_total_v6": 0.8795176148414612, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 921 + }, + { + "epoch": 0.34364517331345507, + "grad_norm": 17.30283546447754, + "learning_rate": 8.054878685929652e-06, + "loss": 10.901667594909668, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4321482479572296, + "loss_total_v6": 0.6125681400299072, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 922 + }, + { + "epoch": 0.3440178904211703, + "grad_norm": 18.31739044189453, + "learning_rate": 8.049868107386696e-06, + "loss": 10.976272583007812, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5851641893386841, + "loss_total_v6": 0.7831622362136841, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 923 + }, + { + "epoch": 0.34439060752888556, + "grad_norm": 17.17140769958496, + "learning_rate": 8.044852646511756e-06, + "loss": 11.044660568237305, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.559794545173645, + "loss_total_v6": 0.745829701423645, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 924 + }, + { + "epoch": 0.3447633246366008, + "grad_norm": 17.66133689880371, + "learning_rate": 8.039832311333753e-06, + "loss": 11.281731605529785, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.41814368963241577, + "loss_total_v6": 0.6019815802574158, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 925 + }, + { + "epoch": 0.34513604174431606, + "grad_norm": 17.822160720825195, + "learning_rate": 8.034807109889405e-06, + "loss": 10.958686828613281, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.25803643465042114, + "loss_total_v6": 0.44065362215042114, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 926 + }, + { + "epoch": 0.3455087588520313, + "grad_norm": 17.5854549407959, + "learning_rate": 8.029777050223226e-06, + "loss": 10.38157844543457, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.436903715133667, + "loss_total_v6": 0.620009183883667, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 927 + }, + { + "epoch": 0.34588147595974655, + "grad_norm": 17.712167739868164, + "learning_rate": 8.024742140387506e-06, + "loss": 11.297198295593262, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5550618171691895, + "loss_total_v6": 0.5972981452941895, + "loss_wm_recon": 0.1689453125, + "loss_wm_recon_w": 0.042236328125, + "step": 928 + }, + { + "epoch": 0.3462541930674618, + "grad_norm": 17.853567123413086, + "learning_rate": 8.019702388442299e-06, + "loss": 11.034507751464844, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.7300142645835876, + "loss_total_v6": 0.9260591864585876, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 929 + }, + { + "epoch": 0.34662691017517705, + "grad_norm": 17.719058990478516, + "learning_rate": 8.01465780245541e-06, + "loss": 10.907594680786133, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.36202725768089294, + "loss_total_v6": 0.5341464281082153, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 930 + }, + { + "epoch": 0.3469996272828923, + "grad_norm": 17.584993362426758, + "learning_rate": 8.00960839050238e-06, + "loss": 10.766534805297852, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5633036494255066, + "loss_total_v6": 0.7508036494255066, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 931 + }, + { + "epoch": 0.34737234439060755, + "grad_norm": 17.453758239746094, + "learning_rate": 8.004554160666487e-06, + "loss": 10.82127857208252, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.7750529050827026, + "loss_total_v6": 0.9598673582077026, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 932 + }, + { + "epoch": 0.3477450614983228, + "grad_norm": 17.364580154418945, + "learning_rate": 7.999495121038702e-06, + "loss": 10.724637985229492, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.49949949979782104, + "loss_total_v6": 0.687731921672821, + "loss_wm_recon": 0.1357421875, + "loss_wm_recon_w": 0.033935546875, + "step": 933 + }, + { + "epoch": 0.34811777860603804, + "grad_norm": 17.272531509399414, + "learning_rate": 7.994431279717715e-06, + "loss": 10.867685317993164, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.40304237604141235, + "loss_total_v6": 0.5827298760414124, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 934 + }, + { + "epoch": 0.34849049571375323, + "grad_norm": 18.75432777404785, + "learning_rate": 7.989362644809892e-06, + "loss": 10.712730407714844, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3305201828479767, + "loss_total_v6": 0.5124049186706543, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 935 + }, + { + "epoch": 0.3488632128214685, + "grad_norm": 17.586214065551758, + "learning_rate": 7.984289224429273e-06, + "loss": 10.994365692138672, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.43930789828300476, + "loss_total_v6": 0.6280286312103271, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 936 + }, + { + "epoch": 0.34923592992918373, + "grad_norm": 17.882625579833984, + "learning_rate": 7.979211026697562e-06, + "loss": 10.787055015563965, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7275271415710449, + "loss_total_v6": 0.9225955009460449, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 937 + }, + { + "epoch": 0.349608647036899, + "grad_norm": 17.149198532104492, + "learning_rate": 7.97412805974411e-06, + "loss": 10.920546531677246, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.47902098298072815, + "loss_total_v6": 0.6623705625534058, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 938 + }, + { + "epoch": 0.3499813641446142, + "grad_norm": 17.167951583862305, + "learning_rate": 7.969040331705904e-06, + "loss": 10.489460945129395, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3138134777545929, + "loss_total_v6": 0.5018017292022705, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 939 + }, + { + "epoch": 0.3503540812523295, + "grad_norm": 18.019512176513672, + "learning_rate": 7.963947850727548e-06, + "loss": 10.853931427001953, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.527168333530426, + "loss_total_v6": 0.704658567905426, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 940 + }, + { + "epoch": 0.3507267983600447, + "grad_norm": 18.178442001342773, + "learning_rate": 7.95885062496126e-06, + "loss": 11.207778930664062, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.2820194363594055, + "loss_total_v6": 0.4513309597969055, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 941 + }, + { + "epoch": 0.35109951546775997, + "grad_norm": 18.481794357299805, + "learning_rate": 7.95374866256685e-06, + "loss": 10.608044624328613, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.5009674429893494, + "loss_total_v6": 0.7087311148643494, + "loss_wm_recon": 0.1669921875, + "loss_wm_recon_w": 0.041748046875, + "step": 942 + }, + { + "epoch": 0.3514722325754752, + "grad_norm": 18.30703353881836, + "learning_rate": 7.948641971711712e-06, + "loss": 11.00224781036377, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6169734597206116, + "loss_total_v6": 0.7969051003456116, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 943 + }, + { + "epoch": 0.35184494968319047, + "grad_norm": 19.094364166259766, + "learning_rate": 7.943530560570809e-06, + "loss": 10.263162612915039, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5512667298316956, + "loss_total_v6": 0.7331514954566956, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 944 + }, + { + "epoch": 0.3522176667909057, + "grad_norm": 17.313627243041992, + "learning_rate": 7.938414437326663e-06, + "loss": 11.058963775634766, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4599321186542511, + "loss_total_v6": 0.6449906826019287, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 945 + }, + { + "epoch": 0.35259038389862096, + "grad_norm": 17.9998779296875, + "learning_rate": 7.933293610169337e-06, + "loss": 10.775196075439453, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.13082614541053772, + "loss_total_v6": 0.3053866922855377, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 946 + }, + { + "epoch": 0.3529631010063362, + "grad_norm": 17.17559242248535, + "learning_rate": 7.92816808729642e-06, + "loss": 10.67563533782959, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.39780372381210327, + "loss_total_v6": 0.5955576300621033, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 947 + }, + { + "epoch": 0.35333581811405146, + "grad_norm": 19.015565872192383, + "learning_rate": 7.923037876913027e-06, + "loss": 11.127739906311035, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5130589604377747, + "loss_total_v6": 0.5487034916877747, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 948 + }, + { + "epoch": 0.3537085352217667, + "grad_norm": 17.655258178710938, + "learning_rate": 7.917902987231769e-06, + "loss": 10.884132385253906, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.3532223105430603, + "loss_total_v6": 0.5492672324180603, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 949 + }, + { + "epoch": 0.3540812523294819, + "grad_norm": 18.08584976196289, + "learning_rate": 7.912763426472754e-06, + "loss": 11.121757507324219, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.12194833904504776, + "loss_total_v6": 0.31384286284446716, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 950 + }, + { + "epoch": 0.35445396943719715, + "grad_norm": 17.904359817504883, + "learning_rate": 7.907619202863561e-06, + "loss": 10.797155380249023, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5758099555969238, + "loss_total_v6": 0.7696576118469238, + "loss_wm_recon": 0.150390625, + "loss_wm_recon_w": 0.03759765625, + "step": 951 + }, + { + "epoch": 0.3548266865449124, + "grad_norm": 18.508485794067383, + "learning_rate": 7.902470324639242e-06, + "loss": 10.672149658203125, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4843885600566864, + "loss_total_v6": 0.6696913242340088, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 952 + }, + { + "epoch": 0.35519940365262764, + "grad_norm": 18.002487182617188, + "learning_rate": 7.897316800042292e-06, + "loss": 10.527928352355957, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.3878730237483978, + "loss_total_v6": 0.5831855535507202, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 953 + }, + { + "epoch": 0.3555721207603429, + "grad_norm": 17.539703369140625, + "learning_rate": 7.892158637322647e-06, + "loss": 10.539676666259766, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5157822370529175, + "loss_total_v6": 0.6944931745529175, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 954 + }, + { + "epoch": 0.35594483786805814, + "grad_norm": 16.777292251586914, + "learning_rate": 7.886995844737668e-06, + "loss": 10.512140274047852, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.2312973141670227, + "loss_total_v6": 0.4393051266670227, + "loss_wm_recon": 0.16015625, + "loss_wm_recon_w": 0.0400390625, + "step": 955 + }, + { + "epoch": 0.3563175549757734, + "grad_norm": 17.355052947998047, + "learning_rate": 7.88182843055213e-06, + "loss": 10.605426788330078, + "loss_alignment": 0.359375, + "loss_alignment_w": 0.1796875, + "loss_sft": 0.6337123513221741, + "loss_total_v6": 0.8573451638221741, + "loss_wm_recon": 0.17578125, + "loss_wm_recon_w": 0.0439453125, + "step": 956 + }, + { + "epoch": 0.35669027208348864, + "grad_norm": 17.845735549926758, + "learning_rate": 7.876656403038204e-06, + "loss": 10.651439666748047, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5003567337989807, + "loss_total_v6": 0.6834622025489807, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 957 + }, + { + "epoch": 0.3570629891912039, + "grad_norm": 17.333742141723633, + "learning_rate": 7.871479770475445e-06, + "loss": 10.823661804199219, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5160931944847107, + "loss_total_v6": 0.5522260069847107, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 958 + }, + { + "epoch": 0.35743570629891913, + "grad_norm": 17.897504806518555, + "learning_rate": 7.866298541150784e-06, + "loss": 10.687593460083008, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3239492177963257, + "loss_total_v6": 0.5053457021713257, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 959 + }, + { + "epoch": 0.3578084234066344, + "grad_norm": 17.537683486938477, + "learning_rate": 7.861112723358506e-06, + "loss": 11.072528839111328, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.34031492471694946, + "loss_total_v6": 0.5341625809669495, + "loss_wm_recon": 0.158203125, + "loss_wm_recon_w": 0.03955078125, + "step": 960 + }, + { + "epoch": 0.3581811405143496, + "grad_norm": 17.54836082458496, + "learning_rate": 7.855922325400245e-06, + "loss": 10.529918670654297, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.8298527002334595, + "loss_total_v6": 0.8613468408584595, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 961 + }, + { + "epoch": 0.3585538576220649, + "grad_norm": 17.77779769897461, + "learning_rate": 7.850727355584963e-06, + "loss": 10.711563110351562, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.2093304991722107, + "loss_total_v6": 0.3851117491722107, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 962 + }, + { + "epoch": 0.3589265747297801, + "grad_norm": 17.79818344116211, + "learning_rate": 7.845527822228945e-06, + "loss": 10.944125175476074, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5374952554702759, + "loss_total_v6": 0.7208448648452759, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 963 + }, + { + "epoch": 0.3592992918374953, + "grad_norm": 17.41572380065918, + "learning_rate": 7.84032373365578e-06, + "loss": 10.617094039916992, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6617028713226318, + "loss_total_v6": 0.8406579494476318, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 964 + }, + { + "epoch": 0.35967200894521056, + "grad_norm": 17.289268493652344, + "learning_rate": 7.835115098196348e-06, + "loss": 11.027383804321289, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5616784691810608, + "loss_total_v6": 0.7503991723060608, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 965 + }, + { + "epoch": 0.3600447260529258, + "grad_norm": 23.11572265625, + "learning_rate": 7.82990192418881e-06, + "loss": 10.626394271850586, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5231831669807434, + "loss_total_v6": 0.7021382451057434, + "loss_wm_recon": 0.1220703125, + "loss_wm_recon_w": 0.030517578125, + "step": 966 + }, + { + "epoch": 0.36041744316064106, + "grad_norm": 18.107769012451172, + "learning_rate": 7.824684219978591e-06, + "loss": 10.656982421875, + "loss_alignment": 0.26171875, + "loss_alignment_w": 0.130859375, + "loss_sft": 0.4820220470428467, + "loss_total_v6": 0.6475493907928467, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 967 + }, + { + "epoch": 0.3607901602683563, + "grad_norm": 17.54461097717285, + "learning_rate": 7.819461993918367e-06, + "loss": 11.070980072021484, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.7073732018470764, + "loss_total_v6": 0.8803468346595764, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 968 + }, + { + "epoch": 0.36116287737607156, + "grad_norm": 17.809764862060547, + "learning_rate": 7.814235254368054e-06, + "loss": 10.884488105773926, + "loss_alignment": 0.38671875, + "loss_alignment_w": 0.193359375, + "loss_sft": 0.8099045157432556, + "loss_total_v6": 1.0552659034729004, + "loss_wm_recon": 0.2080078125, + "loss_wm_recon_w": 0.052001953125, + "step": 969 + }, + { + "epoch": 0.3615355944837868, + "grad_norm": 16.961122512817383, + "learning_rate": 7.809004009694795e-06, + "loss": 11.088051795959473, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.48155003786087036, + "loss_total_v6": 0.6666086316108704, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 970 + }, + { + "epoch": 0.36190831159150205, + "grad_norm": 18.28779411315918, + "learning_rate": 7.803768268272945e-06, + "loss": 10.977994918823242, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.45639631152153015, + "loss_total_v6": 0.6478025913238525, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 971 + }, + { + "epoch": 0.3622810286992173, + "grad_norm": 17.182701110839844, + "learning_rate": 7.79852803848405e-06, + "loss": 10.64803695678711, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3321394622325897, + "loss_total_v6": 0.5186629295349121, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 972 + }, + { + "epoch": 0.36265374580693255, + "grad_norm": 17.697561264038086, + "learning_rate": 7.793283328716853e-06, + "loss": 10.946630477905273, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5672591924667358, + "loss_total_v6": 0.7496322393417358, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 973 + }, + { + "epoch": 0.3630264629146478, + "grad_norm": 17.75999641418457, + "learning_rate": 7.78803414736726e-06, + "loss": 10.627265930175781, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.39998993277549744, + "loss_total_v6": 0.5918844938278198, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 974 + }, + { + "epoch": 0.36339918002236304, + "grad_norm": 16.92551612854004, + "learning_rate": 7.782780502838337e-06, + "loss": 10.470763206481934, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5471495389938354, + "loss_total_v6": 0.7234190702438354, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 975 + }, + { + "epoch": 0.3637718971300783, + "grad_norm": 19.796138763427734, + "learning_rate": 7.7775224035403e-06, + "loss": 10.999799728393555, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.37707921862602234, + "loss_total_v6": 0.5716593265533447, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 976 + }, + { + "epoch": 0.36414461423779354, + "grad_norm": 18.13018035888672, + "learning_rate": 7.772259857890485e-06, + "loss": 10.601585388183594, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.49792924523353577, + "loss_total_v6": 0.6722456216812134, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 977 + }, + { + "epoch": 0.36451733134550873, + "grad_norm": 17.787015914916992, + "learning_rate": 7.76699287431336e-06, + "loss": 10.310718536376953, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5425460338592529, + "loss_total_v6": 0.7305343151092529, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 978 + }, + { + "epoch": 0.364890048453224, + "grad_norm": 17.468338012695312, + "learning_rate": 7.761721461240487e-06, + "loss": 10.672839164733887, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6735244989395142, + "loss_total_v6": 0.8717666864395142, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 979 + }, + { + "epoch": 0.36526276556093923, + "grad_norm": 18.553396224975586, + "learning_rate": 7.756445627110523e-06, + "loss": 10.823522567749023, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3231271207332611, + "loss_total_v6": 0.5081857442855835, + "loss_wm_recon": 0.146484375, + "loss_wm_recon_w": 0.03662109375, + "step": 980 + }, + { + "epoch": 0.3656354826686545, + "grad_norm": 17.95220375061035, + "learning_rate": 7.7511653803692e-06, + "loss": 10.874261856079102, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.42790061235427856, + "loss_total_v6": 0.6129592061042786, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 981 + }, + { + "epoch": 0.3660081997763697, + "grad_norm": 17.16049575805664, + "learning_rate": 7.74588072946932e-06, + "loss": 10.738383293151855, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.3605470359325409, + "loss_total_v6": 0.5387697219848633, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 982 + }, + { + "epoch": 0.366380916884085, + "grad_norm": 18.462646484375, + "learning_rate": 7.740591682870725e-06, + "loss": 10.728446960449219, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6184861660003662, + "loss_total_v6": 0.8199021816253662, + "loss_wm_recon": 0.1494140625, + "loss_wm_recon_w": 0.037353515625, + "step": 983 + }, + { + "epoch": 0.3667536339918002, + "grad_norm": 17.803129196166992, + "learning_rate": 7.735298249040304e-06, + "loss": 10.515602111816406, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3264462351799011, + "loss_total_v6": 0.5115048289299011, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 984 + }, + { + "epoch": 0.36712635109951547, + "grad_norm": 17.130653381347656, + "learning_rate": 7.730000436451963e-06, + "loss": 10.47087287902832, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5807532072067261, + "loss_total_v6": 0.7714270353317261, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 985 + }, + { + "epoch": 0.3674990682072307, + "grad_norm": 17.719287872314453, + "learning_rate": 7.72469825358662e-06, + "loss": 11.062845230102539, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.9036062955856323, + "loss_total_v6": 1.0869559049606323, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 986 + }, + { + "epoch": 0.36787178531494596, + "grad_norm": 18.39997673034668, + "learning_rate": 7.719391708932189e-06, + "loss": 10.39082145690918, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.45502060651779175, + "loss_total_v6": 0.6491124033927917, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 987 + }, + { + "epoch": 0.3682445024226612, + "grad_norm": 17.410694122314453, + "learning_rate": 7.714080810983564e-06, + "loss": 10.287286758422852, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3623964488506317, + "loss_total_v6": 0.5516054630279541, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 988 + }, + { + "epoch": 0.36861721953037646, + "grad_norm": 16.944887161254883, + "learning_rate": 7.708765568242613e-06, + "loss": 11.004501342773438, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5920816659927368, + "loss_total_v6": 0.6243082284927368, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 989 + }, + { + "epoch": 0.3689899366380917, + "grad_norm": 17.573163986206055, + "learning_rate": 7.703445989218154e-06, + "loss": 10.728073120117188, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.7106022238731384, + "loss_total_v6": 0.9007877707481384, + "loss_wm_recon": 0.1357421875, + "loss_wm_recon_w": 0.033935546875, + "step": 990 + }, + { + "epoch": 0.36936265374580696, + "grad_norm": 17.310800552368164, + "learning_rate": 7.69812208242595e-06, + "loss": 10.32693099975586, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.8557078838348389, + "loss_total_v6": 1.0373485088348389, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 991 + }, + { + "epoch": 0.36973537085352215, + "grad_norm": 17.537616729736328, + "learning_rate": 7.692793856388692e-06, + "loss": 10.618480682373047, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.746173620223999, + "loss_total_v6": 0.948077917098999, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 992 + }, + { + "epoch": 0.3701080879612374, + "grad_norm": 17.102365493774414, + "learning_rate": 7.68746131963598e-06, + "loss": 10.887691497802734, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.8276041746139526, + "loss_total_v6": 1.0212076902389526, + "loss_wm_recon": 0.1416015625, + "loss_wm_recon_w": 0.035400390625, + "step": 993 + }, + { + "epoch": 0.37048080506895265, + "grad_norm": 17.892427444458008, + "learning_rate": 7.682124480704328e-06, + "loss": 10.786559104919434, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.7342096567153931, + "loss_total_v6": 0.9251276254653931, + "loss_wm_recon": 0.154296875, + "loss_wm_recon_w": 0.03857421875, + "step": 994 + }, + { + "epoch": 0.3708535221766679, + "grad_norm": 17.52199363708496, + "learning_rate": 7.676783348137119e-06, + "loss": 11.056588172912598, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5258287787437439, + "loss_total_v6": 0.7201647162437439, + "loss_wm_recon": 0.16015625, + "loss_wm_recon_w": 0.0400390625, + "step": 995 + }, + { + "epoch": 0.37122623928438314, + "grad_norm": 17.199201583862305, + "learning_rate": 7.671437930484622e-06, + "loss": 10.750809669494629, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.432812362909317, + "loss_total_v6": 0.6081053018569946, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 996 + }, + { + "epoch": 0.3715989563920984, + "grad_norm": 17.24632453918457, + "learning_rate": 7.66608823630396e-06, + "loss": 10.68938159942627, + "loss_alignment": 0.33984375, + "loss_alignment_w": 0.169921875, + "loss_sft": 0.43621692061424255, + "loss_total_v6": 0.6444689035415649, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 997 + }, + { + "epoch": 0.37197167349981364, + "grad_norm": 18.95355796813965, + "learning_rate": 7.660734274159112e-06, + "loss": 10.665352821350098, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.43905290961265564, + "loss_total_v6": 0.47054705023765564, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 998 + }, + { + "epoch": 0.3723443906075289, + "grad_norm": 16.922672271728516, + "learning_rate": 7.655376052620874e-06, + "loss": 10.630687713623047, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.29818668961524963, + "loss_total_v6": 0.48031559586524963, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 999 + }, + { + "epoch": 0.37271710771524413, + "grad_norm": 17.3173828125, + "learning_rate": 7.650013580266871e-06, + "loss": 10.976400375366211, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.29914140701293945, + "loss_total_v6": 0.47382402420043945, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1000 + }, + { + "epoch": 0.37271710771524413, + "eval_loss": 0.6516885161399841, + "eval_loss_alignment": 0.2831259181356838, + "eval_loss_alignment_w": 0.1415629590678419, + "eval_loss_sft": 0.47775622175481075, + "eval_loss_total_v6": 0.6516885502407184, + "eval_loss_wm_recon": 0.12947747646233973, + "eval_loss_wm_recon_w": 0.03236936911558493, + "eval_runtime": 227.2895, + "eval_samples_per_second": 8.21, + "eval_steps_per_second": 1.03, + "step": 1000 + }, + { + "epoch": 0.3730898248229594, + "grad_norm": 18.567415237426758, + "learning_rate": 7.644646865681528e-06, + "loss": 10.743598937988281, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6862227320671082, + "loss_total_v6": 0.8812910914421082, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 1001 + }, + { + "epoch": 0.37346254193067463, + "grad_norm": 17.819843292236328, + "learning_rate": 7.639275917456068e-06, + "loss": 10.50001335144043, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.46344342827796936, + "loss_total_v6": 0.664126992225647, + "loss_wm_recon": 0.146484375, + "loss_wm_recon_w": 0.03662109375, + "step": 1002 + }, + { + "epoch": 0.3738352590383899, + "grad_norm": 17.44097328186035, + "learning_rate": 7.63390074418848e-06, + "loss": 11.25491714477539, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6404586434364319, + "loss_total_v6": 0.8333297371864319, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1003 + }, + { + "epoch": 0.3742079761461051, + "grad_norm": 16.90302276611328, + "learning_rate": 7.628521354483528e-06, + "loss": 10.843755722045898, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4642047882080078, + "loss_total_v6": 0.6485309600830078, + "loss_wm_recon": 0.1357421875, + "loss_wm_recon_w": 0.033935546875, + "step": 1004 + }, + { + "epoch": 0.3745806932538204, + "grad_norm": 17.848827362060547, + "learning_rate": 7.623137756952719e-06, + "loss": 10.779977798461914, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6131316423416138, + "loss_total_v6": 0.7928191423416138, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1005 + }, + { + "epoch": 0.3749534103615356, + "grad_norm": 17.220117568969727, + "learning_rate": 7.6177499602143e-06, + "loss": 11.081338882446289, + "loss_alignment": 0.3515625, + "loss_alignment_w": 0.17578125, + "loss_sft": 0.8450026512145996, + "loss_total_v6": 1.0659499168395996, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 1006 + }, + { + "epoch": 0.3753261274692508, + "grad_norm": 17.004806518554688, + "learning_rate": 7.612357972893234e-06, + "loss": 10.790410041809082, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.43103450536727905, + "loss_total_v6": 0.616337239742279, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 1007 + }, + { + "epoch": 0.37569884457696606, + "grad_norm": 17.55887222290039, + "learning_rate": 7.606961803621201e-06, + "loss": 10.644010543823242, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.32829520106315613, + "loss_total_v6": 0.49333426356315613, + "loss_wm_recon": 0.11328125, + "loss_wm_recon_w": 0.0283203125, + "step": 1008 + }, + { + "epoch": 0.3760715616846813, + "grad_norm": 17.60965347290039, + "learning_rate": 7.601561461036568e-06, + "loss": 11.23830795288086, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.2866858243942261, + "loss_total_v6": 0.4602698087692261, + "loss_wm_recon": 0.1240234375, + "loss_wm_recon_w": 0.031005859375, + "step": 1009 + }, + { + "epoch": 0.37644427879239656, + "grad_norm": 18.02094078063965, + "learning_rate": 7.596156953784388e-06, + "loss": 10.3942289352417, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.28911346197128296, + "loss_total_v6": 0.47905486822128296, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 1010 + }, + { + "epoch": 0.3768169959001118, + "grad_norm": 16.48710060119629, + "learning_rate": 7.590748290516379e-06, + "loss": 10.706462860107422, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.18350891768932343, + "loss_total_v6": 0.3600226044654846, + "loss_wm_recon": 0.1201171875, + "loss_wm_recon_w": 0.030029296875, + "step": 1011 + }, + { + "epoch": 0.37718971300782705, + "grad_norm": 17.315170288085938, + "learning_rate": 7.58533547989091e-06, + "loss": 10.772911071777344, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.3875993490219116, + "loss_total_v6": 0.5565446615219116, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1012 + }, + { + "epoch": 0.3775624301155423, + "grad_norm": 17.822458267211914, + "learning_rate": 7.579918530572994e-06, + "loss": 10.915589332580566, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6232948303222656, + "loss_total_v6": 0.8090858459472656, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1013 + }, + { + "epoch": 0.37793514722325755, + "grad_norm": 17.407474517822266, + "learning_rate": 7.574497451234264e-06, + "loss": 10.115029335021973, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.5709973573684692, + "loss_total_v6": 0.7699719667434692, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 1014 + }, + { + "epoch": 0.3783078643309728, + "grad_norm": 18.636953353881836, + "learning_rate": 7.569072250552968e-06, + "loss": 10.328688621520996, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5995364189147949, + "loss_total_v6": 0.7845950126647949, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 1015 + }, + { + "epoch": 0.37868058143868805, + "grad_norm": 18.517620086669922, + "learning_rate": 7.563642937213952e-06, + "loss": 10.617883682250977, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.561863899230957, + "loss_total_v6": 0.752049446105957, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 1016 + }, + { + "epoch": 0.3790532985464033, + "grad_norm": 18.042184829711914, + "learning_rate": 7.558209519908644e-06, + "loss": 11.010662078857422, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.4819803237915039, + "loss_total_v6": 0.6521463394165039, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1017 + }, + { + "epoch": 0.37942601565411854, + "grad_norm": 17.092090606689453, + "learning_rate": 7.5527720073350385e-06, + "loss": 10.822088241577148, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3697982728481293, + "loss_total_v6": 0.5533920526504517, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1018 + }, + { + "epoch": 0.3797987327618338, + "grad_norm": 17.81587028503418, + "learning_rate": 7.547330408197695e-06, + "loss": 10.803271293640137, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.447123646736145, + "loss_total_v6": 0.646342396736145, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 1019 + }, + { + "epoch": 0.38017144986954904, + "grad_norm": 17.89579200744629, + "learning_rate": 7.541884731207707e-06, + "loss": 11.239049911499023, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.34276676177978516, + "loss_total_v6": 0.5154962539672852, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1020 + }, + { + "epoch": 0.38054416697726423, + "grad_norm": 19.82714080810547, + "learning_rate": 7.536434985082696e-06, + "loss": 11.156023979187012, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3020428717136383, + "loss_total_v6": 0.4797772467136383, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1021 + }, + { + "epoch": 0.3809168840849795, + "grad_norm": 17.90571403503418, + "learning_rate": 7.5309811785468035e-06, + "loss": 10.200872421264648, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.423387736082077, + "loss_total_v6": 0.6245596408843994, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 1022 + }, + { + "epoch": 0.3812896011926947, + "grad_norm": 18.3035888671875, + "learning_rate": 7.5255233203306665e-06, + "loss": 11.002962112426758, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.39752551913261414, + "loss_total_v6": 0.5860021114349365, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 1023 + }, + { + "epoch": 0.38166231830041, + "grad_norm": 18.325349807739258, + "learning_rate": 7.520061419171407e-06, + "loss": 10.95720100402832, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5462340116500854, + "loss_total_v6": 0.7339781522750854, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1024 + }, + { + "epoch": 0.3820350354081252, + "grad_norm": 17.485700607299805, + "learning_rate": 7.514595483812624e-06, + "loss": 10.562309265136719, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.1815308779478073, + "loss_total_v6": 0.3565797209739685, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1025 + }, + { + "epoch": 0.38240775251584047, + "grad_norm": 19.04399299621582, + "learning_rate": 7.509125523004368e-06, + "loss": 10.696345329284668, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5924531817436218, + "loss_total_v6": 0.7700654864311218, + "loss_wm_recon": 0.12451171875, + "loss_wm_recon_w": 0.0311279296875, + "step": 1026 + }, + { + "epoch": 0.3827804696235557, + "grad_norm": 18.01698875427246, + "learning_rate": 7.5036515455031415e-06, + "loss": 11.057796478271484, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5074198246002197, + "loss_total_v6": 0.7005350589752197, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 1027 + }, + { + "epoch": 0.38315318673127097, + "grad_norm": 18.140588760375977, + "learning_rate": 7.4981735600718685e-06, + "loss": 10.539718627929688, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6570286750793457, + "loss_total_v6": 0.8408665657043457, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1028 + }, + { + "epoch": 0.3835259038389862, + "grad_norm": 16.9239559173584, + "learning_rate": 7.492691575479896e-06, + "loss": 10.414992332458496, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.20103199779987335, + "loss_total_v6": 0.39927417039871216, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 1029 + }, + { + "epoch": 0.38389862094670146, + "grad_norm": 17.221782684326172, + "learning_rate": 7.48720560050297e-06, + "loss": 11.069683074951172, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.7618365287780762, + "loss_total_v6": 0.9766802787780762, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 1030 + }, + { + "epoch": 0.3842713380544167, + "grad_norm": 17.858583450317383, + "learning_rate": 7.481715643923225e-06, + "loss": 10.764959335327148, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6001173853874207, + "loss_total_v6": 0.7835890650749207, + "loss_wm_recon": 0.12451171875, + "loss_wm_recon_w": 0.0311279296875, + "step": 1031 + }, + { + "epoch": 0.38464405516213196, + "grad_norm": 18.058286666870117, + "learning_rate": 7.476221714529167e-06, + "loss": 11.0145263671875, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7994996309280396, + "loss_total_v6": 0.9943238496780396, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 1032 + }, + { + "epoch": 0.3850167722698472, + "grad_norm": 17.99306869506836, + "learning_rate": 7.470723821115666e-06, + "loss": 10.97419548034668, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5032644867897034, + "loss_total_v6": 0.6878347992897034, + "loss_wm_recon": 0.15234375, + "loss_wm_recon_w": 0.0380859375, + "step": 1033 + }, + { + "epoch": 0.38538948937756246, + "grad_norm": 17.163652420043945, + "learning_rate": 7.465221972483936e-06, + "loss": 10.681909561157227, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.5510852336883545, + "loss_total_v6": 0.7546985149383545, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 1034 + }, + { + "epoch": 0.38576220648527765, + "grad_norm": 17.099172592163086, + "learning_rate": 7.45971617744152e-06, + "loss": 10.555244445800781, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.396084189414978, + "loss_total_v6": 0.573086142539978, + "loss_wm_recon": 0.1220703125, + "loss_wm_recon_w": 0.030517578125, + "step": 1035 + }, + { + "epoch": 0.3861349235929929, + "grad_norm": 17.132755279541016, + "learning_rate": 7.454206444802283e-06, + "loss": 10.602062225341797, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.3010954260826111, + "loss_total_v6": 0.4768766760826111, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1036 + }, + { + "epoch": 0.38650764070070814, + "grad_norm": 17.230493545532227, + "learning_rate": 7.448692783386391e-06, + "loss": 10.736900329589844, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.11493906378746033, + "loss_total_v6": 0.2987769544124603, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1037 + }, + { + "epoch": 0.3868803578084234, + "grad_norm": 17.668285369873047, + "learning_rate": 7.443175202020297e-06, + "loss": 10.840531349182129, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.1537497490644455, + "loss_total_v6": 0.3279440999031067, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1038 + }, + { + "epoch": 0.38725307491613864, + "grad_norm": 18.917800903320312, + "learning_rate": 7.437653709536735e-06, + "loss": 10.658489227294922, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6390030384063721, + "loss_total_v6": 0.8338272571563721, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 1039 + }, + { + "epoch": 0.3876257920238539, + "grad_norm": 17.233768463134766, + "learning_rate": 7.432128314774696e-06, + "loss": 10.728355407714844, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4506365656852722, + "loss_total_v6": 0.6317889094352722, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1040 + }, + { + "epoch": 0.38799850913156914, + "grad_norm": 17.037324905395508, + "learning_rate": 7.426599026579419e-06, + "loss": 10.860797882080078, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4775795340538025, + "loss_total_v6": 0.6709389090538025, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 1041 + }, + { + "epoch": 0.3883712262392844, + "grad_norm": 18.8544921875, + "learning_rate": 7.421065853802375e-06, + "loss": 10.916449546813965, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.3355955481529236, + "loss_total_v6": 0.5093016028404236, + "loss_wm_recon": 0.12451171875, + "loss_wm_recon_w": 0.0311279296875, + "step": 1042 + }, + { + "epoch": 0.38874394334699963, + "grad_norm": 17.302427291870117, + "learning_rate": 7.415528805301254e-06, + "loss": 10.7283935546875, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.36402463912963867, + "loss_total_v6": 0.5534777641296387, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 1043 + }, + { + "epoch": 0.3891166604547149, + "grad_norm": 18.61834716796875, + "learning_rate": 7.4099878899399506e-06, + "loss": 10.912252426147461, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4086034595966339, + "loss_total_v6": 0.5975682735443115, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 1044 + }, + { + "epoch": 0.38948937756243013, + "grad_norm": 18.104869842529297, + "learning_rate": 7.404443116588548e-06, + "loss": 10.572248458862305, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.569968044757843, + "loss_total_v6": 0.752585232257843, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1045 + }, + { + "epoch": 0.3898620946701454, + "grad_norm": 17.18991470336914, + "learning_rate": 7.3988944941233095e-06, + "loss": 10.691567420959473, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.8006792664527893, + "loss_total_v6": 0.9991655945777893, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1046 + }, + { + "epoch": 0.3902348117778606, + "grad_norm": 17.275423049926758, + "learning_rate": 7.393342031426656e-06, + "loss": 10.369037628173828, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.8386952877044678, + "loss_total_v6": 1.0222890377044678, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 1047 + }, + { + "epoch": 0.39060752888557587, + "grad_norm": 17.063764572143555, + "learning_rate": 7.3877857373871585e-06, + "loss": 10.341196060180664, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.37883836030960083, + "loss_total_v6": 0.5366752743721008, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1048 + }, + { + "epoch": 0.39098024599329106, + "grad_norm": 17.669897079467773, + "learning_rate": 7.382225620899519e-06, + "loss": 10.52692985534668, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.1629037708044052, + "loss_total_v6": 0.3494272232055664, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 1049 + }, + { + "epoch": 0.3913529631010063, + "grad_norm": 17.26572036743164, + "learning_rate": 7.37666169086456e-06, + "loss": 10.929330825805664, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4124954342842102, + "loss_total_v6": 0.5969436764717102, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1050 + }, + { + "epoch": 0.39172568020872156, + "grad_norm": 17.95763397216797, + "learning_rate": 7.371093956189211e-06, + "loss": 10.725479125976562, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.8407689929008484, + "loss_total_v6": 1.0270483493804932, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1051 + }, + { + "epoch": 0.3920983973164368, + "grad_norm": 16.972267150878906, + "learning_rate": 7.365522425786487e-06, + "loss": 10.87490177154541, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5525348782539368, + "loss_total_v6": 0.7468708157539368, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 1052 + }, + { + "epoch": 0.39247111442415206, + "grad_norm": 17.869823455810547, + "learning_rate": 7.359947108575485e-06, + "loss": 10.844260215759277, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.06734317541122437, + "loss_total_v6": 0.24629825353622437, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1053 + }, + { + "epoch": 0.3928438315318673, + "grad_norm": 16.644657135009766, + "learning_rate": 7.354368013481359e-06, + "loss": 10.739401817321777, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.45732051134109497, + "loss_total_v6": 0.642867386341095, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 1054 + }, + { + "epoch": 0.39321654863958255, + "grad_norm": 17.27267074584961, + "learning_rate": 7.348785149435315e-06, + "loss": 10.498188018798828, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.7244605422019958, + "loss_total_v6": 0.9051246047019958, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1055 + }, + { + "epoch": 0.3935892657472978, + "grad_norm": 17.267332077026367, + "learning_rate": 7.343198525374588e-06, + "loss": 10.709625244140625, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5259100198745728, + "loss_total_v6": 0.7032781839370728, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1056 + }, + { + "epoch": 0.39396198285501305, + "grad_norm": 17.15097427368164, + "learning_rate": 7.337608150242438e-06, + "loss": 10.750032424926758, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5989430546760559, + "loss_total_v6": 0.7774098515510559, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1057 + }, + { + "epoch": 0.3943346999627283, + "grad_norm": 17.599231719970703, + "learning_rate": 7.332014032988123e-06, + "loss": 11.069411277770996, + "loss_alignment": 0.3984375, + "loss_alignment_w": 0.19921875, + "loss_sft": 0.5563538670539856, + "loss_total_v6": 0.8044007420539856, + "loss_wm_recon": 0.1953125, + "loss_wm_recon_w": 0.048828125, + "step": 1058 + }, + { + "epoch": 0.39470741707044354, + "grad_norm": 17.66929054260254, + "learning_rate": 7.326416182566897e-06, + "loss": 10.532222747802734, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.41473227739334106, + "loss_total_v6": 0.5982039570808411, + "loss_wm_recon": 0.12451171875, + "loss_wm_recon_w": 0.0311279296875, + "step": 1059 + }, + { + "epoch": 0.3950801341781588, + "grad_norm": 19.827360153198242, + "learning_rate": 7.3208146079399885e-06, + "loss": 10.55732250213623, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.457489013671875, + "loss_total_v6": 0.635589599609375, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1060 + }, + { + "epoch": 0.39545285128587404, + "grad_norm": 17.405153274536133, + "learning_rate": 7.315209318074587e-06, + "loss": 10.564067840576172, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5018121004104614, + "loss_total_v6": 0.7032281160354614, + "loss_wm_recon": 0.1572265625, + "loss_wm_recon_w": 0.039306640625, + "step": 1061 + }, + { + "epoch": 0.3958255683935893, + "grad_norm": 16.7806453704834, + "learning_rate": 7.30960032194383e-06, + "loss": 10.53411865234375, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5998557806015015, + "loss_total_v6": 0.6347678899765015, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 1062 + }, + { + "epoch": 0.39619828550130454, + "grad_norm": 17.581684112548828, + "learning_rate": 7.303987628526789e-06, + "loss": 10.421403884887695, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.4526228606700897, + "loss_total_v6": 0.6237654685974121, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1063 + }, + { + "epoch": 0.39657100260901973, + "grad_norm": 17.281272888183594, + "learning_rate": 7.298371246808451e-06, + "loss": 10.645332336425781, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6481767892837524, + "loss_total_v6": 0.8361650705337524, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 1064 + }, + { + "epoch": 0.396943719716735, + "grad_norm": 17.509071350097656, + "learning_rate": 7.292751185779713e-06, + "loss": 11.124850273132324, + "loss_alignment": 0.359375, + "loss_alignment_w": 0.1796875, + "loss_sft": 0.7903751134872437, + "loss_total_v6": 1.0152286291122437, + "loss_wm_recon": 0.1806640625, + "loss_wm_recon_w": 0.045166015625, + "step": 1065 + }, + { + "epoch": 0.3973164368244502, + "grad_norm": 17.02750587463379, + "learning_rate": 7.287127454437356e-06, + "loss": 10.51187801361084, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.16681279242038727, + "loss_total_v6": 0.3451575040817261, + "loss_wm_recon": 0.11962890625, + "loss_wm_recon_w": 0.0299072265625, + "step": 1066 + }, + { + "epoch": 0.3976891539321655, + "grad_norm": 16.983491897583008, + "learning_rate": 7.2815000617840406e-06, + "loss": 10.576519012451172, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.4231069087982178, + "loss_total_v6": 0.5965688228607178, + "loss_wm_recon": 0.12353515625, + "loss_wm_recon_w": 0.0308837890625, + "step": 1067 + }, + { + "epoch": 0.3980618710398807, + "grad_norm": 17.1857852935791, + "learning_rate": 7.275869016828288e-06, + "loss": 10.95931625366211, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4963865876197815, + "loss_total_v6": 0.6765623688697815, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 1068 + }, + { + "epoch": 0.39843458814759597, + "grad_norm": 17.494976043701172, + "learning_rate": 7.270234328584463e-06, + "loss": 11.097122192382812, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.42680832743644714, + "loss_total_v6": 0.6028337478637695, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1069 + }, + { + "epoch": 0.3988073052553112, + "grad_norm": 17.74421501159668, + "learning_rate": 7.264596006072766e-06, + "loss": 10.98193645477295, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.3979429304599762, + "loss_total_v6": 0.5745786428451538, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1070 + }, + { + "epoch": 0.39918002236302647, + "grad_norm": 18.0059871673584, + "learning_rate": 7.2589540583192165e-06, + "loss": 10.263498306274414, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5394068956375122, + "loss_total_v6": 0.7281275987625122, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1071 + }, + { + "epoch": 0.3995527394707417, + "grad_norm": 16.84761619567871, + "learning_rate": 7.253308494355633e-06, + "loss": 10.798933982849121, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.38519465923309326, + "loss_total_v6": 0.5687884092330933, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 1072 + }, + { + "epoch": 0.39992545657845696, + "grad_norm": 17.86505699157715, + "learning_rate": 7.247659323219624e-06, + "loss": 10.855916976928711, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.4501543641090393, + "loss_total_v6": 0.6226397156715393, + "loss_wm_recon": 0.11962890625, + "loss_wm_recon_w": 0.0299072265625, + "step": 1073 + }, + { + "epoch": 0.4002981736861722, + "grad_norm": 17.316646575927734, + "learning_rate": 7.242006553954577e-06, + "loss": 10.495223045349121, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.4257664382457733, + "loss_total_v6": 0.5919041633605957, + "loss_wm_recon": 0.10986328125, + "loss_wm_recon_w": 0.0274658203125, + "step": 1074 + }, + { + "epoch": 0.40067089079388746, + "grad_norm": 18.099437713623047, + "learning_rate": 7.236350195609634e-06, + "loss": 10.749977111816406, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.4406420886516571, + "loss_total_v6": 0.6786792278289795, + "loss_wm_recon": 0.2177734375, + "loss_wm_recon_w": 0.054443359375, + "step": 1075 + }, + { + "epoch": 0.4010436079016027, + "grad_norm": 16.80180549621582, + "learning_rate": 7.230690257239684e-06, + "loss": 10.566606521606445, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.4998282194137573, + "loss_total_v6": 0.7031973600387573, + "loss_wm_recon": 0.1416015625, + "loss_wm_recon_w": 0.035400390625, + "step": 1076 + }, + { + "epoch": 0.40141632500931795, + "grad_norm": 16.935218811035156, + "learning_rate": 7.22502674790535e-06, + "loss": 10.589521408081055, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.629986584186554, + "loss_total_v6": 0.826275646686554, + "loss_wm_recon": 0.16015625, + "loss_wm_recon_w": 0.0400390625, + "step": 1077 + }, + { + "epoch": 0.40178904211703315, + "grad_norm": 17.253223419189453, + "learning_rate": 7.219359676672969e-06, + "loss": 10.640119552612305, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.4840894937515259, + "loss_total_v6": 0.6571851968765259, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1078 + }, + { + "epoch": 0.4021617592247484, + "grad_norm": 17.65083885192871, + "learning_rate": 7.213689052614576e-06, + "loss": 10.973089218139648, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.3107328414916992, + "loss_total_v6": 0.4950590133666992, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1079 + }, + { + "epoch": 0.40253447633246364, + "grad_norm": 17.587188720703125, + "learning_rate": 7.208014884807903e-06, + "loss": 10.766681671142578, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.38978204131126404, + "loss_total_v6": 0.5636101961135864, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1080 + }, + { + "epoch": 0.4029071934401789, + "grad_norm": 16.936994552612305, + "learning_rate": 7.20233718233635e-06, + "loss": 10.61393928527832, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4320824146270752, + "loss_total_v6": 0.6227562427520752, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1081 + }, + { + "epoch": 0.40327991054789414, + "grad_norm": 17.33666229248047, + "learning_rate": 7.196655954288969e-06, + "loss": 10.434938430786133, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.2926982045173645, + "loss_total_v6": 0.4833720326423645, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1082 + }, + { + "epoch": 0.4036526276556094, + "grad_norm": 19.713754653930664, + "learning_rate": 7.190971209760469e-06, + "loss": 10.604351043701172, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.7000823616981506, + "loss_total_v6": 0.8895354866981506, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 1083 + }, + { + "epoch": 0.40402534476332463, + "grad_norm": 18.41352653503418, + "learning_rate": 7.185282957851175e-06, + "loss": 11.059188842773438, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.23811258375644684, + "loss_total_v6": 0.41963112354278564, + "loss_wm_recon": 0.11669921875, + "loss_wm_recon_w": 0.0291748046875, + "step": 1084 + }, + { + "epoch": 0.4043980618710399, + "grad_norm": 17.654266357421875, + "learning_rate": 7.179591207667038e-06, + "loss": 10.700677871704102, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.47613275051116943, + "loss_total_v6": 0.6556981801986694, + "loss_wm_recon": 0.12451171875, + "loss_wm_recon_w": 0.0311279296875, + "step": 1085 + }, + { + "epoch": 0.40477077897875513, + "grad_norm": 16.718372344970703, + "learning_rate": 7.173895968319601e-06, + "loss": 10.889939308166504, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.6287546753883362, + "loss_total_v6": 0.8533640503883362, + "loss_wm_recon": 0.171875, + "loss_wm_recon_w": 0.04296875, + "step": 1086 + }, + { + "epoch": 0.4051434960864704, + "grad_norm": 17.047964096069336, + "learning_rate": 7.168197248925996e-06, + "loss": 10.431999206542969, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.7042214870452881, + "loss_total_v6": 0.9085671901702881, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 1087 + }, + { + "epoch": 0.4055162131941856, + "grad_norm": 16.969526290893555, + "learning_rate": 7.162495058608928e-06, + "loss": 10.482378005981445, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.24573001265525818, + "loss_total_v6": 0.4315210282802582, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1088 + }, + { + "epoch": 0.4058889303019009, + "grad_norm": 17.085859298706055, + "learning_rate": 7.156789406496655e-06, + "loss": 10.68331527709961, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.2849806845188141, + "loss_total_v6": 0.4715041220188141, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1089 + }, + { + "epoch": 0.4062616474096161, + "grad_norm": 18.111127853393555, + "learning_rate": 7.151080301722976e-06, + "loss": 10.405298233032227, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5381080508232117, + "loss_total_v6": 0.7153541445732117, + "loss_wm_recon": 0.123046875, + "loss_wm_recon_w": 0.03076171875, + "step": 1090 + }, + { + "epoch": 0.40663436451733137, + "grad_norm": 17.007282257080078, + "learning_rate": 7.14536775342722e-06, + "loss": 10.687498092651367, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6613912582397461, + "loss_total_v6": 0.8615865707397461, + "loss_wm_recon": 0.15234375, + "loss_wm_recon_w": 0.0380859375, + "step": 1091 + }, + { + "epoch": 0.40700708162504656, + "grad_norm": 17.434139251708984, + "learning_rate": 7.13965177075423e-06, + "loss": 10.461121559143066, + "loss_alignment": 0.25390625, + "loss_alignment_w": 0.126953125, + "loss_sft": 0.5172756910324097, + "loss_total_v6": 0.6699856519699097, + "loss_wm_recon": 0.10302734375, + "loss_wm_recon_w": 0.0257568359375, + "step": 1092 + }, + { + "epoch": 0.4073797987327618, + "grad_norm": 16.454137802124023, + "learning_rate": 7.1339323628543385e-06, + "loss": 9.902090072631836, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6585948467254639, + "loss_total_v6": 0.8412120342254639, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1093 + }, + { + "epoch": 0.40775251584047706, + "grad_norm": 16.94434928894043, + "learning_rate": 7.128209538883373e-06, + "loss": 10.207961082458496, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4592481851577759, + "loss_total_v6": 0.6352735757827759, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1094 + }, + { + "epoch": 0.4081252329481923, + "grad_norm": 22.430570602416992, + "learning_rate": 7.1224833080026216e-06, + "loss": 10.44774055480957, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4513774812221527, + "loss_total_v6": 0.4796977937221527, + "loss_wm_recon": 0.11328125, + "loss_wm_recon_w": 0.0283203125, + "step": 1095 + }, + { + "epoch": 0.40849795005590755, + "grad_norm": 17.776992797851562, + "learning_rate": 7.116753679378826e-06, + "loss": 10.605430603027344, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.10921391099691391, + "loss_total_v6": 0.2809668481349945, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1096 + }, + { + "epoch": 0.4088706671636228, + "grad_norm": 18.406383514404297, + "learning_rate": 7.111020662184174e-06, + "loss": 10.679521560668945, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4033612608909607, + "loss_total_v6": 0.5781659483909607, + "loss_wm_recon": 0.12109375, + "loss_wm_recon_w": 0.0302734375, + "step": 1097 + }, + { + "epoch": 0.40924338427133805, + "grad_norm": 17.44175910949707, + "learning_rate": 7.105284265596269e-06, + "loss": 10.829161643981934, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.33517318964004517, + "loss_total_v6": 0.5225511193275452, + "loss_wm_recon": 0.12451171875, + "loss_wm_recon_w": 0.0311279296875, + "step": 1098 + }, + { + "epoch": 0.4096161013790533, + "grad_norm": 17.06223487854004, + "learning_rate": 7.09954449879813e-06, + "loss": 10.674068450927734, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.4519619345664978, + "loss_total_v6": 0.6487392783164978, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1099 + }, + { + "epoch": 0.40998881848676855, + "grad_norm": 17.722970962524414, + "learning_rate": 7.093801370978171e-06, + "loss": 10.872432708740234, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.2000703662633896, + "loss_total_v6": 0.3830537796020508, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1100 + }, + { + "epoch": 0.40998881848676855, + "eval_loss": 0.6466321349143982, + "eval_loss_alignment": 0.28267519698183763, + "eval_loss_alignment_w": 0.14133759849091881, + "eval_loss_sft": 0.4751605675515965, + "eval_loss_total_v6": 0.6466321874187033, + "eval_loss_wm_recon": 0.12053608690571581, + "eval_loss_wm_recon_w": 0.030134021726428952, + "eval_runtime": 228.4065, + "eval_samples_per_second": 8.17, + "eval_steps_per_second": 1.024, + "step": 1100 + }, + { + "epoch": 0.4103615355944838, + "grad_norm": 18.067495346069336, + "learning_rate": 7.0880548913301835e-06, + "loss": 10.84392261505127, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.27179720997810364, + "loss_total_v6": 0.44086459279060364, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1101 + }, + { + "epoch": 0.41073425270219904, + "grad_norm": 18.116464614868164, + "learning_rate": 7.082305069053328e-06, + "loss": 11.221631050109863, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5622119307518005, + "loss_total_v6": 0.7509326338768005, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1102 + }, + { + "epoch": 0.4111069698099143, + "grad_norm": 17.153108596801758, + "learning_rate": 7.076551913352113e-06, + "loss": 10.315801620483398, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.3278060853481293, + "loss_total_v6": 0.5201889276504517, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1103 + }, + { + "epoch": 0.41147968691762954, + "grad_norm": 17.3565616607666, + "learning_rate": 7.070795433436386e-06, + "loss": 10.473627090454102, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4349484443664551, + "loss_total_v6": 0.6082882881164551, + "loss_wm_recon": 0.115234375, + "loss_wm_recon_w": 0.02880859375, + "step": 1104 + }, + { + "epoch": 0.4118524040253448, + "grad_norm": 17.490867614746094, + "learning_rate": 7.065035638521312e-06, + "loss": 10.734489440917969, + "loss_alignment": 0.37109375, + "loss_alignment_w": 0.185546875, + "loss_sft": 0.45365697145462036, + "loss_total_v6": 0.6860788464546204, + "loss_wm_recon": 0.1875, + "loss_wm_recon_w": 0.046875, + "step": 1105 + }, + { + "epoch": 0.41222512113306004, + "grad_norm": 17.776256561279297, + "learning_rate": 7.05927253782737e-06, + "loss": 10.572968482971191, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5885478854179382, + "loss_total_v6": 0.7626201510429382, + "loss_wm_recon": 0.1181640625, + "loss_wm_recon_w": 0.029541015625, + "step": 1106 + }, + { + "epoch": 0.4125978382407752, + "grad_norm": 22.0122013092041, + "learning_rate": 7.053506140580322e-06, + "loss": 10.265579223632812, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.46448010206222534, + "loss_total_v6": 0.6396510004997253, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1107 + }, + { + "epoch": 0.4129705553484905, + "grad_norm": 17.990718841552734, + "learning_rate": 7.0477364560112134e-06, + "loss": 10.265291213989258, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.6278895735740662, + "loss_total_v6": 0.7989100813865662, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1108 + }, + { + "epoch": 0.4133432724562057, + "grad_norm": 18.492900848388672, + "learning_rate": 7.041963493356352e-06, + "loss": 10.9779634475708, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5757163763046265, + "loss_total_v6": 0.7612632513046265, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1109 + }, + { + "epoch": 0.41371598956392097, + "grad_norm": 16.87782096862793, + "learning_rate": 7.036187261857289e-06, + "loss": 10.569257736206055, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.60328209400177, + "loss_total_v6": 0.78760826587677, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1110 + }, + { + "epoch": 0.4140887066716362, + "grad_norm": 17.14132308959961, + "learning_rate": 7.030407770760812e-06, + "loss": 10.804489135742188, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.8248952627182007, + "loss_total_v6": 1.0072683095932007, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1111 + }, + { + "epoch": 0.41446142377935147, + "grad_norm": 17.41312026977539, + "learning_rate": 7.024625029318927e-06, + "loss": 10.72785758972168, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5661876201629639, + "loss_total_v6": 0.7533214092254639, + "loss_wm_recon": 0.12353515625, + "loss_wm_recon_w": 0.0308837890625, + "step": 1112 + }, + { + "epoch": 0.4148341408870667, + "grad_norm": 17.430736541748047, + "learning_rate": 7.018839046788839e-06, + "loss": 10.5933837890625, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6968739032745361, + "loss_total_v6": 0.7305653095245361, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 1113 + }, + { + "epoch": 0.41520685799478196, + "grad_norm": 16.66971778869629, + "learning_rate": 7.013049832432947e-06, + "loss": 10.499914169311523, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6179338097572327, + "loss_total_v6": 0.8100724816322327, + "loss_wm_recon": 0.1435546875, + "loss_wm_recon_w": 0.035888671875, + "step": 1114 + }, + { + "epoch": 0.4155795751024972, + "grad_norm": 17.956628799438477, + "learning_rate": 7.007257395518822e-06, + "loss": 11.064435958862305, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.734413206577301, + "loss_total_v6": 0.915565550327301, + "loss_wm_recon": 0.123046875, + "loss_wm_recon_w": 0.03076171875, + "step": 1115 + }, + { + "epoch": 0.41595229221021246, + "grad_norm": 17.63219451904297, + "learning_rate": 7.001461745319189e-06, + "loss": 10.24232292175293, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.49864664673805237, + "loss_total_v6": 0.67686927318573, + "loss_wm_recon": 0.119140625, + "loss_wm_recon_w": 0.02978515625, + "step": 1116 + }, + { + "epoch": 0.4163250093179277, + "grad_norm": 16.777606964111328, + "learning_rate": 6.995662891111925e-06, + "loss": 10.201018333435059, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.8485682010650635, + "loss_total_v6": 1.0336267948150635, + "loss_wm_recon": 0.123046875, + "loss_wm_recon_w": 0.03076171875, + "step": 1117 + }, + { + "epoch": 0.41669772642564296, + "grad_norm": 16.64278221130371, + "learning_rate": 6.98986084218003e-06, + "loss": 10.555973052978516, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.9403152465820312, + "loss_total_v6": 1.1126785278320312, + "loss_wm_recon": 0.119140625, + "loss_wm_recon_w": 0.02978515625, + "step": 1118 + }, + { + "epoch": 0.4170704435333582, + "grad_norm": 18.167190551757812, + "learning_rate": 6.9840556078116195e-06, + "loss": 10.901226997375488, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.34884801506996155, + "loss_total_v6": 0.5226761102676392, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1119 + }, + { + "epoch": 0.41744316064107345, + "grad_norm": 17.493427276611328, + "learning_rate": 6.97824719729991e-06, + "loss": 10.342931747436523, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5800737142562866, + "loss_total_v6": 0.7629350423812866, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1120 + }, + { + "epoch": 0.41781587774878864, + "grad_norm": 16.819358825683594, + "learning_rate": 6.972435619943198e-06, + "loss": 10.615025520324707, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5114638805389404, + "loss_total_v6": 0.6953017711639404, + "loss_wm_recon": 0.1181640625, + "loss_wm_recon_w": 0.029541015625, + "step": 1121 + }, + { + "epoch": 0.4181885948565039, + "grad_norm": 16.890634536743164, + "learning_rate": 6.9666208850448595e-06, + "loss": 10.467304229736328, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.23843981325626373, + "loss_total_v6": 0.41263413429260254, + "loss_wm_recon": 0.11083984375, + "loss_wm_recon_w": 0.0277099609375, + "step": 1122 + }, + { + "epoch": 0.41856131196421914, + "grad_norm": 17.0390625, + "learning_rate": 6.960803001913315e-06, + "loss": 10.592053413391113, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4340078830718994, + "loss_total_v6": 0.6200430393218994, + "loss_wm_recon": 0.119140625, + "loss_wm_recon_w": 0.02978515625, + "step": 1123 + }, + { + "epoch": 0.4189340290719344, + "grad_norm": 16.09562110900879, + "learning_rate": 6.954981979862027e-06, + "loss": 10.235292434692383, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.42263248562812805, + "loss_total_v6": 0.5842535495758057, + "loss_wm_recon": 0.115234375, + "loss_wm_recon_w": 0.02880859375, + "step": 1124 + }, + { + "epoch": 0.41930674617964964, + "grad_norm": 16.517051696777344, + "learning_rate": 6.9491578282094876e-06, + "loss": 10.812651634216309, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4855557084083557, + "loss_total_v6": 0.6651211380958557, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1125 + }, + { + "epoch": 0.4196794632873649, + "grad_norm": 17.1707763671875, + "learning_rate": 6.9433305562791935e-06, + "loss": 10.461395263671875, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5584352016448975, + "loss_total_v6": 0.7405641078948975, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 1126 + }, + { + "epoch": 0.42005218039508013, + "grad_norm": 17.002899169921875, + "learning_rate": 6.937500173399639e-06, + "loss": 10.63796329498291, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5583993196487427, + "loss_total_v6": 0.7495614290237427, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 1127 + }, + { + "epoch": 0.4204248975027954, + "grad_norm": 17.14313316345215, + "learning_rate": 6.9316666889042995e-06, + "loss": 10.395846366882324, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.34829795360565186, + "loss_total_v6": 0.5232247114181519, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 1128 + }, + { + "epoch": 0.42079761461051063, + "grad_norm": 17.970582962036133, + "learning_rate": 6.925830112131613e-06, + "loss": 10.706218719482422, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.3321184813976288, + "loss_total_v6": 0.5201067924499512, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 1129 + }, + { + "epoch": 0.4211703317182259, + "grad_norm": 19.34734344482422, + "learning_rate": 6.9199904524249695e-06, + "loss": 10.119560241699219, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.3963063359260559, + "loss_total_v6": 0.5833180546760559, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1130 + }, + { + "epoch": 0.4215430488259411, + "grad_norm": 17.89679527282715, + "learning_rate": 6.914147719132692e-06, + "loss": 10.536027908325195, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5148436427116394, + "loss_total_v6": 0.5482909083366394, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1131 + }, + { + "epoch": 0.4219157659336564, + "grad_norm": 18.201383590698242, + "learning_rate": 6.908301921608028e-06, + "loss": 10.039864540100098, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.554750919342041, + "loss_total_v6": 0.723940372467041, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1132 + }, + { + "epoch": 0.4222884830413716, + "grad_norm": 17.907278060913086, + "learning_rate": 6.902453069209128e-06, + "loss": 10.770058631896973, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2690178453922272, + "loss_total_v6": 0.4499260485172272, + "loss_wm_recon": 0.1220703125, + "loss_wm_recon_w": 0.030517578125, + "step": 1133 + }, + { + "epoch": 0.42266120014908687, + "grad_norm": 17.65868377685547, + "learning_rate": 6.8966011712990275e-06, + "loss": 10.60545539855957, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.1959073841571808, + "loss_total_v6": 0.3736417591571808, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1134 + }, + { + "epoch": 0.42303391725680206, + "grad_norm": 18.662405014038086, + "learning_rate": 6.890746237245649e-06, + "loss": 10.551050186157227, + "loss_alignment": 0.359375, + "loss_alignment_w": 0.1796875, + "loss_sft": 0.7526400089263916, + "loss_total_v6": 0.9701693058013916, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 1135 + }, + { + "epoch": 0.4234066343645173, + "grad_norm": 16.418384552001953, + "learning_rate": 6.884888276421766e-06, + "loss": 10.319633483886719, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.40709832310676575, + "loss_total_v6": 0.5892271995544434, + "loss_wm_recon": 0.119140625, + "loss_wm_recon_w": 0.02978515625, + "step": 1136 + }, + { + "epoch": 0.42377935147223256, + "grad_norm": 17.125146865844727, + "learning_rate": 6.879027298205001e-06, + "loss": 10.800848960876465, + "loss_alignment": 0.265625, + "loss_alignment_w": 0.1328125, + "loss_sft": 0.4224453866481781, + "loss_total_v6": 0.5851651430130005, + "loss_wm_recon": 0.11962890625, + "loss_wm_recon_w": 0.0299072265625, + "step": 1137 + }, + { + "epoch": 0.4241520685799478, + "grad_norm": 17.738359451293945, + "learning_rate": 6.873163311977805e-06, + "loss": 10.310513496398926, + "loss_alignment": 0.2578125, + "loss_alignment_w": 0.12890625, + "loss_sft": 0.2952086627483368, + "loss_total_v6": 0.4545104205608368, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 1138 + }, + { + "epoch": 0.42452478568766305, + "grad_norm": 16.50423240661621, + "learning_rate": 6.8672963271274465e-06, + "loss": 10.852043151855469, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4483751952648163, + "loss_total_v6": 0.6327013969421387, + "loss_wm_recon": 0.1201171875, + "loss_wm_recon_w": 0.030029296875, + "step": 1139 + }, + { + "epoch": 0.4248975027953783, + "grad_norm": 17.977075576782227, + "learning_rate": 6.8614263530459925e-06, + "loss": 10.517598152160645, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.250644713640213, + "loss_total_v6": 0.422031432390213, + "loss_wm_recon": 0.107421875, + "loss_wm_recon_w": 0.02685546875, + "step": 1140 + }, + { + "epoch": 0.42527021990309355, + "grad_norm": 20.868741989135742, + "learning_rate": 6.855553399130296e-06, + "loss": 10.744958877563477, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.40119969844818115, + "loss_total_v6": 0.5669711828231812, + "loss_wm_recon": 0.1083984375, + "loss_wm_recon_w": 0.027099609375, + "step": 1141 + }, + { + "epoch": 0.4256429370108088, + "grad_norm": 18.19587516784668, + "learning_rate": 6.849677474781982e-06, + "loss": 10.657075881958008, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.25219476222991943, + "loss_total_v6": 0.42065179347991943, + "loss_wm_recon": 0.103515625, + "loss_wm_recon_w": 0.02587890625, + "step": 1142 + }, + { + "epoch": 0.42601565411852405, + "grad_norm": 17.28180503845215, + "learning_rate": 6.843798589407424e-06, + "loss": 10.295797348022461, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6149150729179382, + "loss_total_v6": 0.8097392916679382, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1143 + }, + { + "epoch": 0.4263883712262393, + "grad_norm": 17.259660720825195, + "learning_rate": 6.837916752417748e-06, + "loss": 10.33930492401123, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6606144905090332, + "loss_total_v6": 0.8392033576965332, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1144 + }, + { + "epoch": 0.42676108833395454, + "grad_norm": 17.279277801513672, + "learning_rate": 6.83203197322879e-06, + "loss": 10.496809005737305, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.7453117966651917, + "loss_total_v6": 0.9263420701026917, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1145 + }, + { + "epoch": 0.4271338054416698, + "grad_norm": 16.86248016357422, + "learning_rate": 6.826144261261108e-06, + "loss": 11.00155258178711, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.29571184515953064, + "loss_total_v6": 0.504696249961853, + "loss_wm_recon": 0.1640625, + "loss_wm_recon_w": 0.041015625, + "step": 1146 + }, + { + "epoch": 0.42750652254938504, + "grad_norm": 17.39400291442871, + "learning_rate": 6.82025362593995e-06, + "loss": 10.574249267578125, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6635074019432068, + "loss_total_v6": 0.8590640425682068, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1147 + }, + { + "epoch": 0.4278792396571003, + "grad_norm": 16.858047485351562, + "learning_rate": 6.814360076695242e-06, + "loss": 10.794854164123535, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6358484625816345, + "loss_total_v6": 0.8274988532066345, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1148 + }, + { + "epoch": 0.4282519567648155, + "grad_norm": 18.053466796875, + "learning_rate": 6.8084636229615786e-06, + "loss": 10.725019454956055, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.3042812645435333, + "loss_total_v6": 0.4713955223560333, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1149 + }, + { + "epoch": 0.4286246738725307, + "grad_norm": 16.335126876831055, + "learning_rate": 6.802564274178203e-06, + "loss": 10.700441360473633, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5697767734527588, + "loss_total_v6": 0.7527601718902588, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1150 + }, + { + "epoch": 0.428997390980246, + "grad_norm": 16.570585250854492, + "learning_rate": 6.7966620397889906e-06, + "loss": 10.631300926208496, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.33201467990875244, + "loss_total_v6": 0.5165849924087524, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1151 + }, + { + "epoch": 0.4293701080879612, + "grad_norm": 17.32374382019043, + "learning_rate": 6.790756929242439e-06, + "loss": 10.598405838012695, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.36100873351097107, + "loss_total_v6": 0.5536357164382935, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1152 + }, + { + "epoch": 0.42974282519567647, + "grad_norm": 21.011760711669922, + "learning_rate": 6.784848951991647e-06, + "loss": 10.421058654785156, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.44354838132858276, + "loss_total_v6": 0.6227476000785828, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1153 + }, + { + "epoch": 0.4301155423033917, + "grad_norm": 18.865949630737305, + "learning_rate": 6.778938117494306e-06, + "loss": 11.056482315063477, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.6354509592056274, + "loss_total_v6": 0.8090349435806274, + "loss_wm_recon": 0.1240234375, + "loss_wm_recon_w": 0.031005859375, + "step": 1154 + }, + { + "epoch": 0.43048825941110697, + "grad_norm": 17.162992477416992, + "learning_rate": 6.773024435212678e-06, + "loss": 10.366262435913086, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4001389741897583, + "loss_total_v6": 0.4302903413772583, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1155 + }, + { + "epoch": 0.4308609765188222, + "grad_norm": 17.102174758911133, + "learning_rate": 6.767107914613589e-06, + "loss": 9.917074203491211, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4748722314834595, + "loss_total_v6": 0.6714054346084595, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1156 + }, + { + "epoch": 0.43123369362653746, + "grad_norm": 17.612443923950195, + "learning_rate": 6.761188565168402e-06, + "loss": 10.990577697753906, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.41650325059890747, + "loss_total_v6": 0.6262200474739075, + "loss_wm_recon": 0.1669921875, + "loss_wm_recon_w": 0.041748046875, + "step": 1157 + }, + { + "epoch": 0.4316064107342527, + "grad_norm": 16.880290985107422, + "learning_rate": 6.755266396353013e-06, + "loss": 10.305078506469727, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5616986751556396, + "loss_total_v6": 0.7378461360931396, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1158 + }, + { + "epoch": 0.43197912784196796, + "grad_norm": 17.70355987548828, + "learning_rate": 6.74934141764783e-06, + "loss": 10.177397727966309, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.3386155366897583, + "loss_total_v6": 0.5170823335647583, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1159 + }, + { + "epoch": 0.4323518449496832, + "grad_norm": 17.25799560546875, + "learning_rate": 6.743413638537757e-06, + "loss": 10.152141571044922, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6214136481285095, + "loss_total_v6": 0.8032984137535095, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1160 + }, + { + "epoch": 0.43272456205739845, + "grad_norm": 18.399295806884766, + "learning_rate": 6.7374830685121876e-06, + "loss": 10.425262451171875, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.47641825675964355, + "loss_total_v6": 0.6607444286346436, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1161 + }, + { + "epoch": 0.4330972791651137, + "grad_norm": 18.467491149902344, + "learning_rate": 6.731549717064975e-06, + "loss": 10.43875503540039, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.4575853943824768, + "loss_total_v6": 0.6239672303199768, + "loss_wm_recon": 0.10302734375, + "loss_wm_recon_w": 0.0257568359375, + "step": 1162 + }, + { + "epoch": 0.43346999627282895, + "grad_norm": 17.71402931213379, + "learning_rate": 6.72561359369443e-06, + "loss": 10.526710510253906, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5529844760894775, + "loss_total_v6": 0.7290098667144775, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1163 + }, + { + "epoch": 0.43384271338054414, + "grad_norm": 17.988664627075195, + "learning_rate": 6.7196747079033006e-06, + "loss": 10.337434768676758, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.3810911178588867, + "loss_total_v6": 0.5566282272338867, + "loss_wm_recon": 0.1162109375, + "loss_wm_recon_w": 0.029052734375, + "step": 1164 + }, + { + "epoch": 0.4342154304882594, + "grad_norm": 17.254762649536133, + "learning_rate": 6.7137330691987565e-06, + "loss": 10.176240921020508, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4949536919593811, + "loss_total_v6": 0.6824536919593811, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 1165 + }, + { + "epoch": 0.43458814759597464, + "grad_norm": 17.308631896972656, + "learning_rate": 6.707788687092371e-06, + "loss": 10.60084342956543, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4860878586769104, + "loss_total_v6": 0.6646767258644104, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1166 + }, + { + "epoch": 0.4349608647036899, + "grad_norm": 18.034875869750977, + "learning_rate": 6.701841571100114e-06, + "loss": 10.310181617736816, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5833525061607361, + "loss_total_v6": 0.7591337561607361, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1167 + }, + { + "epoch": 0.43533358181140513, + "grad_norm": 16.871238708496094, + "learning_rate": 6.6958917307423295e-06, + "loss": 10.946849822998047, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4416462182998657, + "loss_total_v6": 0.6201130151748657, + "loss_wm_recon": 0.1201171875, + "loss_wm_recon_w": 0.030029296875, + "step": 1168 + }, + { + "epoch": 0.4357062989191204, + "grad_norm": 20.118335723876953, + "learning_rate": 6.689939175543723e-06, + "loss": 10.176858901977539, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4037540555000305, + "loss_total_v6": 0.5844181180000305, + "loss_wm_recon": 0.12109375, + "loss_wm_recon_w": 0.0302734375, + "step": 1169 + }, + { + "epoch": 0.43607901602683563, + "grad_norm": 16.929616928100586, + "learning_rate": 6.683983915033345e-06, + "loss": 10.217012405395508, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4128915071487427, + "loss_total_v6": 0.4429208040237427, + "loss_wm_recon": 0.1201171875, + "loss_wm_recon_w": 0.030029296875, + "step": 1170 + }, + { + "epoch": 0.4364517331345509, + "grad_norm": 17.47414207458496, + "learning_rate": 6.67802595874458e-06, + "loss": 10.321277618408203, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.42355459928512573, + "loss_total_v6": 0.6178905367851257, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 1171 + }, + { + "epoch": 0.4368244502422661, + "grad_norm": 17.511262893676758, + "learning_rate": 6.672065316215125e-06, + "loss": 10.797460556030273, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.7879147529602051, + "loss_total_v6": 0.9710202217102051, + "loss_wm_recon": 0.115234375, + "loss_wm_recon_w": 0.02880859375, + "step": 1172 + }, + { + "epoch": 0.4371971673499814, + "grad_norm": 17.246191024780273, + "learning_rate": 6.666101996986978e-06, + "loss": 10.363182067871094, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4877735674381256, + "loss_total_v6": 0.5174366235733032, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1173 + }, + { + "epoch": 0.4375698844576966, + "grad_norm": 16.669649124145508, + "learning_rate": 6.660136010606421e-06, + "loss": 10.463549613952637, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.48447534441947937, + "loss_total_v6": 0.6697781085968018, + "loss_wm_recon": 0.1162109375, + "loss_wm_recon_w": 0.029052734375, + "step": 1174 + }, + { + "epoch": 0.43794260156541187, + "grad_norm": 17.08426284790039, + "learning_rate": 6.654167366624009e-06, + "loss": 10.609320640563965, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5226592421531677, + "loss_total_v6": 0.7055205702781677, + "loss_wm_recon": 0.1220703125, + "loss_wm_recon_w": 0.030517578125, + "step": 1175 + }, + { + "epoch": 0.4383153186731271, + "grad_norm": 16.6566104888916, + "learning_rate": 6.648196074594547e-06, + "loss": 10.709836959838867, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5699298977851868, + "loss_total_v6": 0.7587726712226868, + "loss_wm_recon": 0.11474609375, + "loss_wm_recon_w": 0.0286865234375, + "step": 1176 + }, + { + "epoch": 0.43868803578084237, + "grad_norm": 17.64805030822754, + "learning_rate": 6.642222144077082e-06, + "loss": 10.498024940490723, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.9258622527122498, + "loss_total_v6": 1.1097002029418945, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1177 + }, + { + "epoch": 0.43906075288855756, + "grad_norm": 17.183279037475586, + "learning_rate": 6.636245584634885e-06, + "loss": 10.813837051391602, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6024149656295776, + "loss_total_v6": 0.7857645750045776, + "loss_wm_recon": 0.1240234375, + "loss_wm_recon_w": 0.031005859375, + "step": 1178 + }, + { + "epoch": 0.4394334699962728, + "grad_norm": 19.883617401123047, + "learning_rate": 6.630266405835433e-06, + "loss": 10.527263641357422, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4506520628929138, + "loss_total_v6": 0.6374196410179138, + "loss_wm_recon": 0.1220703125, + "loss_wm_recon_w": 0.030517578125, + "step": 1179 + }, + { + "epoch": 0.43980618710398806, + "grad_norm": 16.8948917388916, + "learning_rate": 6.624284617250397e-06, + "loss": 10.689327239990234, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5936047434806824, + "loss_total_v6": 0.7735363841056824, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1180 + }, + { + "epoch": 0.4401789042117033, + "grad_norm": 17.865345001220703, + "learning_rate": 6.6183002284556285e-06, + "loss": 10.899646759033203, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.5691879391670227, + "loss_total_v6": 0.7405746579170227, + "loss_wm_recon": 0.123046875, + "loss_wm_recon_w": 0.03076171875, + "step": 1181 + }, + { + "epoch": 0.44055162131941855, + "grad_norm": 17.43901824951172, + "learning_rate": 6.612313249031137e-06, + "loss": 10.589815139770508, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5335014462471008, + "loss_total_v6": 0.7233207821846008, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1182 + }, + { + "epoch": 0.4409243384271338, + "grad_norm": 17.45182228088379, + "learning_rate": 6.606323688561082e-06, + "loss": 10.922382354736328, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.7198402881622314, + "loss_total_v6": 0.9229652881622314, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 1183 + }, + { + "epoch": 0.44129705553484905, + "grad_norm": 16.98387908935547, + "learning_rate": 6.600331556633759e-06, + "loss": 10.648002624511719, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6517612338066101, + "loss_total_v6": 0.8478061556816101, + "loss_wm_recon": 0.1357421875, + "loss_wm_recon_w": 0.033935546875, + "step": 1184 + }, + { + "epoch": 0.4416697726425643, + "grad_norm": 17.606292724609375, + "learning_rate": 6.594336862841569e-06, + "loss": 10.566561698913574, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.7516376972198486, + "loss_total_v6": 0.9313251972198486, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1185 + }, + { + "epoch": 0.44204248975027954, + "grad_norm": 16.456079483032227, + "learning_rate": 6.5883396167810235e-06, + "loss": 10.244869232177734, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.5131525993347168, + "loss_total_v6": 0.6833186149597168, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1186 + }, + { + "epoch": 0.4424152068579948, + "grad_norm": 17.820798873901367, + "learning_rate": 6.582339828052717e-06, + "loss": 10.483570098876953, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3686952590942383, + "loss_total_v6": 0.5483827590942383, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1187 + }, + { + "epoch": 0.44278792396571004, + "grad_norm": 17.65382957458496, + "learning_rate": 6.576337506261314e-06, + "loss": 10.36661148071289, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.37847262620925903, + "loss_total_v6": 0.551690399646759, + "loss_wm_recon": 0.10693359375, + "loss_wm_recon_w": 0.0267333984375, + "step": 1188 + }, + { + "epoch": 0.4431606410734253, + "grad_norm": 16.889253616333008, + "learning_rate": 6.570332661015531e-06, + "loss": 10.056190490722656, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.5980343222618103, + "loss_total_v6": 0.7669796347618103, + "loss_wm_recon": 0.10546875, + "loss_wm_recon_w": 0.0263671875, + "step": 1189 + }, + { + "epoch": 0.44353335818114054, + "grad_norm": 17.27818489074707, + "learning_rate": 6.564325301928134e-06, + "loss": 10.492124557495117, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.21411509811878204, + "loss_total_v6": 0.39453500509262085, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1190 + }, + { + "epoch": 0.4439060752888558, + "grad_norm": 17.42177391052246, + "learning_rate": 6.558315438615901e-06, + "loss": 10.514500617980957, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.5293413400650024, + "loss_total_v6": 0.7231889963150024, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 1191 + }, + { + "epoch": 0.444278792396571, + "grad_norm": 17.394527435302734, + "learning_rate": 6.552303080699625e-06, + "loss": 10.440889358520508, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.20537243783473969, + "loss_total_v6": 0.3718763589859009, + "loss_wm_recon": 0.103515625, + "loss_wm_recon_w": 0.02587890625, + "step": 1192 + }, + { + "epoch": 0.4446515095042862, + "grad_norm": 18.042251586914062, + "learning_rate": 6.546288237804093e-06, + "loss": 10.397114753723145, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6590563058853149, + "loss_total_v6": 0.8575426340103149, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1193 + }, + { + "epoch": 0.4450242266120015, + "grad_norm": 18.067691802978516, + "learning_rate": 6.540270919558069e-06, + "loss": 10.70627212524414, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.9843595027923584, + "loss_total_v6": 1.1625821590423584, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 1194 + }, + { + "epoch": 0.4453969437197167, + "grad_norm": 17.879878997802734, + "learning_rate": 6.534251135594281e-06, + "loss": 10.192299842834473, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.21412277221679688, + "loss_total_v6": 0.3962516784667969, + "loss_wm_recon": 0.119140625, + "loss_wm_recon_w": 0.02978515625, + "step": 1195 + }, + { + "epoch": 0.44576966082743197, + "grad_norm": 17.619754791259766, + "learning_rate": 6.528228895549397e-06, + "loss": 10.35364055633545, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.1780381202697754, + "loss_total_v6": 0.3779892921447754, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 1196 + }, + { + "epoch": 0.4461423779351472, + "grad_norm": 17.25165557861328, + "learning_rate": 6.52220420906403e-06, + "loss": 10.709280014038086, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4764205813407898, + "loss_total_v6": 0.6592819094657898, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1197 + }, + { + "epoch": 0.44651509504286246, + "grad_norm": 16.817577362060547, + "learning_rate": 6.516177085782697e-06, + "loss": 10.892976760864258, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.371489942073822, + "loss_total_v6": 0.558623731136322, + "loss_wm_recon": 0.12353515625, + "loss_wm_recon_w": 0.0308837890625, + "step": 1198 + }, + { + "epoch": 0.4468878121505777, + "grad_norm": 17.89738655090332, + "learning_rate": 6.5101475353538225e-06, + "loss": 10.354360580444336, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4878062605857849, + "loss_total_v6": 0.6613902449607849, + "loss_wm_recon": 0.1083984375, + "loss_wm_recon_w": 0.027099609375, + "step": 1199 + }, + { + "epoch": 0.44726052925829296, + "grad_norm": 17.237234115600586, + "learning_rate": 6.504115567429713e-06, + "loss": 10.57895278930664, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.39655113220214844, + "loss_total_v6": 0.5703792572021484, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1200 + }, + { + "epoch": 0.44726052925829296, + "eval_loss": 0.6394014358520508, + "eval_loss_alignment": 0.28232254941239315, + "eval_loss_alignment_w": 0.14116127470619658, + "eval_loss_sft": 0.4697742432387721, + "eval_loss_total_v6": 0.6394013756742845, + "eval_loss_wm_recon": 0.11386343149038461, + "eval_loss_wm_recon_w": 0.028465857872596152, + "eval_runtime": 227.6867, + "eval_samples_per_second": 8.195, + "eval_steps_per_second": 1.028, + "step": 1200 + }, + { + "epoch": 0.4476332463660082, + "grad_norm": 16.412952423095703, + "learning_rate": 6.498081191666549e-06, + "loss": 10.688136100769043, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.2654329836368561, + "loss_total_v6": 0.2998568117618561, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1201 + }, + { + "epoch": 0.44800596347372346, + "grad_norm": 17.940807342529297, + "learning_rate": 6.492044417724362e-06, + "loss": 10.735586166381836, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.23432616889476776, + "loss_total_v6": 0.41022950410842896, + "loss_wm_recon": 0.11767578125, + "loss_wm_recon_w": 0.0294189453125, + "step": 1202 + }, + { + "epoch": 0.4483786805814387, + "grad_norm": 16.65979766845703, + "learning_rate": 6.4860052552670216e-06, + "loss": 10.368973731994629, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4043912887573242, + "loss_total_v6": 0.5762662887573242, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1203 + }, + { + "epoch": 0.44875139768915395, + "grad_norm": 17.56489372253418, + "learning_rate": 6.4799637139622275e-06, + "loss": 9.852102279663086, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.5111886262893677, + "loss_total_v6": 0.7108956575393677, + "loss_wm_recon": 0.134765625, + "loss_wm_recon_w": 0.03369140625, + "step": 1204 + }, + { + "epoch": 0.4491241147968692, + "grad_norm": 17.55339241027832, + "learning_rate": 6.473919803481482e-06, + "loss": 10.570352554321289, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.3646109402179718, + "loss_total_v6": 0.5396597385406494, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1205 + }, + { + "epoch": 0.4494968319045844, + "grad_norm": 17.431354522705078, + "learning_rate": 6.46787353350008e-06, + "loss": 10.491523742675781, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5777965784072876, + "loss_total_v6": 0.7607799768447876, + "loss_wm_recon": 0.11474609375, + "loss_wm_recon_w": 0.0286865234375, + "step": 1206 + }, + { + "epoch": 0.44986954901229964, + "grad_norm": 17.251829147338867, + "learning_rate": 6.461824913697096e-06, + "loss": 10.919958114624023, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.7070039510726929, + "loss_total_v6": 0.9150117635726929, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 1207 + }, + { + "epoch": 0.4502422661200149, + "grad_norm": 17.081205368041992, + "learning_rate": 6.4557739537553644e-06, + "loss": 10.558197021484375, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.409391313791275, + "loss_total_v6": 0.5890787839889526, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1208 + }, + { + "epoch": 0.45061498322773014, + "grad_norm": 18.614641189575195, + "learning_rate": 6.449720663361465e-06, + "loss": 10.836708068847656, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.21111126244068146, + "loss_total_v6": 0.38310831785202026, + "loss_wm_recon": 0.10205078125, + "loss_wm_recon_w": 0.0255126953125, + "step": 1209 + }, + { + "epoch": 0.4509877003354454, + "grad_norm": 16.79824447631836, + "learning_rate": 6.4436650522057135e-06, + "loss": 10.602072715759277, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.31113988161087036, + "loss_total_v6": 0.48582249879837036, + "loss_wm_recon": 0.10498046875, + "loss_wm_recon_w": 0.0262451171875, + "step": 1210 + }, + { + "epoch": 0.45136041744316063, + "grad_norm": 17.542524337768555, + "learning_rate": 6.4376071299821355e-06, + "loss": 10.225059509277344, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5017296671867371, + "loss_total_v6": 0.6850792765617371, + "loss_wm_recon": 0.1162109375, + "loss_wm_recon_w": 0.029052734375, + "step": 1211 + }, + { + "epoch": 0.4517331345508759, + "grad_norm": 16.46800994873047, + "learning_rate": 6.431546906388455e-06, + "loss": 10.76602554321289, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.17814119160175323, + "loss_total_v6": 0.34159332513809204, + "loss_wm_recon": 0.09912109375, + "loss_wm_recon_w": 0.0247802734375, + "step": 1212 + }, + { + "epoch": 0.45210585165859113, + "grad_norm": 16.301870346069336, + "learning_rate": 6.425484391126086e-06, + "loss": 10.078857421875, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.15743142366409302, + "loss_total_v6": 0.345907986164093, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 1213 + }, + { + "epoch": 0.4524785687663064, + "grad_norm": 16.777912139892578, + "learning_rate": 6.419419593900109e-06, + "loss": 10.578461647033691, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.17307403683662415, + "loss_total_v6": 0.35154083371162415, + "loss_wm_recon": 0.1201171875, + "loss_wm_recon_w": 0.030029296875, + "step": 1214 + }, + { + "epoch": 0.4528512858740216, + "grad_norm": 16.588754653930664, + "learning_rate": 6.413352524419252e-06, + "loss": 10.275077819824219, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.2659229338169098, + "loss_total_v6": 0.4507373869419098, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1215 + }, + { + "epoch": 0.4532240029817369, + "grad_norm": 16.88852310180664, + "learning_rate": 6.407283192395888e-06, + "loss": 9.775909423828125, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.39835649728775024, + "loss_total_v6": 0.5763350129127502, + "loss_wm_recon": 0.1181640625, + "loss_wm_recon_w": 0.029541015625, + "step": 1216 + }, + { + "epoch": 0.4535967200894521, + "grad_norm": 17.179841995239258, + "learning_rate": 6.4012116075460085e-06, + "loss": 10.711061477661133, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.58829265832901, + "loss_total_v6": 0.77554851770401, + "loss_wm_recon": 0.1318359375, + "loss_wm_recon_w": 0.032958984375, + "step": 1217 + }, + { + "epoch": 0.45396943719716737, + "grad_norm": 17.02059555053711, + "learning_rate": 6.3951377795892104e-06, + "loss": 10.587371826171875, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.30910518765449524, + "loss_total_v6": 0.34450557827949524, + "loss_wm_recon": 0.1416015625, + "loss_wm_recon_w": 0.035400390625, + "step": 1218 + }, + { + "epoch": 0.4543421543048826, + "grad_norm": 17.505252838134766, + "learning_rate": 6.389061718248686e-06, + "loss": 10.71355152130127, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4216107428073883, + "loss_total_v6": 0.5994671583175659, + "loss_wm_recon": 0.11767578125, + "loss_wm_recon_w": 0.0294189453125, + "step": 1219 + }, + { + "epoch": 0.45471487141259787, + "grad_norm": 17.1263427734375, + "learning_rate": 6.382983433251196e-06, + "loss": 10.299619674682617, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.607566773891449, + "loss_total_v6": 0.782615602016449, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1220 + }, + { + "epoch": 0.45508758852031306, + "grad_norm": 16.822568893432617, + "learning_rate": 6.376902934327067e-06, + "loss": 10.49550724029541, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4778549373149872, + "loss_total_v6": 0.6601059436798096, + "loss_wm_recon": 0.11181640625, + "loss_wm_recon_w": 0.0279541015625, + "step": 1221 + }, + { + "epoch": 0.4554603056280283, + "grad_norm": 17.156972885131836, + "learning_rate": 6.370820231210168e-06, + "loss": 10.057165145874023, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5156586170196533, + "loss_total_v6": 0.6988861560821533, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1222 + }, + { + "epoch": 0.45583302273574355, + "grad_norm": 16.438831329345703, + "learning_rate": 6.364735333637892e-06, + "loss": 10.740126609802246, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.2506571412086487, + "loss_total_v6": 0.4355936646461487, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1223 + }, + { + "epoch": 0.4562057398434588, + "grad_norm": 16.93054962158203, + "learning_rate": 6.358648251351151e-06, + "loss": 9.923982620239258, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4179910719394684, + "loss_total_v6": 0.592795729637146, + "loss_wm_recon": 0.11328125, + "loss_wm_recon_w": 0.0283203125, + "step": 1224 + }, + { + "epoch": 0.45657845695117405, + "grad_norm": 16.796741485595703, + "learning_rate": 6.3525589940943545e-06, + "loss": 10.322351455688477, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.668374240398407, + "loss_total_v6": 0.865151584148407, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1225 + }, + { + "epoch": 0.4569511740588893, + "grad_norm": 17.394895553588867, + "learning_rate": 6.346467571615387e-06, + "loss": 10.336979866027832, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.5405805110931396, + "loss_total_v6": 0.7105023860931396, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1226 + }, + { + "epoch": 0.45732389116660455, + "grad_norm": 17.88475799560547, + "learning_rate": 6.340373993665607e-06, + "loss": 10.20073127746582, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5721272230148315, + "loss_total_v6": 0.7513264417648315, + "loss_wm_recon": 0.123046875, + "loss_wm_recon_w": 0.03076171875, + "step": 1227 + }, + { + "epoch": 0.4576966082743198, + "grad_norm": 17.477563858032227, + "learning_rate": 6.334278269999819e-06, + "loss": 10.889457702636719, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.3937981426715851, + "loss_total_v6": 0.5805656909942627, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1228 + }, + { + "epoch": 0.45806932538203504, + "grad_norm": 17.336469650268555, + "learning_rate": 6.328180410376264e-06, + "loss": 10.60241413116455, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4566687345504761, + "loss_total_v6": 0.6379431486129761, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1229 + }, + { + "epoch": 0.4584420424897503, + "grad_norm": 17.75368881225586, + "learning_rate": 6.322080424556598e-06, + "loss": 10.384379386901855, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4200066924095154, + "loss_total_v6": 0.6039666533470154, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1230 + }, + { + "epoch": 0.45881475959746554, + "grad_norm": 17.222951889038086, + "learning_rate": 6.315978322305892e-06, + "loss": 10.712211608886719, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.35109055042266846, + "loss_total_v6": 0.5195475816726685, + "loss_wm_recon": 0.103515625, + "loss_wm_recon_w": 0.02587890625, + "step": 1231 + }, + { + "epoch": 0.4591874767051808, + "grad_norm": 20.084949493408203, + "learning_rate": 6.30987411339259e-06, + "loss": 10.330175399780273, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4875960350036621, + "loss_total_v6": 0.6705794334411621, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1232 + }, + { + "epoch": 0.45956019381289603, + "grad_norm": 17.177602767944336, + "learning_rate": 6.303767807588521e-06, + "loss": 10.082014083862305, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6062329411506653, + "loss_total_v6": 0.7810376286506653, + "loss_wm_recon": 0.11328125, + "loss_wm_recon_w": 0.0283203125, + "step": 1233 + }, + { + "epoch": 0.4599329109206113, + "grad_norm": 17.033445358276367, + "learning_rate": 6.297659414668865e-06, + "loss": 10.683338165283203, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.4322946071624756, + "loss_total_v6": 0.5989205837249756, + "loss_wm_recon": 0.10400390625, + "loss_wm_recon_w": 0.0260009765625, + "step": 1234 + }, + { + "epoch": 0.4603056280283265, + "grad_norm": 15.818922996520996, + "learning_rate": 6.291548944412141e-06, + "loss": 10.151865005493164, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3289870619773865, + "loss_total_v6": 0.4963454604148865, + "loss_wm_recon": 0.10693359375, + "loss_wm_recon_w": 0.0267333984375, + "step": 1235 + }, + { + "epoch": 0.4606783451360417, + "grad_norm": 16.378446578979492, + "learning_rate": 6.285436406600201e-06, + "loss": 10.487459182739258, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.48240697383880615, + "loss_total_v6": 0.6701511144638062, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1236 + }, + { + "epoch": 0.46105106224375697, + "grad_norm": 16.973045349121094, + "learning_rate": 6.2793218110181974e-06, + "loss": 10.513952255249023, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4341038763523102, + "loss_total_v6": 0.6184300184249878, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1237 + }, + { + "epoch": 0.4614237793514722, + "grad_norm": 17.923770904541016, + "learning_rate": 6.273205167454585e-06, + "loss": 9.928689956665039, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.5326316952705383, + "loss_total_v6": 0.7069481015205383, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 1238 + }, + { + "epoch": 0.46179649645918747, + "grad_norm": 16.352067947387695, + "learning_rate": 6.267086485701093e-06, + "loss": 9.916044235229492, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.34604963660240173, + "loss_total_v6": 0.5335496664047241, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1239 + }, + { + "epoch": 0.4621692135669027, + "grad_norm": 17.11444664001465, + "learning_rate": 6.2609657755527135e-06, + "loss": 10.218713760375977, + "loss_alignment": 0.34765625, + "loss_alignment_w": 0.173828125, + "loss_sft": 0.3896227478981018, + "loss_total_v6": 0.6012926697731018, + "loss_wm_recon": 0.1513671875, + "loss_wm_recon_w": 0.037841796875, + "step": 1240 + }, + { + "epoch": 0.46254193067461796, + "grad_norm": 18.078594207763672, + "learning_rate": 6.2548430468076835e-06, + "loss": 10.5157470703125, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.425858736038208, + "loss_total_v6": 0.454423189163208, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1241 + }, + { + "epoch": 0.4629146477823332, + "grad_norm": 17.561416625976562, + "learning_rate": 6.2487183092674805e-06, + "loss": 10.538007736206055, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.5690774321556091, + "loss_total_v6": 0.7400979399681091, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1242 + }, + { + "epoch": 0.46328736489004846, + "grad_norm": 20.406557083129883, + "learning_rate": 6.242591572736786e-06, + "loss": 10.659696578979492, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.5282015800476074, + "loss_total_v6": 0.6944613456726074, + "loss_wm_recon": 0.1025390625, + "loss_wm_recon_w": 0.025634765625, + "step": 1243 + }, + { + "epoch": 0.4636600819977637, + "grad_norm": 17.85334587097168, + "learning_rate": 6.236462847023488e-06, + "loss": 10.318655967712402, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6491593718528748, + "loss_total_v6": 0.8307999968528748, + "loss_wm_recon": 0.1328125, + "loss_wm_recon_w": 0.033203125, + "step": 1244 + }, + { + "epoch": 0.46403279910547895, + "grad_norm": 17.56736183166504, + "learning_rate": 6.230332141938661e-06, + "loss": 10.241683959960938, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.21671493351459503, + "loss_total_v6": 0.40983015298843384, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 1245 + }, + { + "epoch": 0.4644055162131942, + "grad_norm": 17.73232650756836, + "learning_rate": 6.224199467296542e-06, + "loss": 10.358070373535156, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.09099162369966507, + "loss_total_v6": 0.2587162256240845, + "loss_wm_recon": 0.1005859375, + "loss_wm_recon_w": 0.025146484375, + "step": 1246 + }, + { + "epoch": 0.46477823332090945, + "grad_norm": 17.36038589477539, + "learning_rate": 6.218064832914523e-06, + "loss": 10.25395393371582, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6169142723083496, + "loss_total_v6": 0.8132033348083496, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1247 + }, + { + "epoch": 0.4651509504286247, + "grad_norm": 17.053728103637695, + "learning_rate": 6.211928248613138e-06, + "loss": 10.521441459655762, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.3277434706687927, + "loss_total_v6": 0.4958342909812927, + "loss_wm_recon": 0.10205078125, + "loss_wm_recon_w": 0.0255126953125, + "step": 1248 + }, + { + "epoch": 0.4655236675363399, + "grad_norm": 17.001110076904297, + "learning_rate": 6.205789724216034e-06, + "loss": 10.2483491897583, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.3779701888561249, + "loss_total_v6": 0.5743813514709473, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 1249 + }, + { + "epoch": 0.46589638464405514, + "grad_norm": 16.40261459350586, + "learning_rate": 6.199649269549972e-06, + "loss": 10.421024322509766, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.8353612422943115, + "loss_total_v6": 1.0397069454193115, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 1250 + }, + { + "epoch": 0.4662691017517704, + "grad_norm": 17.982391357421875, + "learning_rate": 6.1935068944448e-06, + "loss": 10.305112838745117, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.07237844914197922, + "loss_total_v6": 0.24449759721755981, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1251 + }, + { + "epoch": 0.46664181885948564, + "grad_norm": 16.77159881591797, + "learning_rate": 6.187362608733436e-06, + "loss": 10.30129623413086, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6424242854118347, + "loss_total_v6": 0.8271166682243347, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 1252 + }, + { + "epoch": 0.4670145359672009, + "grad_norm": 17.373125076293945, + "learning_rate": 6.1812164222518626e-06, + "loss": 9.869396209716797, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.40783730149269104, + "loss_total_v6": 0.5876468420028687, + "loss_wm_recon": 0.11767578125, + "loss_wm_recon_w": 0.0294189453125, + "step": 1253 + }, + { + "epoch": 0.46738725307491613, + "grad_norm": 17.62310791015625, + "learning_rate": 6.175068344839102e-06, + "loss": 10.019057273864746, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.35273101925849915, + "loss_total_v6": 0.5357143878936768, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1254 + }, + { + "epoch": 0.4677599701826314, + "grad_norm": 17.110464096069336, + "learning_rate": 6.168918386337206e-06, + "loss": 10.414738655090332, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.597472071647644, + "loss_total_v6": 0.764464259147644, + "loss_wm_recon": 0.09765625, + "loss_wm_recon_w": 0.0244140625, + "step": 1255 + }, + { + "epoch": 0.4681326872903466, + "grad_norm": 17.062000274658203, + "learning_rate": 6.162766556591236e-06, + "loss": 10.404722213745117, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.49172544479370117, + "loss_total_v6": 0.6687273979187012, + "loss_wm_recon": 0.1064453125, + "loss_wm_recon_w": 0.026611328125, + "step": 1256 + }, + { + "epoch": 0.4685054043980619, + "grad_norm": 16.82146453857422, + "learning_rate": 6.156612865449245e-06, + "loss": 10.13945198059082, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6454940438270569, + "loss_total_v6": 0.8292098641395569, + "loss_wm_recon": 0.11767578125, + "loss_wm_recon_w": 0.0294189453125, + "step": 1257 + }, + { + "epoch": 0.4688781215057771, + "grad_norm": 16.595706939697266, + "learning_rate": 6.150457322762274e-06, + "loss": 10.580530166625977, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.25670626759529114, + "loss_total_v6": 0.43224337697029114, + "loss_wm_recon": 0.1162109375, + "loss_wm_recon_w": 0.029052734375, + "step": 1258 + }, + { + "epoch": 0.46925083861349237, + "grad_norm": 16.993120193481445, + "learning_rate": 6.144299938384325e-06, + "loss": 10.620462417602539, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.177963525056839, + "loss_total_v6": 0.344833642244339, + "loss_wm_recon": 0.09716796875, + "loss_wm_recon_w": 0.0242919921875, + "step": 1259 + }, + { + "epoch": 0.4696235557212076, + "grad_norm": 16.737598419189453, + "learning_rate": 6.138140722172344e-06, + "loss": 10.349276542663574, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5969968438148499, + "loss_total_v6": 0.7807126641273499, + "loss_wm_recon": 0.11767578125, + "loss_wm_recon_w": 0.0294189453125, + "step": 1260 + }, + { + "epoch": 0.46999627282892287, + "grad_norm": 17.39426040649414, + "learning_rate": 6.131979683986213e-06, + "loss": 10.531099319458008, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6066464185714722, + "loss_total_v6": 0.7901180982589722, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1261 + }, + { + "epoch": 0.4703689899366381, + "grad_norm": 16.87352180480957, + "learning_rate": 6.125816833688731e-06, + "loss": 10.432950973510742, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5791639685630798, + "loss_total_v6": 0.7713026404380798, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1262 + }, + { + "epoch": 0.47074170704435336, + "grad_norm": 16.790424346923828, + "learning_rate": 6.1196521811455975e-06, + "loss": 10.277976989746094, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.45877987146377563, + "loss_total_v6": 0.48710018396377563, + "loss_wm_recon": 0.11328125, + "loss_wm_recon_w": 0.0283203125, + "step": 1263 + }, + { + "epoch": 0.47111442415206856, + "grad_norm": 17.04224395751953, + "learning_rate": 6.113485736225394e-06, + "loss": 9.99694538116455, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.36179476976394653, + "loss_total_v6": 0.5431912541389465, + "loss_wm_recon": 0.1162109375, + "loss_wm_recon_w": 0.029052734375, + "step": 1264 + }, + { + "epoch": 0.4714871412597838, + "grad_norm": 17.66082000732422, + "learning_rate": 6.107317508799578e-06, + "loss": 10.344388008117676, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6821146607398987, + "loss_total_v6": 0.8694925904273987, + "loss_wm_recon": 0.12451171875, + "loss_wm_recon_w": 0.0311279296875, + "step": 1265 + }, + { + "epoch": 0.47185985836749905, + "grad_norm": 16.806259155273438, + "learning_rate": 6.101147508742456e-06, + "loss": 10.79809856414795, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5815466046333313, + "loss_total_v6": 0.7670934796333313, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1266 + }, + { + "epoch": 0.4722325754752143, + "grad_norm": 17.381324768066406, + "learning_rate": 6.0949757459311675e-06, + "loss": 10.24765396118164, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.542694628238678, + "loss_total_v6": 0.727020800113678, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1267 + }, + { + "epoch": 0.47260529258292955, + "grad_norm": 16.394886016845703, + "learning_rate": 6.088802230245686e-06, + "loss": 10.187963485717773, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3735679090023041, + "loss_total_v6": 0.5497153997421265, + "loss_wm_recon": 0.10302734375, + "loss_wm_recon_w": 0.0257568359375, + "step": 1268 + }, + { + "epoch": 0.4729780096906448, + "grad_norm": 16.851482391357422, + "learning_rate": 6.082626971568781e-06, + "loss": 10.330684661865234, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6997241377830505, + "loss_total_v6": 0.8759936690330505, + "loss_wm_recon": 0.119140625, + "loss_wm_recon_w": 0.02978515625, + "step": 1269 + }, + { + "epoch": 0.47335072679836004, + "grad_norm": 17.29526138305664, + "learning_rate": 6.076449979786014e-06, + "loss": 10.89992904663086, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.417670875787735, + "loss_total_v6": 0.6030956506729126, + "loss_wm_recon": 0.11669921875, + "loss_wm_recon_w": 0.0291748046875, + "step": 1270 + }, + { + "epoch": 0.4737234439060753, + "grad_norm": 17.113920211791992, + "learning_rate": 6.070271264785723e-06, + "loss": 10.79045295715332, + "loss_alignment": 0.33203125, + "loss_alignment_w": 0.166015625, + "loss_sft": 0.1510978788137436, + "loss_total_v6": 0.3475090265274048, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 1271 + }, + { + "epoch": 0.47409616101379054, + "grad_norm": 16.77372169494629, + "learning_rate": 6.064090836459007e-06, + "loss": 10.318670272827148, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.2830559313297272, + "loss_total_v6": 0.4631096422672272, + "loss_wm_recon": 0.11083984375, + "loss_wm_recon_w": 0.0277099609375, + "step": 1272 + }, + { + "epoch": 0.4744688781215058, + "grad_norm": 16.91921043395996, + "learning_rate": 6.0579087046996995e-06, + "loss": 9.89693832397461, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.38144198060035706, + "loss_total_v6": 0.5638149976730347, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1273 + }, + { + "epoch": 0.47484159522922104, + "grad_norm": 18.35114097595215, + "learning_rate": 6.051724879404371e-06, + "loss": 10.586312294006348, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.46564534306526184, + "loss_total_v6": 0.6354451179504395, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1274 + }, + { + "epoch": 0.4752143123369363, + "grad_norm": 17.592891693115234, + "learning_rate": 6.045539370472296e-06, + "loss": 10.58924388885498, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.278068482875824, + "loss_total_v6": 0.462150514125824, + "loss_wm_recon": 0.119140625, + "loss_wm_recon_w": 0.02978515625, + "step": 1275 + }, + { + "epoch": 0.47558702944465153, + "grad_norm": 17.76186180114746, + "learning_rate": 6.039352187805445e-06, + "loss": 9.998405456542969, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.46810534596443176, + "loss_total_v6": 0.6508445739746094, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1276 + }, + { + "epoch": 0.4759597465523668, + "grad_norm": 18.017410278320312, + "learning_rate": 6.0331633413084725e-06, + "loss": 10.477418899536133, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5140932202339172, + "loss_total_v6": 0.6870668530464172, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1277 + }, + { + "epoch": 0.476332463660082, + "grad_norm": 16.681184768676758, + "learning_rate": 6.026972840888688e-06, + "loss": 10.279533386230469, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.18228107690811157, + "loss_total_v6": 0.3581843972206116, + "loss_wm_recon": 0.10986328125, + "loss_wm_recon_w": 0.0274658203125, + "step": 1278 + }, + { + "epoch": 0.4767051807677972, + "grad_norm": 16.984127044677734, + "learning_rate": 6.020780696456059e-06, + "loss": 10.472041130065918, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.14608953893184662, + "loss_total_v6": 0.3102741241455078, + "loss_wm_recon": 0.09423828125, + "loss_wm_recon_w": 0.0235595703125, + "step": 1279 + }, + { + "epoch": 0.47707789787551247, + "grad_norm": 17.791589736938477, + "learning_rate": 6.014586917923178e-06, + "loss": 9.958934783935547, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6249496936798096, + "loss_total_v6": 0.6531479358673096, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1280 + }, + { + "epoch": 0.4774506149832277, + "grad_norm": 17.023557662963867, + "learning_rate": 6.008391515205253e-06, + "loss": 10.08154296875, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.33168545365333557, + "loss_total_v6": 0.5035604238510132, + "loss_wm_recon": 0.1015625, + "loss_wm_recon_w": 0.025390625, + "step": 1281 + }, + { + "epoch": 0.47782333209094296, + "grad_norm": 16.885162353515625, + "learning_rate": 6.002194498220097e-06, + "loss": 10.347357749938965, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.6377387642860413, + "loss_total_v6": 0.8633247017860413, + "loss_wm_recon": 0.16796875, + "loss_wm_recon_w": 0.0419921875, + "step": 1282 + }, + { + "epoch": 0.4781960491986582, + "grad_norm": 16.57992935180664, + "learning_rate": 5.995995876888105e-06, + "loss": 10.048495292663574, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4104045629501343, + "loss_total_v6": 0.5864299535751343, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1283 + }, + { + "epoch": 0.47856876630637346, + "grad_norm": 17.348098754882812, + "learning_rate": 5.989795661132236e-06, + "loss": 9.935827255249023, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2784000337123871, + "loss_total_v6": 0.4607730805873871, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1284 + }, + { + "epoch": 0.4789414834140887, + "grad_norm": 17.965818405151367, + "learning_rate": 5.983593860878006e-06, + "loss": 10.950199127197266, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6881844997406006, + "loss_total_v6": 0.8694589138031006, + "loss_wm_recon": 0.10791015625, + "loss_wm_recon_w": 0.0269775390625, + "step": 1285 + }, + { + "epoch": 0.47931420052180396, + "grad_norm": 17.012239456176758, + "learning_rate": 5.977390486053469e-06, + "loss": 10.757706642150879, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.47563058137893677, + "loss_total_v6": 0.6626423001289368, + "loss_wm_recon": 0.123046875, + "loss_wm_recon_w": 0.03076171875, + "step": 1286 + }, + { + "epoch": 0.4796869176295192, + "grad_norm": 17.973100662231445, + "learning_rate": 5.971185546589196e-06, + "loss": 10.367738723754883, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.6705665588378906, + "loss_total_v6": 0.8707618713378906, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 1287 + }, + { + "epoch": 0.48005963473723445, + "grad_norm": 16.35359001159668, + "learning_rate": 5.964979052418263e-06, + "loss": 10.493894577026367, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.37408918142318726, + "loss_total_v6": 0.5473069548606873, + "loss_wm_recon": 0.10693359375, + "loss_wm_recon_w": 0.0267333984375, + "step": 1288 + }, + { + "epoch": 0.4804323518449497, + "grad_norm": 16.927797317504883, + "learning_rate": 5.958771013476235e-06, + "loss": 10.157206535339355, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.15849699079990387, + "loss_total_v6": 0.3473397493362427, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1289 + }, + { + "epoch": 0.48080506895266495, + "grad_norm": 16.62738609313965, + "learning_rate": 5.9525614397011535e-06, + "loss": 10.162463188171387, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.7110183238983154, + "loss_total_v6": 0.8872878551483154, + "loss_wm_recon": 0.103515625, + "loss_wm_recon_w": 0.02587890625, + "step": 1290 + }, + { + "epoch": 0.4811777860603802, + "grad_norm": 16.75748062133789, + "learning_rate": 5.9463503410335125e-06, + "loss": 10.266379356384277, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.7964014410972595, + "loss_total_v6": 0.9672998785972595, + "loss_wm_recon": 0.11328125, + "loss_wm_recon_w": 0.0283203125, + "step": 1291 + }, + { + "epoch": 0.4815505031680954, + "grad_norm": 17.246313095092773, + "learning_rate": 5.940137727416247e-06, + "loss": 10.79493522644043, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6273192763328552, + "loss_total_v6": 0.8129882216453552, + "loss_wm_recon": 0.10986328125, + "loss_wm_recon_w": 0.0274658203125, + "step": 1292 + }, + { + "epoch": 0.48192322027581064, + "grad_norm": 18.231891632080078, + "learning_rate": 5.93392360879472e-06, + "loss": 10.225713729858398, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4175191819667816, + "loss_total_v6": 0.598183274269104, + "loss_wm_recon": 0.11328125, + "loss_wm_recon_w": 0.0283203125, + "step": 1293 + }, + { + "epoch": 0.4822959373835259, + "grad_norm": 24.766307830810547, + "learning_rate": 5.927707995116706e-06, + "loss": 10.018957138061523, + "loss_alignment": 0.25, + "loss_alignment_w": 0.125, + "loss_sft": 0.24953727424144745, + "loss_total_v6": 0.40346795320510864, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1294 + }, + { + "epoch": 0.48266865449124113, + "grad_norm": 18.17049789428711, + "learning_rate": 5.921490896332367e-06, + "loss": 9.916997909545898, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.15488852560520172, + "loss_total_v6": 0.3282283544540405, + "loss_wm_recon": 0.099609375, + "loss_wm_recon_w": 0.02490234375, + "step": 1295 + }, + { + "epoch": 0.4830413715989564, + "grad_norm": 17.75370216369629, + "learning_rate": 5.915272322394243e-06, + "loss": 10.239646911621094, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.46837079524993896, + "loss_total_v6": 0.650011420249939, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1296 + }, + { + "epoch": 0.48341408870667163, + "grad_norm": 16.97684669494629, + "learning_rate": 5.909052283257239e-06, + "loss": 10.339414596557617, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.6232264637947083, + "loss_total_v6": 0.7985194325447083, + "loss_wm_recon": 0.107421875, + "loss_wm_recon_w": 0.02685546875, + "step": 1297 + }, + { + "epoch": 0.4837868058143869, + "grad_norm": 16.094112396240234, + "learning_rate": 5.902830788878605e-06, + "loss": 10.130970001220703, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.19281093776226044, + "loss_total_v6": 0.36737149953842163, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1298 + }, + { + "epoch": 0.4841595229221021, + "grad_norm": 16.811063766479492, + "learning_rate": 5.896607849217919e-06, + "loss": 10.650876998901367, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.39796313643455505, + "loss_total_v6": 0.5749651193618774, + "loss_wm_recon": 0.1064453125, + "loss_wm_recon_w": 0.026611328125, + "step": 1299 + }, + { + "epoch": 0.4845322400298174, + "grad_norm": 16.493314743041992, + "learning_rate": 5.890383474237074e-06, + "loss": 10.128194808959961, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.5846633315086365, + "loss_total_v6": 0.7789992690086365, + "loss_wm_recon": 0.13671875, + "loss_wm_recon_w": 0.0341796875, + "step": 1300 + }, + { + "epoch": 0.4845322400298174, + "eval_loss": 0.6352217793464661, + "eval_loss_alignment": 0.28188226161858976, + "eval_loss_alignment_w": 0.14094113080929488, + "eval_loss_sft": 0.4672273244533656, + "eval_loss_total_v6": 0.6352217668014714, + "eval_loss_wm_recon": 0.10821324535924146, + "eval_loss_wm_recon_w": 0.027053311339810364, + "eval_runtime": 229.7707, + "eval_samples_per_second": 8.121, + "eval_steps_per_second": 1.018, + "step": 1300 + }, + { + "epoch": 0.4849049571375326, + "grad_norm": 17.452075958251953, + "learning_rate": 5.884157673900257e-06, + "loss": 10.247725486755371, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.19856688380241394, + "loss_total_v6": 0.36800047755241394, + "loss_wm_recon": 0.099609375, + "loss_wm_recon_w": 0.02490234375, + "step": 1301 + }, + { + "epoch": 0.48527767424524787, + "grad_norm": 17.156169891357422, + "learning_rate": 5.877930458173942e-06, + "loss": 10.388874053955078, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.1898632049560547, + "loss_total_v6": 0.3636913299560547, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1302 + }, + { + "epoch": 0.4856503913529631, + "grad_norm": 16.59684181213379, + "learning_rate": 5.871701837026866e-06, + "loss": 10.509932518005371, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.465036004781723, + "loss_total_v6": 0.6637665033340454, + "loss_wm_recon": 0.138671875, + "loss_wm_recon_w": 0.03466796875, + "step": 1303 + }, + { + "epoch": 0.48602310846067837, + "grad_norm": 17.3280029296875, + "learning_rate": 5.8654718204300164e-06, + "loss": 10.571672439575195, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.7381670475006104, + "loss_total_v6": 0.9155352115631104, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1304 + }, + { + "epoch": 0.4863958255683936, + "grad_norm": 17.000438690185547, + "learning_rate": 5.859240418356614e-06, + "loss": 10.58577823638916, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.2871635854244232, + "loss_total_v6": 0.4624565541744232, + "loss_wm_recon": 0.107421875, + "loss_wm_recon_w": 0.02685546875, + "step": 1305 + }, + { + "epoch": 0.4867685426761088, + "grad_norm": 17.035587310791016, + "learning_rate": 5.8530076407821006e-06, + "loss": 10.299149513244629, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.27884340286254883, + "loss_total_v6": 0.31155824661254883, + "loss_wm_recon": 0.130859375, + "loss_wm_recon_w": 0.03271484375, + "step": 1306 + }, + { + "epoch": 0.48714125978382405, + "grad_norm": 17.62495994567871, + "learning_rate": 5.8467734976841125e-06, + "loss": 10.014762878417969, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4716772437095642, + "loss_total_v6": 0.6572241187095642, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1307 + }, + { + "epoch": 0.4875139768915393, + "grad_norm": 17.003808975219727, + "learning_rate": 5.84053799904248e-06, + "loss": 10.194040298461914, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.47643858194351196, + "loss_total_v6": 0.649046003818512, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1308 + }, + { + "epoch": 0.48788669399925455, + "grad_norm": 17.536890029907227, + "learning_rate": 5.8343011548392e-06, + "loss": 10.02444839477539, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.20865695178508759, + "loss_total_v6": 0.3891989588737488, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1309 + }, + { + "epoch": 0.4882594111069698, + "grad_norm": 17.32396697998047, + "learning_rate": 5.828062975058422e-06, + "loss": 10.344228744506836, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.43173274397850037, + "loss_total_v6": 0.45895442366600037, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1310 + }, + { + "epoch": 0.48863212821468505, + "grad_norm": 17.358768463134766, + "learning_rate": 5.8218234696864366e-06, + "loss": 9.904109954833984, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.95223069190979, + "loss_total_v6": 1.14436936378479, + "loss_wm_recon": 0.1357421875, + "loss_wm_recon_w": 0.033935546875, + "step": 1311 + }, + { + "epoch": 0.4890048453224003, + "grad_norm": 17.248077392578125, + "learning_rate": 5.815582648711654e-06, + "loss": 10.594047546386719, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.13242828845977783, + "loss_total_v6": 0.30198395252227783, + "loss_wm_recon": 0.10009765625, + "loss_wm_recon_w": 0.0250244140625, + "step": 1312 + }, + { + "epoch": 0.48937756243011554, + "grad_norm": 16.614831924438477, + "learning_rate": 5.809340522124592e-06, + "loss": 10.653511047363281, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7450410723686218, + "loss_total_v6": 0.7759248614311218, + "loss_wm_recon": 0.12353515625, + "loss_wm_recon_w": 0.0308837890625, + "step": 1313 + }, + { + "epoch": 0.4897502795378308, + "grad_norm": 16.608064651489258, + "learning_rate": 5.803097099917859e-06, + "loss": 10.402359008789062, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4490947723388672, + "loss_total_v6": 0.6217021942138672, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1314 + }, + { + "epoch": 0.49012299664554604, + "grad_norm": 16.668062210083008, + "learning_rate": 5.796852392086133e-06, + "loss": 10.009309768676758, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4662911891937256, + "loss_total_v6": 0.6397531032562256, + "loss_wm_recon": 0.10009765625, + "loss_wm_recon_w": 0.0250244140625, + "step": 1315 + }, + { + "epoch": 0.4904957137532613, + "grad_norm": 17.24980926513672, + "learning_rate": 5.790606408626156e-06, + "loss": 10.727090835571289, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6083480715751648, + "loss_total_v6": 0.7988998293876648, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 1316 + }, + { + "epoch": 0.49086843086097653, + "grad_norm": 17.112119674682617, + "learning_rate": 5.7843591595367095e-06, + "loss": 10.32588005065918, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.36861392855644226, + "loss_total_v6": 0.5426862239837646, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1317 + }, + { + "epoch": 0.4912411479686918, + "grad_norm": 17.347373962402344, + "learning_rate": 5.778110654818602e-06, + "loss": 10.17872428894043, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.531408965587616, + "loss_total_v6": 0.700232207775116, + "loss_wm_recon": 0.09716796875, + "loss_wm_recon_w": 0.0242919921875, + "step": 1318 + }, + { + "epoch": 0.49161386507640703, + "grad_norm": 16.746065139770508, + "learning_rate": 5.771860904474646e-06, + "loss": 10.279237747192383, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.17815807461738586, + "loss_total_v6": 0.35870006680488586, + "loss_wm_recon": 0.10498046875, + "loss_wm_recon_w": 0.0262451171875, + "step": 1319 + }, + { + "epoch": 0.4919865821841223, + "grad_norm": 16.88494873046875, + "learning_rate": 5.765609918509661e-06, + "loss": 10.132633209228516, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.16081015765666962, + "loss_total_v6": 0.3275582194328308, + "loss_wm_recon": 0.0966796875, + "loss_wm_recon_w": 0.024169921875, + "step": 1320 + }, + { + "epoch": 0.49235929929183747, + "grad_norm": 16.979740142822266, + "learning_rate": 5.759357706930431e-06, + "loss": 10.621543884277344, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6826216578483582, + "loss_total_v6": 0.8601118922233582, + "loss_wm_recon": 0.1083984375, + "loss_wm_recon_w": 0.027099609375, + "step": 1321 + }, + { + "epoch": 0.4927320163995527, + "grad_norm": 16.654394149780273, + "learning_rate": 5.753104279745708e-06, + "loss": 10.136348724365234, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.20049415528774261, + "loss_total_v6": 0.23491798341274261, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1322 + }, + { + "epoch": 0.49310473350726797, + "grad_norm": 16.308231353759766, + "learning_rate": 5.746849646966192e-06, + "loss": 10.407222747802734, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6262395977973938, + "loss_total_v6": 0.8141058087348938, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1323 + }, + { + "epoch": 0.4934774506149832, + "grad_norm": 16.631616592407227, + "learning_rate": 5.7405938186045104e-06, + "loss": 10.84427261352539, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.6744630932807922, + "loss_total_v6": 0.8529298901557922, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1324 + }, + { + "epoch": 0.49385016772269846, + "grad_norm": 17.74072265625, + "learning_rate": 5.734336804675203e-06, + "loss": 10.28736686706543, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.10902072489261627, + "loss_total_v6": 0.2895627021789551, + "loss_wm_recon": 0.10498046875, + "loss_wm_recon_w": 0.0262451171875, + "step": 1325 + }, + { + "epoch": 0.4942228848304137, + "grad_norm": 16.878116607666016, + "learning_rate": 5.7280786151947095e-06, + "loss": 10.28127670288086, + "loss_alignment": 0.36328125, + "loss_alignment_w": 0.181640625, + "loss_sft": 0.5588571429252625, + "loss_total_v6": 0.7756540179252625, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 1326 + }, + { + "epoch": 0.49459560193812896, + "grad_norm": 17.237485885620117, + "learning_rate": 5.721819260181354e-06, + "loss": 10.339120864868164, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.44789358973503113, + "loss_total_v6": 0.6242852210998535, + "loss_wm_recon": 0.11962890625, + "loss_wm_recon_w": 0.0299072265625, + "step": 1327 + }, + { + "epoch": 0.4949683190458442, + "grad_norm": 16.77572250366211, + "learning_rate": 5.715558749655319e-06, + "loss": 10.507745742797852, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4803534150123596, + "loss_total_v6": 0.6657782196998596, + "loss_wm_recon": 0.11669921875, + "loss_wm_recon_w": 0.0291748046875, + "step": 1328 + }, + { + "epoch": 0.49534103615355946, + "grad_norm": 17.042203903198242, + "learning_rate": 5.709297093638646e-06, + "loss": 10.509167671203613, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.6616376638412476, + "loss_total_v6": 0.8348554372787476, + "loss_wm_recon": 0.11474609375, + "loss_wm_recon_w": 0.0286865234375, + "step": 1329 + }, + { + "epoch": 0.4957137532612747, + "grad_norm": 17.618812561035156, + "learning_rate": 5.703034302155202e-06, + "loss": 10.154228210449219, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.49649637937545776, + "loss_total_v6": 0.6699582934379578, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1330 + }, + { + "epoch": 0.49608647036898995, + "grad_norm": 18.670902252197266, + "learning_rate": 5.696770385230679e-06, + "loss": 10.403589248657227, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5687615275382996, + "loss_total_v6": 0.7458855509757996, + "loss_wm_recon": 0.11474609375, + "loss_wm_recon_w": 0.0286865234375, + "step": 1331 + }, + { + "epoch": 0.4964591874767052, + "grad_norm": 16.886594772338867, + "learning_rate": 5.690505352892563e-06, + "loss": 10.186381340026855, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5816243886947632, + "loss_total_v6": 0.7611898183822632, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1332 + }, + { + "epoch": 0.49683190458442045, + "grad_norm": 17.768762588500977, + "learning_rate": 5.684239215170133e-06, + "loss": 10.803712844848633, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6001715660095215, + "loss_total_v6": 0.7882819175720215, + "loss_wm_recon": 0.11181640625, + "loss_wm_recon_w": 0.0279541015625, + "step": 1333 + }, + { + "epoch": 0.4972046216921357, + "grad_norm": 17.53451919555664, + "learning_rate": 5.677971982094431e-06, + "loss": 10.534555435180664, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.18602724373340607, + "loss_total_v6": 0.3531414866447449, + "loss_wm_recon": 0.09814453125, + "loss_wm_recon_w": 0.0245361328125, + "step": 1334 + }, + { + "epoch": 0.4975773387998509, + "grad_norm": 17.020185470581055, + "learning_rate": 5.671703663698259e-06, + "loss": 10.490424156188965, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.2524610459804535, + "loss_total_v6": 0.4282422959804535, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1335 + }, + { + "epoch": 0.49795005590756614, + "grad_norm": 16.778940200805664, + "learning_rate": 5.66543427001615e-06, + "loss": 10.183841705322266, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.40110334753990173, + "loss_total_v6": 0.5757859945297241, + "loss_wm_recon": 0.10498046875, + "loss_wm_recon_w": 0.0262451171875, + "step": 1336 + }, + { + "epoch": 0.4983227730152814, + "grad_norm": 16.402978897094727, + "learning_rate": 5.659163811084364e-06, + "loss": 10.22172737121582, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.23126927018165588, + "loss_total_v6": 0.3985055983066559, + "loss_wm_recon": 0.0986328125, + "loss_wm_recon_w": 0.024658203125, + "step": 1337 + }, + { + "epoch": 0.49869549012299663, + "grad_norm": 16.635391235351562, + "learning_rate": 5.652892296940862e-06, + "loss": 10.902405738830566, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.6053150296211243, + "loss_total_v6": 0.7779224514961243, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1338 + }, + { + "epoch": 0.4990682072307119, + "grad_norm": 17.12713623046875, + "learning_rate": 5.646619737625295e-06, + "loss": 10.738195419311523, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.39448192715644836, + "loss_total_v6": 0.573436975479126, + "loss_wm_recon": 0.1064453125, + "loss_wm_recon_w": 0.026611328125, + "step": 1339 + }, + { + "epoch": 0.49944092433842713, + "grad_norm": 17.707550048828125, + "learning_rate": 5.640346143178989e-06, + "loss": 10.215357780456543, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.8331546187400818, + "loss_total_v6": 1.0274906158447266, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1340 + }, + { + "epoch": 0.4998136414461424, + "grad_norm": 16.146320343017578, + "learning_rate": 5.634071523644929e-06, + "loss": 10.562803268432617, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.33876466751098633, + "loss_total_v6": 0.5057568550109863, + "loss_wm_recon": 0.09765625, + "loss_wm_recon_w": 0.0244140625, + "step": 1341 + }, + { + "epoch": 0.5001863585538576, + "grad_norm": 17.084869384765625, + "learning_rate": 5.627795889067734e-06, + "loss": 10.092464447021484, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5571233034133911, + "loss_total_v6": 0.7459660768508911, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1342 + }, + { + "epoch": 0.5005590756615729, + "grad_norm": 17.553634643554688, + "learning_rate": 5.621519249493654e-06, + "loss": 10.050928115844727, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4686339199542999, + "loss_total_v6": 0.6409971714019775, + "loss_wm_recon": 0.103515625, + "loss_wm_recon_w": 0.02587890625, + "step": 1343 + }, + { + "epoch": 0.5009317927692881, + "grad_norm": 17.011211395263672, + "learning_rate": 5.615241614970546e-06, + "loss": 10.4750337600708, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6539735794067383, + "loss_total_v6": 0.8463563919067383, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1344 + }, + { + "epoch": 0.5013045098770034, + "grad_norm": 17.06108283996582, + "learning_rate": 5.608962995547856e-06, + "loss": 10.039405822753906, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.719345211982727, + "loss_total_v6": 0.893051266670227, + "loss_wm_recon": 0.11669921875, + "loss_wm_recon_w": 0.0291748046875, + "step": 1345 + }, + { + "epoch": 0.5016772269847186, + "grad_norm": 17.129228591918945, + "learning_rate": 5.6026834012766155e-06, + "loss": 10.65127944946289, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.47618231177330017, + "loss_total_v6": 0.6646589040756226, + "loss_wm_recon": 0.12109375, + "loss_wm_recon_w": 0.0302734375, + "step": 1346 + }, + { + "epoch": 0.5020499440924339, + "grad_norm": 16.474292755126953, + "learning_rate": 5.59640284220941e-06, + "loss": 9.860002517700195, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.8733617067337036, + "loss_total_v6": 1.0537816286087036, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1347 + }, + { + "epoch": 0.5024226612001491, + "grad_norm": 16.776615142822266, + "learning_rate": 5.590121328400367e-06, + "loss": 10.571758270263672, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6533851623535156, + "loss_total_v6": 0.8479652404785156, + "loss_wm_recon": 0.1455078125, + "loss_wm_recon_w": 0.036376953125, + "step": 1348 + }, + { + "epoch": 0.5027953783078644, + "grad_norm": 16.552703857421875, + "learning_rate": 5.583838869905151e-06, + "loss": 10.304553985595703, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.6029348373413086, + "loss_total_v6": 0.7859182357788086, + "loss_wm_recon": 0.11474609375, + "loss_wm_recon_w": 0.0286865234375, + "step": 1349 + }, + { + "epoch": 0.5031680954155796, + "grad_norm": 17.411012649536133, + "learning_rate": 5.577555476780933e-06, + "loss": 10.233346939086914, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.48485568165779114, + "loss_total_v6": 0.6578292846679688, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1350 + }, + { + "epoch": 0.5035408125232949, + "grad_norm": 17.188556671142578, + "learning_rate": 5.5712711590863775e-06, + "loss": 9.882758140563965, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3473011255264282, + "loss_total_v6": 0.5241810083389282, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1351 + }, + { + "epoch": 0.5039135296310101, + "grad_norm": 17.41350555419922, + "learning_rate": 5.564985926881638e-06, + "loss": 10.655367851257324, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.40537315607070923, + "loss_total_v6": 0.5767598748207092, + "loss_wm_recon": 0.107421875, + "loss_wm_recon_w": 0.02685546875, + "step": 1352 + }, + { + "epoch": 0.5042862467387254, + "grad_norm": 16.85430145263672, + "learning_rate": 5.558699790228321e-06, + "loss": 10.045042037963867, + "loss_alignment": 0.37890625, + "loss_alignment_w": 0.189453125, + "loss_sft": 0.2426208257675171, + "loss_total_v6": 0.4664977788925171, + "loss_wm_recon": 0.1376953125, + "loss_wm_recon_w": 0.034423828125, + "step": 1353 + }, + { + "epoch": 0.5046589638464406, + "grad_norm": 17.134469985961914, + "learning_rate": 5.552412759189491e-06, + "loss": 10.168112754821777, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.3852279484272003, + "loss_total_v6": 0.5586898326873779, + "loss_wm_recon": 0.10009765625, + "loss_wm_recon_w": 0.0250244140625, + "step": 1354 + }, + { + "epoch": 0.5050316809541558, + "grad_norm": 22.123640060424805, + "learning_rate": 5.54612484382964e-06, + "loss": 10.105599403381348, + "loss_alignment": 0.35546875, + "loss_alignment_w": 0.177734375, + "loss_sft": 0.5052927136421204, + "loss_total_v6": 0.7213571667671204, + "loss_wm_recon": 0.1533203125, + "loss_wm_recon_w": 0.038330078125, + "step": 1355 + }, + { + "epoch": 0.505404398061871, + "grad_norm": 17.330251693725586, + "learning_rate": 5.539836054214673e-06, + "loss": 9.93215560913086, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.37899458408355713, + "loss_total_v6": 0.5458647012710571, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1356 + }, + { + "epoch": 0.5057771151695862, + "grad_norm": 16.916601181030273, + "learning_rate": 5.533546400411899e-06, + "loss": 10.16706657409668, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5086618661880493, + "loss_total_v6": 0.6935983896255493, + "loss_wm_recon": 0.12255859375, + "loss_wm_recon_w": 0.0306396484375, + "step": 1357 + }, + { + "epoch": 0.5061498322773015, + "grad_norm": 16.826181411743164, + "learning_rate": 5.527255892490007e-06, + "loss": 10.1495361328125, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4042436182498932, + "loss_total_v6": 0.5852738618850708, + "loss_wm_recon": 0.10693359375, + "loss_wm_recon_w": 0.0267333984375, + "step": 1358 + }, + { + "epoch": 0.5065225493850167, + "grad_norm": 16.401124954223633, + "learning_rate": 5.520964540519055e-06, + "loss": 10.400693893432617, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4415476322174072, + "loss_total_v6": 0.6120798587799072, + "loss_wm_recon": 0.09619140625, + "loss_wm_recon_w": 0.0240478515625, + "step": 1359 + }, + { + "epoch": 0.506895266492732, + "grad_norm": 16.393606185913086, + "learning_rate": 5.5146723545704515e-06, + "loss": 10.451128005981445, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.410399466753006, + "loss_total_v6": 0.5860586166381836, + "loss_wm_recon": 0.10107421875, + "loss_wm_recon_w": 0.0252685546875, + "step": 1360 + }, + { + "epoch": 0.5072679836004472, + "grad_norm": 17.396493911743164, + "learning_rate": 5.508379344716943e-06, + "loss": 10.228527069091797, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.4362643361091614, + "loss_total_v6": 0.6121676564216614, + "loss_wm_recon": 0.10205078125, + "loss_wm_recon_w": 0.0255126953125, + "step": 1361 + }, + { + "epoch": 0.5076407007081625, + "grad_norm": 16.4190731048584, + "learning_rate": 5.502085521032589e-06, + "loss": 9.995100021362305, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5252956748008728, + "loss_total_v6": 0.7081570029258728, + "loss_wm_recon": 0.1064453125, + "loss_wm_recon_w": 0.026611328125, + "step": 1362 + }, + { + "epoch": 0.5080134178158777, + "grad_norm": 16.776058197021484, + "learning_rate": 5.4957908935927594e-06, + "loss": 10.218711853027344, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6231021881103516, + "loss_total_v6": 0.8154850006103516, + "loss_wm_recon": 0.12890625, + "loss_wm_recon_w": 0.0322265625, + "step": 1363 + }, + { + "epoch": 0.508386134923593, + "grad_norm": 16.839378356933594, + "learning_rate": 5.489495472474103e-06, + "loss": 10.465564727783203, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.39827755093574524, + "loss_total_v6": 0.5735704898834229, + "loss_wm_recon": 0.107421875, + "loss_wm_recon_w": 0.02685546875, + "step": 1364 + }, + { + "epoch": 0.5087588520313082, + "grad_norm": 16.235549926757812, + "learning_rate": 5.483199267754544e-06, + "loss": 10.086824417114258, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.4017825126647949, + "loss_total_v6": 0.5676760673522949, + "loss_wm_recon": 0.10107421875, + "loss_wm_recon_w": 0.0252685546875, + "step": 1365 + }, + { + "epoch": 0.5091315691390235, + "grad_norm": 17.17607307434082, + "learning_rate": 5.476902289513261e-06, + "loss": 9.96031379699707, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.24184350669384003, + "loss_total_v6": 0.4364235997200012, + "loss_wm_recon": 0.1298828125, + "loss_wm_recon_w": 0.032470703125, + "step": 1366 + }, + { + "epoch": 0.5095042862467387, + "grad_norm": 17.682065963745117, + "learning_rate": 5.470604547830667e-06, + "loss": 10.17663860321045, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.7439139485359192, + "loss_total_v6": 0.7751639485359192, + "loss_wm_recon": 0.125, + "loss_wm_recon_w": 0.03125, + "step": 1367 + }, + { + "epoch": 0.509877003354454, + "grad_norm": 17.97627067565918, + "learning_rate": 5.4643060527884e-06, + "loss": 10.308371543884277, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.1143956258893013, + "loss_total_v6": 0.2850499153137207, + "loss_wm_recon": 0.0966796875, + "loss_wm_recon_w": 0.024169921875, + "step": 1368 + }, + { + "epoch": 0.5102497204621692, + "grad_norm": 16.903657913208008, + "learning_rate": 5.4580068144693075e-06, + "loss": 10.324789047241211, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.30463284254074097, + "loss_total_v6": 0.47528713941574097, + "loss_wm_recon": 0.1044921875, + "loss_wm_recon_w": 0.026123046875, + "step": 1369 + }, + { + "epoch": 0.5106224375698845, + "grad_norm": 17.20456886291504, + "learning_rate": 5.4517068429574215e-06, + "loss": 10.531135559082031, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.5248925089836121, + "loss_total_v6": 0.5544335246086121, + "loss_wm_recon": 0.1181640625, + "loss_wm_recon_w": 0.029541015625, + "step": 1370 + }, + { + "epoch": 0.5109951546775997, + "grad_norm": 17.938201904296875, + "learning_rate": 5.445406148337947e-06, + "loss": 9.940923690795898, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.6935635209083557, + "loss_total_v6": 0.8673916459083557, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1371 + }, + { + "epoch": 0.511367871785315, + "grad_norm": 17.156076431274414, + "learning_rate": 5.439104740697252e-06, + "loss": 9.934163093566895, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5473647117614746, + "loss_total_v6": 0.7261977195739746, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1372 + }, + { + "epoch": 0.5117405888930302, + "grad_norm": 16.728878021240234, + "learning_rate": 5.432802630122841e-06, + "loss": 10.180205345153809, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5437478423118591, + "loss_total_v6": 0.7141579985618591, + "loss_wm_recon": 0.103515625, + "loss_wm_recon_w": 0.02587890625, + "step": 1373 + }, + { + "epoch": 0.5121133060007454, + "grad_norm": 17.643980026245117, + "learning_rate": 5.4264998267033445e-06, + "loss": 9.827570915222168, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4378565549850464, + "loss_total_v6": 0.6132715940475464, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1374 + }, + { + "epoch": 0.5124860231084607, + "grad_norm": 17.024789810180664, + "learning_rate": 5.420196340528507e-06, + "loss": 10.106704711914062, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4119548201560974, + "loss_total_v6": 0.5888347029685974, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1375 + }, + { + "epoch": 0.512858740216176, + "grad_norm": 17.77651023864746, + "learning_rate": 5.413892181689161e-06, + "loss": 10.458944320678711, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4361637234687805, + "loss_total_v6": 0.6168277859687805, + "loss_wm_recon": 0.10546875, + "loss_wm_recon_w": 0.0263671875, + "step": 1376 + }, + { + "epoch": 0.5132314573238912, + "grad_norm": 17.156265258789062, + "learning_rate": 5.407587360277215e-06, + "loss": 10.169942855834961, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.33759066462516785, + "loss_total_v6": 0.5152029991149902, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1377 + }, + { + "epoch": 0.5136041744316064, + "grad_norm": 17.25198745727539, + "learning_rate": 5.401281886385643e-06, + "loss": 10.34376049041748, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.06478101760149002, + "loss_total_v6": 0.23567944765090942, + "loss_wm_recon": 0.08984375, + "loss_wm_recon_w": 0.0224609375, + "step": 1378 + }, + { + "epoch": 0.5139768915393217, + "grad_norm": 17.945472717285156, + "learning_rate": 5.394975770108459e-06, + "loss": 10.770307540893555, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.47837379574775696, + "loss_total_v6": 0.6569626331329346, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1379 + }, + { + "epoch": 0.5143496086470369, + "grad_norm": 16.851024627685547, + "learning_rate": 5.388669021540707e-06, + "loss": 10.477510452270508, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.09585303068161011, + "loss_total_v6": 0.2713901400566101, + "loss_wm_recon": 0.1005859375, + "loss_wm_recon_w": 0.025146484375, + "step": 1380 + }, + { + "epoch": 0.5147223257547522, + "grad_norm": 17.168701171875, + "learning_rate": 5.382361650778445e-06, + "loss": 10.133992195129395, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.45041656494140625, + "loss_total_v6": 0.6362075805664062, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1381 + }, + { + "epoch": 0.5150950428624674, + "grad_norm": 17.22865104675293, + "learning_rate": 5.376053667918726e-06, + "loss": 10.394444465637207, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3774145245552063, + "loss_total_v6": 0.5430639386177063, + "loss_wm_recon": 0.10009765625, + "loss_wm_recon_w": 0.0250244140625, + "step": 1382 + }, + { + "epoch": 0.5154677599701827, + "grad_norm": 18.028812408447266, + "learning_rate": 5.3697450830595775e-06, + "loss": 10.37080192565918, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.49150195717811584, + "loss_total_v6": 0.6776591539382935, + "loss_wm_recon": 0.11181640625, + "loss_wm_recon_w": 0.0279541015625, + "step": 1383 + }, + { + "epoch": 0.5158404770778979, + "grad_norm": 18.100452423095703, + "learning_rate": 5.363435906300001e-06, + "loss": 9.996591567993164, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4109286963939667, + "loss_total_v6": 0.4382724463939667, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1384 + }, + { + "epoch": 0.5162131941856131, + "grad_norm": 16.524654388427734, + "learning_rate": 5.3571261477399375e-06, + "loss": 10.349093437194824, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.4623173475265503, + "loss_total_v6": 0.6444462537765503, + "loss_wm_recon": 0.111328125, + "loss_wm_recon_w": 0.02783203125, + "step": 1385 + }, + { + "epoch": 0.5165859112933283, + "grad_norm": 17.48662757873535, + "learning_rate": 5.3508158174802615e-06, + "loss": 10.449224472045898, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.4086511433124542, + "loss_total_v6": 0.5954186916351318, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1386 + }, + { + "epoch": 0.5169586284010436, + "grad_norm": 19.082412719726562, + "learning_rate": 5.344504925622763e-06, + "loss": 10.174187660217285, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5208583474159241, + "loss_total_v6": 0.7017665505409241, + "loss_wm_recon": 0.0986328125, + "loss_wm_recon_w": 0.024658203125, + "step": 1387 + }, + { + "epoch": 0.5173313455087588, + "grad_norm": 17.855915069580078, + "learning_rate": 5.338193482270132e-06, + "loss": 10.169182777404785, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.5798107385635376, + "loss_total_v6": 0.7511974573135376, + "loss_wm_recon": 0.115234375, + "loss_wm_recon_w": 0.02880859375, + "step": 1388 + }, + { + "epoch": 0.5177040626164741, + "grad_norm": 18.178749084472656, + "learning_rate": 5.33188149752594e-06, + "loss": 10.189960479736328, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5445351004600525, + "loss_total_v6": 0.7155556082725525, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1389 + }, + { + "epoch": 0.5180767797241893, + "grad_norm": 16.162416458129883, + "learning_rate": 5.325568981494627e-06, + "loss": 9.891843795776367, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.2200716733932495, + "loss_total_v6": 0.3923128843307495, + "loss_wm_recon": 0.11083984375, + "loss_wm_recon_w": 0.0277099609375, + "step": 1390 + }, + { + "epoch": 0.5184494968319046, + "grad_norm": 19.097801208496094, + "learning_rate": 5.319255944281479e-06, + "loss": 10.562864303588867, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.377565860748291, + "loss_total_v6": 0.556398868560791, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1391 + }, + { + "epoch": 0.5188222139396198, + "grad_norm": 17.391277313232422, + "learning_rate": 5.312942395992621e-06, + "loss": 10.28958797454834, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.33605411648750305, + "loss_total_v6": 0.5076849460601807, + "loss_wm_recon": 0.1083984375, + "loss_wm_recon_w": 0.027099609375, + "step": 1392 + }, + { + "epoch": 0.519194931047335, + "grad_norm": 17.23242950439453, + "learning_rate": 5.306628346734995e-06, + "loss": 10.396268844604492, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5766513347625732, + "loss_total_v6": 0.7663486003875732, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1393 + }, + { + "epoch": 0.5195676481550503, + "grad_norm": 16.84100914001465, + "learning_rate": 5.300313806616344e-06, + "loss": 10.529237747192383, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.42187580466270447, + "loss_total_v6": 0.5871590375900269, + "loss_wm_recon": 0.0908203125, + "loss_wm_recon_w": 0.022705078125, + "step": 1394 + }, + { + "epoch": 0.5199403652627655, + "grad_norm": 17.073883056640625, + "learning_rate": 5.293998785745198e-06, + "loss": 10.822439193725586, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.5745688080787659, + "loss_total_v6": 0.7427816987037659, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1395 + }, + { + "epoch": 0.5203130823704808, + "grad_norm": 16.608360290527344, + "learning_rate": 5.287683294230855e-06, + "loss": 10.30115795135498, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.39631134271621704, + "loss_total_v6": 0.577829897403717, + "loss_wm_recon": 0.11669921875, + "loss_wm_recon_w": 0.0291748046875, + "step": 1396 + }, + { + "epoch": 0.520685799478196, + "grad_norm": 17.317058563232422, + "learning_rate": 5.281367342183368e-06, + "loss": 9.952058792114258, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.3583812415599823, + "loss_total_v6": 0.5458812713623047, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1397 + }, + { + "epoch": 0.5210585165859113, + "grad_norm": 17.013647079467773, + "learning_rate": 5.2750509397135276e-06, + "loss": 9.728972434997559, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.1685602366924286, + "loss_total_v6": 0.3726617991924286, + "loss_wm_recon": 0.14453125, + "loss_wm_recon_w": 0.0361328125, + "step": 1398 + }, + { + "epoch": 0.5214312336936265, + "grad_norm": 16.912878036499023, + "learning_rate": 5.268734096932842e-06, + "loss": 10.494929313659668, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.7940477728843689, + "loss_total_v6": 0.9896044135093689, + "loss_wm_recon": 0.1337890625, + "loss_wm_recon_w": 0.033447265625, + "step": 1399 + }, + { + "epoch": 0.5218039508013418, + "grad_norm": 16.92681121826172, + "learning_rate": 5.262416823953528e-06, + "loss": 10.114194869995117, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.364433616399765, + "loss_total_v6": 0.5599902868270874, + "loss_wm_recon": 0.1259765625, + "loss_wm_recon_w": 0.031494140625, + "step": 1400 + }, + { + "epoch": 0.5218039508013418, + "eval_loss": 0.630719006061554, + "eval_loss_alignment": 0.2815504807692308, + "eval_loss_alignment_w": 0.1407752403846154, + "eval_loss_sft": 0.4638918252407104, + "eval_loss_total_v6": 0.630718969907134, + "eval_loss_wm_recon": 0.10420761760483441, + "eval_loss_wm_recon_w": 0.026051904401208602, + "eval_runtime": 228.6228, + "eval_samples_per_second": 8.162, + "eval_steps_per_second": 1.024, + "step": 1400 + }, + { + "epoch": 0.522176667909057, + "grad_norm": 16.780054092407227, + "learning_rate": 5.256099130888489e-06, + "loss": 10.302962303161621, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5299973487854004, + "loss_total_v6": 0.7129807472229004, + "loss_wm_recon": 0.10693359375, + "loss_wm_recon_w": 0.0267333984375, + "step": 1401 + }, + { + "epoch": 0.5225493850167723, + "grad_norm": 16.365549087524414, + "learning_rate": 5.249781027851303e-06, + "loss": 10.278633117675781, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.6719831228256226, + "loss_total_v6": 0.8641217947006226, + "loss_wm_recon": 0.1201171875, + "loss_wm_recon_w": 0.030029296875, + "step": 1402 + }, + { + "epoch": 0.5229221021244875, + "grad_norm": 16.391220092773438, + "learning_rate": 5.243462524956202e-06, + "loss": 10.129487037658691, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5742408037185669, + "loss_total_v6": 0.7475806474685669, + "loss_wm_recon": 0.115234375, + "loss_wm_recon_w": 0.02880859375, + "step": 1403 + }, + { + "epoch": 0.5232948192322028, + "grad_norm": 17.5615291595459, + "learning_rate": 5.23714363231806e-06, + "loss": 10.123636245727539, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5920219421386719, + "loss_total_v6": 0.7811088562011719, + "loss_wm_recon": 0.12353515625, + "loss_wm_recon_w": 0.0308837890625, + "step": 1404 + }, + { + "epoch": 0.523667536339918, + "grad_norm": 17.44624900817871, + "learning_rate": 5.230824360052375e-06, + "loss": 10.748806953430176, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.4362679719924927, + "loss_total_v6": 0.6025277376174927, + "loss_wm_recon": 0.0947265625, + "loss_wm_recon_w": 0.023681640625, + "step": 1405 + }, + { + "epoch": 0.5240402534476333, + "grad_norm": 17.48640251159668, + "learning_rate": 5.224504718275248e-06, + "loss": 10.206950187683105, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.46445417404174805, + "loss_total_v6": 0.649512767791748, + "loss_wm_recon": 0.115234375, + "loss_wm_recon_w": 0.02880859375, + "step": 1406 + }, + { + "epoch": 0.5244129705553485, + "grad_norm": 16.658931732177734, + "learning_rate": 5.218184717103378e-06, + "loss": 10.040266036987305, + "loss_alignment": 0.3359375, + "loss_alignment_w": 0.16796875, + "loss_sft": 0.3948005437850952, + "loss_total_v6": 0.5984138250350952, + "loss_wm_recon": 0.142578125, + "loss_wm_recon_w": 0.03564453125, + "step": 1407 + }, + { + "epoch": 0.5247856876630638, + "grad_norm": 17.214384078979492, + "learning_rate": 5.211864366654034e-06, + "loss": 10.712383270263672, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5873196721076965, + "loss_total_v6": 0.7760403752326965, + "loss_wm_recon": 0.1220703125, + "loss_wm_recon_w": 0.030517578125, + "step": 1408 + }, + { + "epoch": 0.525158404770779, + "grad_norm": 17.030059814453125, + "learning_rate": 5.20554367704505e-06, + "loss": 10.60942268371582, + "loss_alignment": 0.3515625, + "loss_alignment_w": 0.17578125, + "loss_sft": 0.5287947058677673, + "loss_total_v6": 0.7414411902427673, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 1409 + }, + { + "epoch": 0.5255311218784943, + "grad_norm": 16.394262313842773, + "learning_rate": 5.199222658394798e-06, + "loss": 10.145343780517578, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6800296902656555, + "loss_total_v6": 0.8702152371406555, + "loss_wm_recon": 0.1279296875, + "loss_wm_recon_w": 0.031982421875, + "step": 1410 + }, + { + "epoch": 0.5259038389862095, + "grad_norm": 17.153291702270508, + "learning_rate": 5.192901320822179e-06, + "loss": 9.88950252532959, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.35250183939933777, + "loss_total_v6": 0.5168085098266602, + "loss_wm_recon": 0.1025390625, + "loss_wm_recon_w": 0.025634765625, + "step": 1411 + }, + { + "epoch": 0.5262765560939248, + "grad_norm": 17.919509887695312, + "learning_rate": 5.1865796744466026e-06, + "loss": 9.801154136657715, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4509769082069397, + "loss_total_v6": 0.6229739785194397, + "loss_wm_recon": 0.10205078125, + "loss_wm_recon_w": 0.0255126953125, + "step": 1412 + }, + { + "epoch": 0.5266492732016399, + "grad_norm": 17.269332885742188, + "learning_rate": 5.180257729387976e-06, + "loss": 10.397186279296875, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.16617879271507263, + "loss_total_v6": 0.35258015990257263, + "loss_wm_recon": 0.10498046875, + "loss_wm_recon_w": 0.0262451171875, + "step": 1413 + }, + { + "epoch": 0.5270219903093551, + "grad_norm": 17.129371643066406, + "learning_rate": 5.173935495766681e-06, + "loss": 10.150861740112305, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.8570056557655334, + "loss_total_v6": 1.0326647758483887, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1414 + }, + { + "epoch": 0.5273947074170704, + "grad_norm": 17.350862503051758, + "learning_rate": 5.167612983703563e-06, + "loss": 10.555521011352539, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.14118927717208862, + "loss_total_v6": 0.3113552927970886, + "loss_wm_recon": 0.1025390625, + "loss_wm_recon_w": 0.025634765625, + "step": 1415 + }, + { + "epoch": 0.5277674245247856, + "grad_norm": 17.2403621673584, + "learning_rate": 5.161290203319914e-06, + "loss": 10.257464408874512, + "loss_alignment": 0.3828125, + "loss_alignment_w": 0.19140625, + "loss_sft": 0.6491318941116333, + "loss_total_v6": 0.8776475191116333, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 1416 + }, + { + "epoch": 0.5281401416325009, + "grad_norm": 17.448965072631836, + "learning_rate": 5.1549671647374545e-06, + "loss": 10.369327545166016, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.5999833941459656, + "loss_total_v6": 0.7937089800834656, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1417 + }, + { + "epoch": 0.5285128587402161, + "grad_norm": 17.06736946105957, + "learning_rate": 5.148643878078318e-06, + "loss": 10.446495056152344, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5597270131111145, + "loss_total_v6": 0.7388041615486145, + "loss_wm_recon": 0.11474609375, + "loss_wm_recon_w": 0.0286865234375, + "step": 1418 + }, + { + "epoch": 0.5288855758479314, + "grad_norm": 16.708959579467773, + "learning_rate": 5.142320353465037e-06, + "loss": 9.667779922485352, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.19763851165771484, + "loss_total_v6": 0.38159847259521484, + "loss_wm_recon": 0.11865234375, + "loss_wm_recon_w": 0.0296630859375, + "step": 1419 + }, + { + "epoch": 0.5292582929556466, + "grad_norm": 17.430252075195312, + "learning_rate": 5.13599660102052e-06, + "loss": 10.39162826538086, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.22789232432842255, + "loss_total_v6": 0.40965503454208374, + "loss_wm_recon": 0.11767578125, + "loss_wm_recon_w": 0.0294189453125, + "step": 1420 + }, + { + "epoch": 0.5296310100633619, + "grad_norm": 17.042932510375977, + "learning_rate": 5.1296726308680455e-06, + "loss": 10.424389839172363, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.10810207575559616, + "loss_total_v6": 0.29059720039367676, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1421 + }, + { + "epoch": 0.5300037271710771, + "grad_norm": 17.112728118896484, + "learning_rate": 5.123348453131242e-06, + "loss": 9.888665199279785, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6075462102890015, + "loss_total_v6": 0.6365989446640015, + "loss_wm_recon": 0.1162109375, + "loss_wm_recon_w": 0.029052734375, + "step": 1422 + }, + { + "epoch": 0.5303764442787924, + "grad_norm": 17.229894638061523, + "learning_rate": 5.117024077934062e-06, + "loss": 10.083894729614258, + "loss_alignment": 0.27734375, + "loss_alignment_w": 0.138671875, + "loss_sft": 0.5662339329719543, + "loss_total_v6": 0.7301743626594543, + "loss_wm_recon": 0.10107421875, + "loss_wm_recon_w": 0.0252685546875, + "step": 1423 + }, + { + "epoch": 0.5307491613865076, + "grad_norm": 16.9143123626709, + "learning_rate": 5.110699515400782e-06, + "loss": 10.306045532226562, + "loss_alignment": 0.3671875, + "loss_alignment_w": 0.18359375, + "loss_sft": 0.7261213064193726, + "loss_total_v6": 0.9468244314193726, + "loss_wm_recon": 0.1484375, + "loss_wm_recon_w": 0.037109375, + "step": 1424 + }, + { + "epoch": 0.5311218784942229, + "grad_norm": 16.920337677001953, + "learning_rate": 5.104374775655976e-06, + "loss": 9.913840293884277, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.33974581956863403, + "loss_total_v6": 0.524194061756134, + "loss_wm_recon": 0.12060546875, + "loss_wm_recon_w": 0.0301513671875, + "step": 1425 + }, + { + "epoch": 0.5314945956019381, + "grad_norm": 17.415847778320312, + "learning_rate": 5.098049868824499e-06, + "loss": 9.883225440979004, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.6458891034126282, + "loss_total_v6": 0.8281400799751282, + "loss_wm_recon": 0.11962890625, + "loss_wm_recon_w": 0.0299072265625, + "step": 1426 + }, + { + "epoch": 0.5318673127096534, + "grad_norm": 17.43448257446289, + "learning_rate": 5.091724805031476e-06, + "loss": 10.056139945983887, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.6078718304634094, + "loss_total_v6": 0.7983015179634094, + "loss_wm_recon": 0.12109375, + "loss_wm_recon_w": 0.0302734375, + "step": 1427 + }, + { + "epoch": 0.5322400298173686, + "grad_norm": 17.142066955566406, + "learning_rate": 5.085399594402285e-06, + "loss": 9.951634407043457, + "loss_alignment": 0.328125, + "loss_alignment_w": 0.1640625, + "loss_sft": 0.47471022605895996, + "loss_total_v6": 0.67392897605896, + "loss_wm_recon": 0.140625, + "loss_wm_recon_w": 0.03515625, + "step": 1428 + }, + { + "epoch": 0.5326127469250839, + "grad_norm": 17.274564743041992, + "learning_rate": 5.079074247062534e-06, + "loss": 10.184625625610352, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.4886099100112915, + "loss_total_v6": 0.6516958475112915, + "loss_wm_recon": 0.08984375, + "loss_wm_recon_w": 0.0224609375, + "step": 1429 + }, + { + "epoch": 0.5329854640327991, + "grad_norm": 17.270706176757812, + "learning_rate": 5.072748773138057e-06, + "loss": 10.457313537597656, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5589438080787659, + "loss_total_v6": 0.7471762299537659, + "loss_wm_recon": 0.1201171875, + "loss_wm_recon_w": 0.030029296875, + "step": 1430 + }, + { + "epoch": 0.5333581811405144, + "grad_norm": 16.440441131591797, + "learning_rate": 5.066423182754885e-06, + "loss": 10.859634399414062, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.7021427154541016, + "loss_total_v6": 0.8912296295166016, + "loss_wm_recon": 0.11572265625, + "loss_wm_recon_w": 0.0289306640625, + "step": 1431 + }, + { + "epoch": 0.5337308982482296, + "grad_norm": 17.42292594909668, + "learning_rate": 5.060097486039234e-06, + "loss": 10.277467727661133, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4300083518028259, + "loss_total_v6": 0.6116489768028259, + "loss_wm_recon": 0.1171875, + "loss_wm_recon_w": 0.029296875, + "step": 1432 + }, + { + "epoch": 0.5341036153559449, + "grad_norm": 17.808074951171875, + "learning_rate": 5.053771693117497e-06, + "loss": 10.638483047485352, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.4869716167449951, + "loss_total_v6": 0.5124843120574951, + "loss_wm_recon": 0.10205078125, + "loss_wm_recon_w": 0.0255126953125, + "step": 1433 + }, + { + "epoch": 0.5344763324636601, + "grad_norm": 16.867521286010742, + "learning_rate": 5.047445814116215e-06, + "loss": 10.396547317504883, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.18558084964752197, + "loss_total_v6": 0.374179482460022, + "loss_wm_recon": 0.12158203125, + "loss_wm_recon_w": 0.0303955078125, + "step": 1434 + }, + { + "epoch": 0.5348490495713754, + "grad_norm": 17.092636108398438, + "learning_rate": 5.041119859162068e-06, + "loss": 10.353718757629395, + "loss_alignment": 0.2734375, + "loss_alignment_w": 0.13671875, + "loss_sft": 0.5047329664230347, + "loss_total_v6": 0.6651333570480347, + "loss_wm_recon": 0.0947265625, + "loss_wm_recon_w": 0.023681640625, + "step": 1435 + }, + { + "epoch": 0.5352217666790906, + "grad_norm": 16.447092056274414, + "learning_rate": 5.0347938383818615e-06, + "loss": 10.241361618041992, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.5111874938011169, + "loss_total_v6": 0.6778134703636169, + "loss_wm_recon": 0.10400390625, + "loss_wm_recon_w": 0.0260009765625, + "step": 1436 + }, + { + "epoch": 0.5355944837868059, + "grad_norm": 16.640155792236328, + "learning_rate": 5.028467761902504e-06, + "loss": 10.298415184020996, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.11278899013996124, + "loss_total_v6": 0.28564053773880005, + "loss_wm_recon": 0.10546875, + "loss_wm_recon_w": 0.0263671875, + "step": 1437 + }, + { + "epoch": 0.5359672008945211, + "grad_norm": 17.732738494873047, + "learning_rate": 5.022141639850988e-06, + "loss": 10.091898918151855, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.3533915877342224, + "loss_total_v6": 0.5314921736717224, + "loss_wm_recon": 0.11083984375, + "loss_wm_recon_w": 0.0277099609375, + "step": 1438 + }, + { + "epoch": 0.5363399180022363, + "grad_norm": 17.656518936157227, + "learning_rate": 5.015815482354387e-06, + "loss": 10.269341468811035, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.6087992191314697, + "loss_total_v6": 0.7934916019439697, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1439 + }, + { + "epoch": 0.5367126351099516, + "grad_norm": 16.688953399658203, + "learning_rate": 5.009489299539827e-06, + "loss": 9.940900802612305, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.17115017771720886, + "loss_total_v6": 0.34595486521720886, + "loss_wm_recon": 0.08984375, + "loss_wm_recon_w": 0.0224609375, + "step": 1440 + }, + { + "epoch": 0.5370853522176668, + "grad_norm": 16.54940414428711, + "learning_rate": 5.003163101534476e-06, + "loss": 10.187724113464355, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.2685883641242981, + "loss_total_v6": 0.4476655125617981, + "loss_wm_recon": 0.09912109375, + "loss_wm_recon_w": 0.0247802734375, + "step": 1441 + }, + { + "epoch": 0.537458069325382, + "grad_norm": 16.964208602905273, + "learning_rate": 4.996836898465525e-06, + "loss": 9.958062171936035, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.8348045945167542, + "loss_total_v6": 1.0210838317871094, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1442 + }, + { + "epoch": 0.5378307864330972, + "grad_norm": 17.50809097290039, + "learning_rate": 4.9905107004601745e-06, + "loss": 10.207344055175781, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5852465629577637, + "loss_total_v6": 0.7562670707702637, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1443 + }, + { + "epoch": 0.5382035035408125, + "grad_norm": 17.66860580444336, + "learning_rate": 4.984184517645614e-06, + "loss": 10.507787704467773, + "loss_alignment": 0.375, + "loss_alignment_w": 0.1875, + "loss_sft": 0.4945456385612488, + "loss_total_v6": 0.7189108729362488, + "loss_wm_recon": 0.1474609375, + "loss_wm_recon_w": 0.036865234375, + "step": 1444 + }, + { + "epoch": 0.5385762206485277, + "grad_norm": 16.55927085876465, + "learning_rate": 4.9778583601490125e-06, + "loss": 9.969511032104492, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5340031981468201, + "loss_total_v6": 0.7213811278343201, + "loss_wm_recon": 0.11669921875, + "loss_wm_recon_w": 0.0291748046875, + "step": 1445 + }, + { + "epoch": 0.538948937756243, + "grad_norm": 17.279499053955078, + "learning_rate": 4.971532238097498e-06, + "loss": 10.239740371704102, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.138638436794281, + "loss_total_v6": 0.324429452419281, + "loss_wm_recon": 0.1025390625, + "loss_wm_recon_w": 0.025634765625, + "step": 1446 + }, + { + "epoch": 0.5393216548639582, + "grad_norm": 16.601478576660156, + "learning_rate": 4.965206161618138e-06, + "loss": 9.827241897583008, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.631016194820404, + "loss_total_v6": 0.815708577632904, + "loss_wm_recon": 0.11376953125, + "loss_wm_recon_w": 0.0284423828125, + "step": 1447 + }, + { + "epoch": 0.5396943719716735, + "grad_norm": 17.419506072998047, + "learning_rate": 4.958880140837934e-06, + "loss": 10.075591087341309, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4491579234600067, + "loss_total_v6": 0.6190798282623291, + "loss_wm_recon": 0.1015625, + "loss_wm_recon_w": 0.025390625, + "step": 1448 + }, + { + "epoch": 0.5400670890793887, + "grad_norm": 17.219751358032227, + "learning_rate": 4.9525541858837865e-06, + "loss": 9.887103080749512, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.2714982032775879, + "loss_total_v6": 0.4414200782775879, + "loss_wm_recon": 0.09375, + "loss_wm_recon_w": 0.0234375, + "step": 1449 + }, + { + "epoch": 0.540439806187104, + "grad_norm": 17.098310470581055, + "learning_rate": 4.946228306882506e-06, + "loss": 10.080693244934082, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.4498211741447449, + "loss_total_v6": 0.6336590647697449, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1450 + }, + { + "epoch": 0.5408125232948192, + "grad_norm": 17.72126579284668, + "learning_rate": 4.9399025139607675e-06, + "loss": 10.043828964233398, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5104578137397766, + "loss_total_v6": 0.6973474621772766, + "loss_wm_recon": 0.11474609375, + "loss_wm_recon_w": 0.0286865234375, + "step": 1451 + }, + { + "epoch": 0.5411852404025345, + "grad_norm": 17.005563735961914, + "learning_rate": 4.933576817245117e-06, + "loss": 10.416884422302246, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.6262914538383484, + "loss_total_v6": 0.8107396960258484, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1452 + }, + { + "epoch": 0.5415579575102497, + "grad_norm": 16.525440216064453, + "learning_rate": 4.927251226861943e-06, + "loss": 9.940679550170898, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.3758538067340851, + "loss_total_v6": 0.5478508472442627, + "loss_wm_recon": 0.10986328125, + "loss_wm_recon_w": 0.0274658203125, + "step": 1453 + }, + { + "epoch": 0.541930674617965, + "grad_norm": 17.50279426574707, + "learning_rate": 4.9209257529374656e-06, + "loss": 10.533735275268555, + "loss_alignment": 0.31640625, + "loss_alignment_w": 0.158203125, + "loss_sft": 0.5021689534187317, + "loss_total_v6": 0.6952841877937317, + "loss_wm_recon": 0.1396484375, + "loss_wm_recon_w": 0.034912109375, + "step": 1454 + }, + { + "epoch": 0.5423033917256802, + "grad_norm": 17.12127113342285, + "learning_rate": 4.914600405597717e-06, + "loss": 10.309932708740234, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.27278435230255127, + "loss_total_v6": 0.45003044605255127, + "loss_wm_recon": 0.099609375, + "loss_wm_recon_w": 0.02490234375, + "step": 1455 + }, + { + "epoch": 0.5426761088333955, + "grad_norm": 17.03752899169922, + "learning_rate": 4.908275194968525e-06, + "loss": 10.068859100341797, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.27589014172554016, + "loss_total_v6": 0.45081689953804016, + "loss_wm_recon": 0.09814453125, + "loss_wm_recon_w": 0.0245361328125, + "step": 1456 + }, + { + "epoch": 0.5430488259411107, + "grad_norm": 16.547481536865234, + "learning_rate": 4.901950131175504e-06, + "loss": 10.056930541992188, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.42732933163642883, + "loss_total_v6": 0.6011574268341064, + "loss_wm_recon": 0.1015625, + "loss_wm_recon_w": 0.025390625, + "step": 1457 + }, + { + "epoch": 0.543421543048826, + "grad_norm": 17.486995697021484, + "learning_rate": 4.895625224344026e-06, + "loss": 10.107830047607422, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.40174850821495056, + "loss_total_v6": 0.571060061454773, + "loss_wm_recon": 0.09130859375, + "loss_wm_recon_w": 0.0228271484375, + "step": 1458 + }, + { + "epoch": 0.5437942601565412, + "grad_norm": 17.109329223632812, + "learning_rate": 4.889300484599218e-06, + "loss": 10.125511169433594, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3076169788837433, + "loss_total_v6": 0.4702146351337433, + "loss_wm_recon": 0.087890625, + "loss_wm_recon_w": 0.02197265625, + "step": 1459 + }, + { + "epoch": 0.5441669772642564, + "grad_norm": 16.875133514404297, + "learning_rate": 4.88297592206594e-06, + "loss": 10.195962905883789, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.14959509670734406, + "loss_total_v6": 0.32952672243118286, + "loss_wm_recon": 0.1103515625, + "loss_wm_recon_w": 0.027587890625, + "step": 1460 + }, + { + "epoch": 0.5445396943719717, + "grad_norm": 17.680713653564453, + "learning_rate": 4.876651546868759e-06, + "loss": 10.017012596130371, + "loss_alignment": 0.28515625, + "loss_alignment_w": 0.142578125, + "loss_sft": 0.6666998267173767, + "loss_total_v6": 0.8374761939048767, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1461 + }, + { + "epoch": 0.5449124114796869, + "grad_norm": 16.41217041015625, + "learning_rate": 4.870327369131956e-06, + "loss": 10.247119903564453, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.09213844686746597, + "loss_total_v6": 0.2668210566043854, + "loss_wm_recon": 0.08935546875, + "loss_wm_recon_w": 0.0223388671875, + "step": 1462 + }, + { + "epoch": 0.5452851285874022, + "grad_norm": 18.5192928314209, + "learning_rate": 4.864003398979482e-06, + "loss": 9.642123222351074, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.38810041546821594, + "loss_total_v6": 0.5553367137908936, + "loss_wm_recon": 0.1064453125, + "loss_wm_recon_w": 0.026611328125, + "step": 1463 + }, + { + "epoch": 0.5456578456951174, + "grad_norm": 18.142776489257812, + "learning_rate": 4.857679646534966e-06, + "loss": 9.888472557067871, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.7412986755371094, + "loss_total_v6": 0.9063377380371094, + "loss_wm_recon": 0.09765625, + "loss_wm_recon_w": 0.0244140625, + "step": 1464 + }, + { + "epoch": 0.5460305628028327, + "grad_norm": 17.544519424438477, + "learning_rate": 4.851356121921683e-06, + "loss": 10.843194007873535, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.48631608486175537, + "loss_total_v6": 0.6674684286117554, + "loss_wm_recon": 0.099609375, + "loss_wm_recon_w": 0.02490234375, + "step": 1465 + }, + { + "epoch": 0.5464032799105479, + "grad_norm": 17.050071716308594, + "learning_rate": 4.845032835262546e-06, + "loss": 10.24783992767334, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.49664509296417236, + "loss_total_v6": 0.6730366945266724, + "loss_wm_recon": 0.10400390625, + "loss_wm_recon_w": 0.0260009765625, + "step": 1466 + }, + { + "epoch": 0.5467759970182632, + "grad_norm": 17.178546905517578, + "learning_rate": 4.838709796680087e-06, + "loss": 10.044878005981445, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.2967475950717926, + "loss_total_v6": 0.4727729856967926, + "loss_wm_recon": 0.1025390625, + "loss_wm_recon_w": 0.025634765625, + "step": 1467 + }, + { + "epoch": 0.5471487141259784, + "grad_norm": 16.65440559387207, + "learning_rate": 4.832387016296438e-06, + "loss": 9.749710083007812, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.34611526131629944, + "loss_total_v6": 0.523117184638977, + "loss_wm_recon": 0.1064453125, + "loss_wm_recon_w": 0.026611328125, + "step": 1468 + }, + { + "epoch": 0.5475214312336937, + "grad_norm": 16.9832820892334, + "learning_rate": 4.8260645042333206e-06, + "loss": 10.677248001098633, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.42494580149650574, + "loss_total_v6": 0.5960884094238281, + "loss_wm_recon": 0.1064453125, + "loss_wm_recon_w": 0.026611328125, + "step": 1469 + }, + { + "epoch": 0.5478941483414088, + "grad_norm": 16.772672653198242, + "learning_rate": 4.819742270612025e-06, + "loss": 10.265817642211914, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.6992482542991638, + "loss_total_v6": 0.8650197386741638, + "loss_wm_recon": 0.1005859375, + "loss_wm_recon_w": 0.025146484375, + "step": 1470 + }, + { + "epoch": 0.5482668654491241, + "grad_norm": 16.922321319580078, + "learning_rate": 4.8134203255534e-06, + "loss": 10.30811882019043, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.5125373601913452, + "loss_total_v6": 0.6908820867538452, + "loss_wm_recon": 0.10400390625, + "loss_wm_recon_w": 0.0260009765625, + "step": 1471 + }, + { + "epoch": 0.5486395825568393, + "grad_norm": 17.504133224487305, + "learning_rate": 4.807098679177822e-06, + "loss": 10.256022453308105, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.4610460102558136, + "loss_total_v6": 0.6338975429534912, + "loss_wm_recon": 0.09765625, + "loss_wm_recon_w": 0.0244140625, + "step": 1472 + }, + { + "epoch": 0.5490122996645546, + "grad_norm": 17.48508071899414, + "learning_rate": 4.800777341605203e-06, + "loss": 9.886524200439453, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.19815996289253235, + "loss_total_v6": 0.37052324414253235, + "loss_wm_recon": 0.103515625, + "loss_wm_recon_w": 0.02587890625, + "step": 1473 + }, + { + "epoch": 0.5493850167722698, + "grad_norm": 17.578657150268555, + "learning_rate": 4.794456322954953e-06, + "loss": 9.960111618041992, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.4004424810409546, + "loss_total_v6": 0.5785430669784546, + "loss_wm_recon": 0.10302734375, + "loss_wm_recon_w": 0.0257568359375, + "step": 1474 + }, + { + "epoch": 0.549757733879985, + "grad_norm": 16.70738983154297, + "learning_rate": 4.788135633345968e-06, + "loss": 10.351863861083984, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.29814067482948303, + "loss_total_v6": 0.46952739357948303, + "loss_wm_recon": 0.099609375, + "loss_wm_recon_w": 0.02490234375, + "step": 1475 + }, + { + "epoch": 0.5501304509877003, + "grad_norm": 16.902799606323242, + "learning_rate": 4.781815282896626e-06, + "loss": 9.838668823242188, + "loss_alignment": 0.32421875, + "loss_alignment_w": 0.162109375, + "loss_sft": 0.4505279064178467, + "loss_total_v6": 0.6403472423553467, + "loss_wm_recon": 0.11083984375, + "loss_wm_recon_w": 0.0277099609375, + "step": 1476 + }, + { + "epoch": 0.5505031680954156, + "grad_norm": 16.993459701538086, + "learning_rate": 4.775495281724755e-06, + "loss": 10.034942626953125, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.39914971590042114, + "loss_total_v6": 0.5806682705879211, + "loss_wm_recon": 0.10888671875, + "loss_wm_recon_w": 0.0272216796875, + "step": 1477 + }, + { + "epoch": 0.5508758852031308, + "grad_norm": 16.736127853393555, + "learning_rate": 4.7691756399476285e-06, + "loss": 10.208099365234375, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.22980520129203796, + "loss_total_v6": 0.40168020129203796, + "loss_wm_recon": 0.09375, + "loss_wm_recon_w": 0.0234375, + "step": 1478 + }, + { + "epoch": 0.551248602310846, + "grad_norm": 17.351234436035156, + "learning_rate": 4.7628563676819415e-06, + "loss": 10.185823440551758, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5285663604736328, + "loss_total_v6": 0.7050800323486328, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1479 + }, + { + "epoch": 0.5516213194185613, + "grad_norm": 16.674076080322266, + "learning_rate": 4.756537475043798e-06, + "loss": 9.959480285644531, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.4990588128566742, + "loss_total_v6": 0.6715441942214966, + "loss_wm_recon": 0.11181640625, + "loss_wm_recon_w": 0.0279541015625, + "step": 1480 + }, + { + "epoch": 0.5519940365262765, + "grad_norm": 17.49563980102539, + "learning_rate": 4.750218972148698e-06, + "loss": 9.969337463378906, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.2059308886528015, + "loss_total_v6": 0.3900129199028015, + "loss_wm_recon": 0.126953125, + "loss_wm_recon_w": 0.03173828125, + "step": 1481 + }, + { + "epoch": 0.5523667536339918, + "grad_norm": 18.199268341064453, + "learning_rate": 4.743900869111512e-06, + "loss": 10.068344116210938, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.41711604595184326, + "loss_total_v6": 0.5905779600143433, + "loss_wm_recon": 0.10009765625, + "loss_wm_recon_w": 0.0250244140625, + "step": 1482 + }, + { + "epoch": 0.552739470741707, + "grad_norm": 17.441619873046875, + "learning_rate": 4.737583176046475e-06, + "loss": 10.36115837097168, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.5644321441650391, + "loss_total_v6": 0.7471714019775391, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1483 + }, + { + "epoch": 0.5531121878494223, + "grad_norm": 17.118545532226562, + "learning_rate": 4.731265903067159e-06, + "loss": 10.658059120178223, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.6410008072853088, + "loss_total_v6": 0.8139744400978088, + "loss_wm_recon": 0.10595703125, + "loss_wm_recon_w": 0.0264892578125, + "step": 1484 + }, + { + "epoch": 0.5534849049571375, + "grad_norm": 17.05489730834961, + "learning_rate": 4.724949060286476e-06, + "loss": 10.331992149353027, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.22299745678901672, + "loss_total_v6": 0.4036615192890167, + "loss_wm_recon": 0.10546875, + "loss_wm_recon_w": 0.0263671875, + "step": 1485 + }, + { + "epoch": 0.5538576220648528, + "grad_norm": 16.830881118774414, + "learning_rate": 4.718632657816634e-06, + "loss": 10.126501083374023, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.1496148407459259, + "loss_total_v6": 0.3162408173084259, + "loss_wm_recon": 0.08837890625, + "loss_wm_recon_w": 0.0220947265625, + "step": 1486 + }, + { + "epoch": 0.554230339172568, + "grad_norm": 17.18436622619629, + "learning_rate": 4.7123167057691446e-06, + "loss": 10.28664779663086, + "loss_alignment": 0.0, + "loss_alignment_w": 0.0, + "loss_sft": 0.6783028841018677, + "loss_total_v6": 0.7163888216018677, + "loss_wm_recon": 0.15234375, + "loss_wm_recon_w": 0.0380859375, + "step": 1487 + }, + { + "epoch": 0.5546030562802833, + "grad_norm": 16.95575523376465, + "learning_rate": 4.706001214254804e-06, + "loss": 10.387626647949219, + "loss_alignment": 0.28125, + "loss_alignment_w": 0.140625, + "loss_sft": 0.3242768943309784, + "loss_total_v6": 0.4893159568309784, + "loss_wm_recon": 0.09765625, + "loss_wm_recon_w": 0.0244140625, + "step": 1488 + }, + { + "epoch": 0.5549757733879985, + "grad_norm": 17.020174026489258, + "learning_rate": 4.699686193383657e-06, + "loss": 10.198393821716309, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.5796828866004944, + "loss_total_v6": 0.7543655037879944, + "loss_wm_recon": 0.09716796875, + "loss_wm_recon_w": 0.0242919921875, + "step": 1489 + }, + { + "epoch": 0.5553484904957138, + "grad_norm": 16.552915573120117, + "learning_rate": 4.693371653265007e-06, + "loss": 10.3776216506958, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.36892351508140564, + "loss_total_v6": 0.550808310508728, + "loss_wm_recon": 0.1025390625, + "loss_wm_recon_w": 0.025634765625, + "step": 1490 + }, + { + "epoch": 0.555721207603429, + "grad_norm": 17.40482521057129, + "learning_rate": 4.6870576040073805e-06, + "loss": 10.254693984985352, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.8493688106536865, + "loss_total_v6": 1.0296666622161865, + "loss_wm_recon": 0.10400390625, + "loss_wm_recon_w": 0.0260009765625, + "step": 1491 + }, + { + "epoch": 0.5560939247111443, + "grad_norm": 17.47450065612793, + "learning_rate": 4.680744055718524e-06, + "loss": 10.34219741821289, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.4443928897380829, + "loss_total_v6": 0.6183431148529053, + "loss_wm_recon": 0.10986328125, + "loss_wm_recon_w": 0.0274658203125, + "step": 1492 + }, + { + "epoch": 0.5564666418188595, + "grad_norm": 17.268291473388672, + "learning_rate": 4.674431018505376e-06, + "loss": 10.11860466003418, + "loss_alignment": 0.30859375, + "loss_alignment_w": 0.154296875, + "loss_sft": 0.5360517501831055, + "loss_total_v6": 0.7185468673706055, + "loss_wm_recon": 0.11279296875, + "loss_wm_recon_w": 0.0281982421875, + "step": 1493 + }, + { + "epoch": 0.5568393589265748, + "grad_norm": 16.590681076049805, + "learning_rate": 4.6681185024740604e-06, + "loss": 9.649552345275879, + "loss_alignment": 0.3125, + "loss_alignment_w": 0.15625, + "loss_sft": 0.20665031671524048, + "loss_total_v6": 0.3898778557777405, + "loss_wm_recon": 0.10791015625, + "loss_wm_recon_w": 0.0269775390625, + "step": 1494 + }, + { + "epoch": 0.55721207603429, + "grad_norm": 16.835283279418945, + "learning_rate": 4.661806517729869e-06, + "loss": 9.796398162841797, + "loss_alignment": 0.296875, + "loss_alignment_w": 0.1484375, + "loss_sft": 0.5242726802825928, + "loss_total_v6": 0.7007863521575928, + "loss_wm_recon": 0.1123046875, + "loss_wm_recon_w": 0.028076171875, + "step": 1495 + }, + { + "epoch": 0.5575847931420053, + "grad_norm": 17.012706756591797, + "learning_rate": 4.655495074377237e-06, + "loss": 10.377008438110352, + "loss_alignment": 0.30078125, + "loss_alignment_w": 0.150390625, + "loss_sft": 0.565509557723999, + "loss_total_v6": 0.744708776473999, + "loss_wm_recon": 0.115234375, + "loss_wm_recon_w": 0.02880859375, + "step": 1496 + }, + { + "epoch": 0.5579575102497205, + "grad_norm": 17.809965133666992, + "learning_rate": 4.649184182519742e-06, + "loss": 10.11745834350586, + "loss_alignment": 0.2890625, + "loss_alignment_w": 0.14453125, + "loss_sft": 0.5337611436843872, + "loss_total_v6": 0.7044154405593872, + "loss_wm_recon": 0.1044921875, + "loss_wm_recon_w": 0.026123046875, + "step": 1497 + }, + { + "epoch": 0.5583302273574358, + "grad_norm": 17.36597442626953, + "learning_rate": 4.642873852260065e-06, + "loss": 10.275375366210938, + "loss_alignment": 0.3046875, + "loss_alignment_w": 0.15234375, + "loss_sft": 0.2534146010875702, + "loss_total_v6": 0.4335903823375702, + "loss_wm_recon": 0.111328125, + "loss_wm_recon_w": 0.02783203125, + "step": 1498 + }, + { + "epoch": 0.5587029444651509, + "grad_norm": 18.0018367767334, + "learning_rate": 4.6365640937000015e-06, + "loss": 10.168351173400879, + "loss_alignment": 0.29296875, + "loss_alignment_w": 0.146484375, + "loss_sft": 0.5828635692596436, + "loss_total_v6": 0.7579123973846436, + "loss_wm_recon": 0.1142578125, + "loss_wm_recon_w": 0.028564453125, + "step": 1499 + }, + { + "epoch": 0.5590756615728661, + "grad_norm": 17.41851043701172, + "learning_rate": 4.630254916940424e-06, + "loss": 10.258989334106445, + "loss_alignment": 0.3203125, + "loss_alignment_w": 0.16015625, + "loss_sft": 0.4793674349784851, + "loss_total_v6": 0.6668674349784851, + "loss_wm_recon": 0.109375, + "loss_wm_recon_w": 0.02734375, + "step": 1500 + }, + { + "epoch": 0.5590756615728661, + "eval_loss": 0.6262960433959961, + "eval_loss_alignment": 0.2812249599358974, + "eval_loss_alignment_w": 0.1406124799679487, + "eval_loss_sft": 0.4607423880161383, + "eval_loss_total_v6": 0.6262960112693473, + "eval_loss_wm_recon": 0.09976457123063569, + "eval_loss_wm_recon_w": 0.024941142807658922, + "eval_runtime": 229.9434, + "eval_samples_per_second": 8.115, + "eval_steps_per_second": 1.018, + "step": 1500 + } + ], + "logging_steps": 1, + "max_steps": 2683, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.5691777043022742e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +} diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/training_args.bin b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..036aa1ac5e8970ff5eb92dd991979f22baa04d6c --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45ef6c3e0a0e8ae04df2b9927b2031622467c90c0b8fee0f06567c4b606bbd89 +size 8913 diff --git a/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/zero_to_fp32.py b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/zero_to_fp32.py new file mode 100644 index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694 --- /dev/null +++ b/qwen-cua__runs__v6-recurtorch-recon-img3-hw2-k16-12k__v0-20260605-222754__checkpoint-1500/zero_to_fp32.py @@ -0,0 +1,760 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets +# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in +# the future. Once extracted, the weights don't require DeepSpeed and can be used in any +# application. +# +# example: +# python zero_to_fp32.py . output_dir/ +# or +# python zero_to_fp32.py . output_dir/ --safe_serialization + +import argparse +import torch +import glob +import math +import os +import re +import gc +import json +import numpy as np +from tqdm import tqdm +from collections import OrderedDict +from dataclasses import dataclass + +# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with +# DeepSpeed data structures it has to be available in the current python environment. +from deepspeed.utils import logger +from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS, + FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES, + FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS) + + +@dataclass +class zero_model_state: + buffers: dict() + param_shapes: dict() + shared_params: list + ds_version: int + frozen_param_shapes: dict() + frozen_param_fragments: dict() + + +debug = 0 + +# load to cpu +device = torch.device('cpu') + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + ''' + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + ''' + return [atoi(c) for c in re.split(r'(\d+)', text)] + + +def get_model_state_file(checkpoint_dir, zero_stage): + if not os.path.isdir(checkpoint_dir): + raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist") + + # there should be only one file + if zero_stage <= 2: + file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt") + elif zero_stage == 3: + file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt") + + if not os.path.exists(file): + raise FileNotFoundError(f"can't find model states file at '{file}'") + + return file + + +def get_checkpoint_files(checkpoint_dir, glob_pattern): + # XXX: need to test that this simple glob rule works for multi-node setup too + ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys) + + if len(ckpt_files) == 0: + raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'") + + return ckpt_files + + +def get_optim_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def get_model_state_files(checkpoint_dir): + return get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def parse_model_states(files): + zero_model_states = [] + for file in files: + state_dict = torch.load(file, map_location=device, weights_only=False) + + if BUFFER_NAMES not in state_dict: + raise ValueError(f"{file} is not a model state checkpoint") + buffer_names = state_dict[BUFFER_NAMES] + if debug: + print("Found buffers:", buffer_names) + + # recover just the buffers while restoring them to fp32 if they were saved in fp16 + buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names} + param_shapes = state_dict[PARAM_SHAPES] + + # collect parameters that are included in param_shapes + param_names = [] + for s in param_shapes: + for name in s.keys(): + param_names.append(name) + + # update with frozen parameters + frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None) + if frozen_param_shapes is not None: + if debug: + print(f"Found frozen_param_shapes: {frozen_param_shapes}") + param_names += list(frozen_param_shapes.keys()) + + # handle shared params + shared_params = [[k, v] for k, v in state_dict["shared_params"].items()] + + ds_version = state_dict.get(DS_VERSION, None) + + frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None) + + z_model_state = zero_model_state(buffers=buffers, + param_shapes=param_shapes, + shared_params=shared_params, + ds_version=ds_version, + frozen_param_shapes=frozen_param_shapes, + frozen_param_fragments=frozen_param_fragments) + zero_model_states.append(z_model_state) + + return zero_model_states + + +def parse_optim_states(files, ds_checkpoint_dir): + total_files = len(files) + state_dicts = [] + for f in tqdm(files, desc='Loading checkpoint shards'): + state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False) + # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights + # and also handle the case where it was already removed by another helper script + state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None) + state_dicts.append(state_dict) + + if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]: + raise ValueError(f"{files[0]} is not a zero checkpoint") + zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE] + world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT] + + # For ZeRO-2 each param group can have different partition_count as data parallelism for expert + # parameters can be different from data parallelism for non-expert parameters. So we can just + # use the max of the partition_count to get the dp world_size. + + if type(world_size) is list: + world_size = max(world_size) + + if world_size != total_files: + raise ValueError( + f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. " + "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes." + ) + + # the groups are named differently in each stage + if zero_stage <= 2: + fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS + elif zero_stage == 3: + fp32_groups_key = FP32_FLAT_GROUPS + else: + raise ValueError(f"unknown zero stage {zero_stage}") + + fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))] + return zero_stage, world_size, fp32_flat_groups + + +def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters): + """ + Returns fp32 state_dict reconstructed from ds checkpoint + + Args: + - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are) + + """ + print(f"Processing zero checkpoint '{ds_checkpoint_dir}'") + + optim_files = get_optim_files(ds_checkpoint_dir) + zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir) + print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}") + + model_files = get_model_state_files(ds_checkpoint_dir) + + zero_model_states = parse_model_states(model_files) + print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}') + + if zero_stage <= 2: + return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + elif zero_stage == 3: + return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters) + + +def _zero2_merge_frozen_params(state_dict, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + frozen_param_fragments = zero_model_states[0].frozen_param_fragments + + if debug: + num_elem = sum(s.numel() for s in frozen_param_shapes.values()) + print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in frozen_param_fragments.values()]) + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + state_dict[name] = frozen_param_fragments[name] + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +def _has_callable(obj, fn): + attr = getattr(obj, fn, None) + return callable(attr) + + +def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + + # Reconstruction protocol: + # + # XXX: document this + + if debug: + for i in range(world_size): + for j in range(len(fp32_flat_groups[0])): + print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}") + + # XXX: memory usage doubles here (zero2) + num_param_groups = len(fp32_flat_groups[0]) + merged_single_partition_of_fp32_groups = [] + for i in range(num_param_groups): + merged_partitions = [sd[i] for sd in fp32_flat_groups] + full_single_fp32_vector = torch.cat(merged_partitions, 0) + merged_single_partition_of_fp32_groups.append(full_single_fp32_vector) + avail_numel = sum( + [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups]) + + if debug: + wanted_params = sum([len(shapes) for shapes in param_shapes]) + wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes]) + # not asserting if there is a mismatch due to possible padding + print(f"Have {avail_numel} numels to process.") + print(f"Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + total_numel = 0 + total_params = 0 + for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups): + offset = 0 + avail_numel = full_single_fp32_vector.numel() + for name, shape in shapes.items(): + + unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape) + total_numel += unpartitioned_numel + total_params += 1 + + if debug: + print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ") + state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape) + offset += unpartitioned_numel + + # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and + # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex + # paddings performed in the code it's almost impossible to predict the exact numbers w/o the + # live optimizer object, so we are checking that the numbers are within the right range + align_to = 2 * world_size + + def zero2_align(x): + return align_to * math.ceil(x / align_to) + + if debug: + print(f"original offset={offset}, avail_numel={avail_numel}") + + offset = zero2_align(offset) + avail_numel = zero2_align(avail_numel) + + if debug: + print(f"aligned offset={offset}, avail_numel={avail_numel}") + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero2_merge_frozen_params(state_dict, zero_model_states) + + _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def zero3_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states): + if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0: + return + + if debug: + for i in range(world_size): + num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values()) + print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}') + + frozen_param_shapes = zero_model_states[0].frozen_param_shapes + wanted_params = len(frozen_param_shapes) + wanted_numel = sum(s.numel() for s in frozen_param_shapes.values()) + avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size + print(f'Frozen params: Have {avail_numel} numels to process.') + print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params') + + total_params = 0 + total_numel = 0 + for name, shape in zero_model_states[0].frozen_param_shapes.items(): + total_params += 1 + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + + param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states) + state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape) + + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements") + + +class GatheredTensor: + """ + A pseudo tensor that collects partitioned weights. + It is more memory efficient when there are multiple groups. + """ + + def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape): + self.flat_groups = flat_groups + self.flat_groups_offset = flat_groups_offset + self.offset = offset + self.partitioned_numel = partitioned_numel + self.shape = shape + self.dtype = self.flat_groups[0][0].dtype + + def contiguous(self): + """ + Merge partitioned weights from flat_groups into a single tensor. + """ + end_idx = self.offset + self.partitioned_numel + world_size = len(self.flat_groups) + pad_flat_param_chunks = [] + + for rank_i in range(world_size): + # for each rank, we need to collect weights from related group/groups + flat_groups_at_rank_i = self.flat_groups[rank_i] + start_group_id = None + end_group_id = None + for group_id in range(len(self.flat_groups_offset)): + if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]: + start_group_id = group_id + if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]: + end_group_id = group_id + break + # collect weights from related group/groups + for group_id in range(start_group_id, end_group_id + 1): + flat_tensor = flat_groups_at_rank_i[group_id] + start_offset = self.offset - self.flat_groups_offset[group_id] + end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id] + pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset]) + + # collect weights from all ranks + pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0) + param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous() + return param + + +def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states): + param_shapes = zero_model_states[0].param_shapes + avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size + + # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each + # param, re-consolidating each param, while dealing with padding if any + + # merge list of dicts, preserving order + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + + if debug: + for i in range(world_size): + print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}") + + wanted_params = len(param_shapes) + wanted_numel = sum(shape.numel() for shape in param_shapes.values()) + # not asserting if there is a mismatch due to possible padding + avail_numel = fp32_flat_groups[0].numel() * world_size + print(f"Trainable params: Have {avail_numel} numels to process.") + print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.") + + # params + # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support + # out-of-core computing solution + offset = 0 + total_numel = 0 + total_params = 0 + flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]])) + for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'): + unpartitioned_numel = shape.numel() + total_numel += unpartitioned_numel + total_params += 1 + partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size) + + if debug: + print( + f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}" + ) + + # memory efficient tensor + tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape) + state_dict[name] = tensor + offset += partitioned_numel + + offset *= world_size + + # Sanity check + if offset != avail_numel: + raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong") + + print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements") + + +def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states, + exclude_frozen_parameters): + state_dict = OrderedDict() + + # buffers + buffers = zero_model_states[0].buffers + state_dict.update(buffers) + if debug: + print(f"added {len(buffers)} buffers") + + if not exclude_frozen_parameters: + _zero3_merge_frozen_params(state_dict, world_size, zero_model_states) + + _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states) + + # recover shared parameters + for pair in zero_model_states[0].shared_params: + if pair[1] in state_dict: + state_dict[pair[0]] = state_dict[pair[1]] + + return state_dict + + +def to_torch_tensor(state_dict, return_empty_tensor=False): + """ + Convert state_dict of GatheredTensor to torch tensor + """ + torch_state_dict = {} + converted_tensors = {} + for name, tensor in state_dict.items(): + tensor_id = id(tensor) + if tensor_id in converted_tensors: # shared tensors + shared_tensor = torch_state_dict[converted_tensors[tensor_id]] + torch_state_dict[name] = shared_tensor + else: + converted_tensors[tensor_id] = name + if return_empty_tensor: + torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype) + else: + torch_state_dict[name] = tensor.contiguous() + return torch_state_dict + + +def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag=None, + exclude_frozen_parameters=False, + lazy_mode=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with + ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example + via a model hub. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient. + Convert the pesduo tensor to torch tensor by ``.contiguous()`` + + Returns: + - pytorch ``state_dict`` + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + # do the training and checkpoint saving + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu + model = model.cpu() # move to cpu + model.load_state_dict(state_dict) + # submit to model hub or save the model to share with others + + In this example the ``model`` will no longer be usable in the deepspeed context of the same + application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead. + + Note: the above usage may not work if your application doesn't have sufficient free CPU memory. + You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with + the checkpoint. Or you can load state_dict in lazy mode :: + + from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu + for name, lazy_tensor in state_dict.item(): + tensor = lazy_tensor.contiguous() # to cpu + print(name, tensor) + # del tensor to release memory if it no longer in use + """ + if tag is None: + latest_path = os.path.join(checkpoint_dir, 'latest') + if os.path.isfile(latest_path): + with open(latest_path, 'r') as fd: + tag = fd.read().strip() + else: + raise ValueError(f"Unable to find 'latest' file at {latest_path}") + + ds_checkpoint_dir = os.path.join(checkpoint_dir, tag) + + if not os.path.isdir(ds_checkpoint_dir): + raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist") + + state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters) + if lazy_mode: + return state_dict + else: + return to_torch_tensor(state_dict) + + +def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, + output_dir, + max_shard_size="5GB", + safe_serialization=False, + tag=None, + exclude_frozen_parameters=False): + """ + Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be + loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed. + + Args: + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``output_dir``: directory to the pytorch fp32 state_dict output files + - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB + - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`). + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + - ``exclude_frozen_parameters``: exclude frozen parameters + """ + + # Dependency pre-check + if safe_serialization: + try: + from safetensors.torch import save_file + except ImportError: + print('If you want to use `safe_serialization`, please `pip install safetensors`') + raise + if max_shard_size is not None: + try: + from huggingface_hub import split_torch_state_dict_into_shards + except ImportError: + print('If you want to use `max_shard_size`, please `pip install huggingface_hub`') + raise + + # Convert zero checkpoint to state_dict + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, + tag, + exclude_frozen_parameters, + lazy_mode=True) + + # Shard the model if it is too big. + weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin" + if max_shard_size is not None: + filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors") + # an memory-efficient approach for sharding + empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True) + state_dict_split = split_torch_state_dict_into_shards(empty_state_dict, + filename_pattern=filename_pattern, + max_shard_size=max_shard_size) + else: + from collections import namedtuple + StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"]) + state_dict_split = StateDictSplit(is_sharded=False, + filename_to_tensors={weights_name: list(state_dict.keys())}) + + # Save the model by shard + os.makedirs(output_dir, exist_ok=True) + filename_to_tensors = state_dict_split.filename_to_tensors.items() + for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"): + shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors} + shard_state_dict = to_torch_tensor(shard_state_dict) + output_path = os.path.join(output_dir, shard_file) + if safe_serialization: + save_file(shard_state_dict, output_path, metadata={"format": "pt"}) + else: + torch.save(shard_state_dict, output_path) + # release the memory of current shard + for tensor_name in list(shard_state_dict.keys()): + del state_dict[tensor_name] + del shard_state_dict[tensor_name] + del shard_state_dict + gc.collect() + + # Save index if sharded + if state_dict_split.is_sharded: + index = { + "metadata": state_dict_split.metadata, + "weight_map": state_dict_split.tensor_to_filename, + } + save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json" + save_index_file = os.path.join(output_dir, save_index_file) + with open(save_index_file, "w", encoding="utf-8") as f: + content = json.dumps(index, indent=2, sort_keys=True) + "\n" + f.write(content) + + +def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None): + """ + 1. Put the provided model to cpu + 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` + 3. Load it into the provided model + + Args: + - ``model``: the model object to update + - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``) + - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14`` + + Returns: + - ``model`: modified model + + Make sure you have plenty of CPU memory available before you call this function. If you don't + have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it + conveniently placed for you in the checkpoint folder. + + A typical usage might be :: + + from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint + model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir) + # submit to model hub or save the model to share with others + + Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context + of the same application. i.e. you will need to re-initialize the deepspeed engine, since + ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it. + + """ + logger.info("Extracting fp32 weights") + state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag) + + logger.info("Overwriting model with fp32 weights") + model = model.cpu() + model.load_state_dict(state_dict, strict=False) + + return model + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument("checkpoint_dir", + type=str, + help="path to the desired checkpoint folder, e.g., path/checkpoint-12") + parser.add_argument("output_dir", + type=str, + help="directory to the pytorch fp32 state_dict output files" + "(e.g. path/checkpoint-12-output/)") + parser.add_argument( + "--max_shard_size", + type=str, + default="5GB", + help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size" + "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`" + "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances" + "without CPU OOM issues.") + parser.add_argument( + "--safe_serialization", + default=False, + action='store_true', + help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).") + parser.add_argument("-t", + "--tag", + type=str, + default=None, + help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1") + parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters") + parser.add_argument("-d", "--debug", action='store_true', help="enable debug") + args = parser.parse_args() + + debug = args.debug + + convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, + args.output_dir, + max_shard_size=args.max_shard_size, + safe_serialization=args.safe_serialization, + tag=args.tag, + exclude_frozen_parameters=args.exclude_frozen_parameters) diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523 +size 19989325 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1d134cd298be1e3be25db393d93a1cefe80e3214 --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,33 @@ +{ + "add_prefix_space": false, + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "backend": "tokenizers", + "bos_token": null, + "clean_up_tokenization_spaces": false, + "eos_token": "<|im_end|>", + "errors": "replace", + "image_token": "<|image_pad|>", + "is_local": true, + "local_files_only": false, + "model_max_length": 262144, + "model_specific_special_tokens": { + "audio_bos_token": "<|audio_start|>", + "audio_eos_token": "<|audio_end|>", + "audio_token": "<|audio_pad|>", + "image_token": "<|image_pad|>", + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" + }, + "pad_token": "<|endoftext|>", + "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+", + "processor_class": "Qwen3VLProcessor", + "split_special_tokens": false, + "tokenizer_class": "Qwen2Tokenizer", + "unk_token": null, + "video_token": "<|video_pad|>", + "vision_bos_token": "<|vision_start|>", + "vision_eos_token": "<|vision_end|>" +}