Package Modujo-1B-A0.75B cleanly and remove superseded root weights

Move the pretrained model into a checkpoint-free standalone folder, omit training-only args, and remove the old root model shards.

model-00001-of-00005.safetensors

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:98405b200b930250ef662a9ed5e4d8115676b45ea69cde0ea313df4b40d9fe07
-size 4000068520

model-00002-of-00005.safetensors

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:794b8574cecc9107332e4e3a607de952559a36e1f10ecdaf184ad42b2b6abfc8
-size 3998606504

model-00003-of-00005.safetensors

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:84d53b361527e08b95c79e984300bd088e750517a82700732334d8c3082c83c0
-size 4000119024

model-00004-of-00005.safetensors

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2ee6db67e997b64aa2f422274a6723e621039289e68c86717479e8d288054ea9
-size 3991529952

model-00005-of-00005.safetensors

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ccfa2a7060a360c5dcf848acd6301c43da0ef826d69a21b22290488ba4db563a
-size 2010707064

pretrain/Modujo-1B-A0.75B-checkpoint-78900/args.json

@@ -1,359 +0,0 @@
-{
-  "output_dir": "/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/output/modujo-1b-domain-continue-bs128/v3-20260922-135620",
-  "per_device_train_batch_size": 4,
-  "num_train_epochs": 1.0,
-  "max_steps": -1,
-  "learning_rate": 0.0003,
-  "lr_scheduler_type": "cosine_with_min_lr",
-  "lr_scheduler_kwargs": "{\"min_lr_rate\": 0.1}",
-  "warmup_steps": 0,
-  "optim": "adamw_torch_fused",
-  "optim_args": null,
-  "weight_decay": 0.1,
-  "adam_beta1": 0.9,
-  "adam_beta2": 0.95,
-  "adam_epsilon": 1e-08,
-  "optim_target_modules": null,
-  "gradient_accumulation_steps": 8,
-  "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": "/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/output/modujo-1b-domain-continue-bs128/v3-20260922-135620",
-  "project": "huggingface",
-  "trackio_space_id": null,
-  "trackio_bucket_id": null,
-  "trackio_static_space_id": null,
-  "eval_strategy": "no",
-  "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": 2,
-  "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,
-  "dataloader_multiprocessing_context": null,
-  "dataloader_in_order": true,
-  "remove_unused_columns": true,
-  "label_names": null,
-  "train_sampling_strategy": "random",
-  "length_column_name": "length",
-  "ddp_find_unused_parameters": true,
-  "ddp_bucket_cap_mb": null,
-  "ddp_broadcast_buffers": null,
-  "ddp_static_graph": null,
-  "ddp_backend": null,
-  "ddp_timeout": 18000000,
-  "fsdp": [],
-  "fsdp_config": null,
-  "deepspeed": null,
-  "debug": null,
-  "skip_memory_metrics": true,
-  "do_train": false,
-  "do_eval": false,
-  "do_predict": false,
-  "resume_from_checkpoint": "/tmp/modujo-resume-9500-grouped-v2",
-  "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": true,
-  "router_aux_loss_coef": 0.001,
-  "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": "modujo_muon",
-  "loss_type": null,
-  "mrl_dims": 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,
-  "logging_dir": "/tmp/modujo-domain-bs128-logging",
-  "warmup_ratio": 0.02,
-  "use_ray": false,
-  "ray_exp_name": null,
-  "device_groups": null,
-  "model": "/tmp/modujo-resume-9500-grouped-v2",
-  "model_type": "modujo_qwen4_exp",
-  "model_revision": null,
-  "task_type": "causal_lm",
-  "torch_dtype": "bfloat16",
-  "attn_impl": null,
-  "experts_impl": "grouped_mm",
-  "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",
-  "system": null,
-  "max_length": 2048,
-  "truncation_strategy": "delete",
-  "max_pixels": null,
-  "agent_template": null,
-  "norm_bbox": null,
-  "use_chat_template": false,
-  "padding_side": "right",
-  "padding_free": false,
-  "loss_scale": "all+ignore_empty_think",
-  "sequence_parallel_size": 1,
-  "is_binary_loss_scale": null,
-  "template_backend": "swift",
-  "response_prefix": null,
-  "enable_thinking": null,
-  "preserve_thinking": null,
-  "add_non_thinking_prefix": true,
-  "disable_ignore_empty_think": false,
-  "dataset": [
-    "/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/datasets/pretrain_standard/minimind_pretrain_t2t.jsonl",
-    "/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/datasets/pretrain_standard/amazon_esci_products.jsonl",
-    "/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/datasets/pretrain_standard/wayfair_wands_products.jsonl"
-  ],
-  "val_dataset": [],
-  "cached_dataset": [],
-  "cached_val_dataset": [],
-  "split_dataset_ratio": 0.0,
-  "dataset_num_proc": 8,
-  "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": [
-    "/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/common/modujo_swift_plugin.py"
-  ],
-  "custom_register_path": [],
-  "model_kwargs": {
-    "output_router_logits": true
-  },
-  "enable_npu_model_patch": true,
-  "load_args": false,
-  "load_data_args": false,
-  "packing": false,
-  "packing_length": null,
-  "packing_num_proc": 1,
-  "packing_strategy": "binpack",
-  "lazy_tokenize": false,
-  "use_hf": true,
-  "ignore_args_error": false,
-  "use_swift_lora": false,
-  "freeze_parameters": [],
-  "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": 32.0,
-  "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": "4.5.3",
-  "ckpt_dir": "/tmp/modujo-resume-9500-grouped-v2",
-  "rank": 0,
-  "global_world_size": 4,
-  "local_world_size": 4,
-  "model_suffix": "modujo-resume-9500-grouped-v2",
-  "model_info": "ModelInfo(model_type='modujo_qwen4_exp', model_dir='/tmp/modujo-resume-9500-grouped-v2', torch_dtype=torch.bfloat16, max_model_len=32768, quant_method=None, quant_bits=None, rope_scaling={'partial_rotary_factor': 0.25, 'rope_theta': 10000000, 'rope_type': 'default'}, is_moe_model=True, is_multimodal=False, config=None, task_type='causal_lm', num_labels=None)",
-  "model_meta": "ModelMeta(model_type='modujo_qwen4_exp', model_groups=[ModelGroup(models=[Model(ms_model_id=None, hf_model_id='Alexhu1999/Modujo-9B-A1B', model_path=None, ms_revision=None, hf_revision=None)], template=None, ignore_patterns=None, requires=None, tags=[])], loader=<class 'swift.model.register.ModelLoader'>, template='qwen3', model_arch=ModelKeys(arch_name='llama', embedding='model.embed_tokens', module_list='model.layers', lm_head='lm_head', q_proj='model.layers.{}.self_attn.q_proj', k_proj='model.layers.{}.self_attn.k_proj', v_proj='model.layers.{}.self_attn.v_proj', o_proj='model.layers.{}.self_attn.o_proj', attention='model.layers.{}.self_attn', mlp='model.layers.{}.mlp', down_proj='model.layers.{}.mlp.down_proj', qkv_proj=None, qk_proj=None, qa_proj=None, qb_proj=None, kv_proj=None, kva_proj=None, kvb_proj=None), mcore_model_type=None, architectures=['Qwen4ExpForCausalLM'], additional_saved_files=[], torch_dtype=None, is_multimodal=False, is_reward=False, task_type=None, ignore_patterns=None, requires=['transformers>=5.16.0'], tags=['text-generation', 'moe'])",
-  "model_dir": "/tmp/modujo-resume-9500-grouped-v2",
-  "template_meta": "Qwen3MixedTemplateMeta(template_type='qwen3', 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=<class 'swift.template.base.Template'>, system_prefix=['<|im_start|>system\\n{{SYSTEM}}<|im_end|>\\n'], default_system=None, auto_add_bos=False, stop_words=['<|endoftext|>'], agent_template='hermes', is_thinking=True, thinking_prefix='', non_thinking_prefix='<think>\\n\\n</think>\\n\\n', preserve_thinking=None, history_thinking_prefix='')",
-  "_val_dataset_exists": false,
-  "hub": "<class 'swift.hub.hub.HFHub'>",
-  "evaluation_strategy": "steps",
-  "training_args": "Seq2SeqTrainingArguments(output_dir='/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/output/modujo-1b-domain-continue-bs128/v3-20260922-135620', per_device_train_batch_size=4, num_train_epochs=1.0, max_steps=-1, learning_rate=0.0003, lr_scheduler_type=<SchedulerType.COSINE_WITH_MIN_LR: 'cosine_with_min_lr'>, lr_scheduler_kwargs={'min_lr_rate': 0.1}, warmup_steps=0.02, optim=<OptimizerNames.ADAMW_TORCH_FUSED: 'adamw_torch_fused'>, optim_args=None, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, optim_target_modules=None, gradient_accumulation_steps=8, 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=<IntervalStrategy.STEPS: 'steps'>, 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='/mnt/bn/ecom-ai-platform-1/yunhai/ModujoMoe/output/modujo-1b-domain-continue-bs128/v3-20260922-135620', project='huggingface', trackio_space_id=None, trackio_bucket_id=None, trackio_static_space_id=None, eval_strategy=<IntervalStrategy.NO: 'no'>, 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=None, include_for_metrics=[], batch_eval_metrics=False, save_only_model=False, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=100, save_on_each_node=False, save_total_limit=2, enable_jit_checkpoint=False, push_to_hub=False, hub_token=None, hub_private_repo=None, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, 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, dataloader_multiprocessing_context=None, dataloader_in_order=True, remove_unused_columns=False, label_names=None, train_sampling_strategy='random', length_column_name='length', ddp_find_unused_parameters=True, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, ddp_static_graph=None, ddp_backend=None, ddp_timeout=18000000, fsdp=[], fsdp_config=None, deepspeed=None, debug=[], skip_memory_metrics=True, do_train=False, do_eval=False, do_predict=False, resume_from_checkpoint='/tmp/modujo-resume-9500-grouped-v2', 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=True, router_aux_loss_coef=0.001, 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='modujo_muon', loss_type=None, mrl_dims=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, logging_dir='/tmp/modujo-domain-bs128-logging', warmup_ratio=0.0)"
-}

pretrain/{Modujo-1B-A0.75B-checkpoint-78900 → Modujo-1B-A0.75B}/parameter_summary.json

@@ -1,6 +1,5 @@
 {
   "model_label": "Modujo-1B-A0.75B",
-  "source_checkpoint": "checkpoint-78900",
   "total_parameters": 1003821539,
   "active_parameters_per_token": 749017571,
   "total_parameters_billions": 1.003821539,