| { |
| "loss_args": { |
| "losses": "img_distill,retrieval", |
| "retrieval_loss_weight": 0.5, |
| "retrieval_queue_size": 8192, |
| "retrieval_loss_type": "sigmoid", |
| "retrieval_scale": 100.0, |
| "retrieval_bias": -15.0, |
| "retrieval_causal_positive_base": 0.2, |
| "distill_loss_weight": 1.0, |
| "temporal_memory_loss_weight": 1.0, |
| "temporal_memory_temperature": 0.05, |
| "temporal_memory_gamma": 0.8, |
| "order_loss_weight": 0.5, |
| "order_temperature": 0.07, |
| "order_num_negatives": 1, |
| "order_num_frames_per_video": 8, |
| "order_ema_decay": 0.0, |
| "order_save_prob": 0.0, |
| "order_save_dir": "order_debug_samples", |
| "order_save_max_samples": 200, |
| "temporal_state_loss_weight": 1.0, |
| "temporal_state_alpha": 1.0, |
| "temporal_state_stable_weight": 1.0, |
| "temporal_state_delta_weight": 1.0, |
| "temporal_state_teacher_chunk_size": 32, |
| "temporal_state_use_projection": true, |
| "pred_loss_weight": 1.0, |
| "history_decode_loss_weight": 1.0, |
| "img_distill_loss_weight": 1.0, |
| "img_distill_cos_weight": 1.0, |
| "img_distill_map_weight": 0.01, |
| "img_distill_retrieval_kl_weight": 1.0, |
| "img_distill_queue_size": 8192, |
| "img_distill_teacher_temperature": 0.05, |
| "img_distill_student_temperature": 0.1, |
| "img_distill_map_teacher_temperature": 0.07, |
| "img_distill_map_student_temperature": 0.07, |
| "img_distill_num_layers": [ |
| "-1" |
| ] |
| }, |
| "model_args": { |
| "model_name_or_path": "encoder_only", |
| "dataset_type": "image_video_webdataset", |
| "vision_tower": "gatedeltanet_siglip_2cross_history", |
| "pretrain_encoder": "/home/pubw/proj/encoder/playground/models/siglip2-base-patch16-224", |
| "mm_projector_type": "state_proj", |
| "vision_checkpoint_path": null, |
| "use_video_boundaries": true, |
| "video_cut": false, |
| "num_temporal_layers": 4, |
| "temporal_layer_indices": [ |
| -1, |
| -2, |
| -3, |
| -4 |
| ], |
| "conv_kernel_size": 4, |
| "state_expansion_factor": 1, |
| "temporal_state_init_mode": "learn", |
| "temporal_mlp_ratio": "2", |
| "image_token_index": 151655, |
| "video_token_index": 151656, |
| "vision_token_index": 151654, |
| "mm_hidden_size": null, |
| "collator_type": "video_instruction", |
| "target_frames_per_batch": 0, |
| "state_select_layer": -2, |
| "frames_per_stage": 16, |
| "max_stages": null, |
| "min_stages": 1, |
| "max_visual_stages": 16, |
| "model_mode": "encoder", |
| "vision_minimal_forward": false, |
| "frozen_spatial_chunk_size": 128, |
| "sparse_temporal_patch_ratio": 1.0 |
| }, |
| "data_args": { |
| "dataset_use": null, |
| "data_path": "./data/train.jsonl", |
| "image_data_path": "/home/pubw/proj/encoder/playground/images/cc12m_imagenet21k_recap_256_20m_wds/{00000..00045}.tar", |
| "video_data_path": "/home/pubw/proj/encoder/playground/videos/internvid-2m/shards_fps2_max32_jpgzip/data-{000001..000446}.tar", |
| "image_mix_ratio": 0.5, |
| "mixed_image_batch_size": 64, |
| "mixed_video_batch_size": 10, |
| "mixed_image_num_workers": 2, |
| "mixed_video_num_workers": 4, |
| "mixed_video_decode_threads": 1, |
| "mixed_video_oversample_factor": 1.0, |
| "mixed_video_target_frames": 180, |
| "mixed_min_frames_per_video": 8, |
| "mixed_max_image_repeats": 2, |
| "mixed_image_jitter_px": 0, |
| "mixed_image_jitter_brightness": 0.0, |
| "mixed_image_jitter_contrast": 0.0, |
| "mixed_image_jitter_noise_std": 0.0, |
| "step_mixed_image_steps": 1, |
| "step_mixed_video_steps": 4, |
| "video_root": null, |
| "video_metadata_cache_path": null, |
| "image_static_frames": 2, |
| "allowed_image_grids": null, |
| "num_frames": 4, |
| "fps": 2.0, |
| "eval_data_path": null, |
| "image_folder": "./data/images", |
| "image_min_pixels": 3136, |
| "image_max_pixels": 12845056, |
| "image_resized_width": null, |
| "image_resized_height": null, |
| "video_min_pixels": 100352, |
| "video_max_pixels": 602112, |
| "video_resized_width": null, |
| "video_resized_height": null, |
| "max_frames": 24 |
| }, |
| "training_args": { |
| "output_dir": "./output/siglip_1234_vid30_vf160_20000_sigmoid_scale100_bias015_ImgDistill4_nocut_causalretrieval", |
| "overwrite_output_dir": false, |
| "do_train": false, |
| "do_eval": false, |
| "do_predict": false, |
| "eval_strategy": "no", |
| "prediction_loss_only": false, |
| "per_device_train_batch_size": 1, |
| "per_device_eval_batch_size": 8, |
| "per_gpu_train_batch_size": null, |
| "per_gpu_eval_batch_size": null, |
| "gradient_accumulation_steps": 5, |
| "eval_accumulation_steps": null, |
| "eval_delay": 0, |
| "torch_empty_cache_steps": null, |
| "learning_rate": 2e-05, |
| "weight_decay": 0.05, |
| "adam_beta1": 0.9, |
| "adam_beta2": 0.95, |
| "adam_epsilon": 1e-08, |
| "max_grad_norm": 1.0, |
| "num_train_epochs": 3.0, |
| "max_steps": 20000, |
| "lr_scheduler_type": "cosine", |
| "lr_scheduler_kwargs": {}, |
| "warmup_ratio": 0.03, |
| "warmup_steps": 0, |
| "log_level": "passive", |
| "log_level_replica": "warning", |
| "log_on_each_node": true, |
| "logging_dir": "./output/siglip_1234_vid30_vf160_20000_sigmoid_scale100_bias015_ImgDistill4_nocut_causalretrieval/runs/Aug11_13-43-12_612d7ecd0b4e", |
| "logging_strategy": "steps", |
| "logging_first_step": false, |
| "logging_steps": 1.0, |
| "logging_nan_inf_filter": true, |
| "save_strategy": "steps", |
| "save_steps": 1000, |
| "save_total_limit": 3, |
| "save_safetensors": true, |
| "save_on_each_node": false, |
| "save_only_model": false, |
| "restore_callback_states_from_checkpoint": false, |
| "no_cuda": false, |
| "use_cpu": false, |
| "use_mps_device": false, |
| "seed": 42, |
| "data_seed": null, |
| "jit_mode_eval": false, |
| "bf16": true, |
| "fp16": false, |
| "fp16_opt_level": "O1", |
| "half_precision_backend": "auto", |
| "bf16_full_eval": false, |
| "fp16_full_eval": false, |
| "tf32": true, |
| "local_rank": 1, |
| "ddp_backend": null, |
| "tpu_num_cores": null, |
| "tpu_metrics_debug": false, |
| "debug": [], |
| "dataloader_drop_last": true, |
| "eval_steps": null, |
| "dataloader_num_workers": 2, |
| "dataloader_prefetch_factor": 2, |
| "past_index": -1, |
| "run_name": "siglip_1234_vid30_vf160_20000_sigmoid_scale100_bias015_ImgDistill4_nocut_causalretrieval", |
| "disable_tqdm": false, |
| "remove_unused_columns": false, |
| "label_names": null, |
| "load_best_model_at_end": false, |
| "metric_for_best_model": null, |
| "greater_is_better": null, |
| "ignore_data_skip": true, |
| "fsdp": [], |
| "fsdp_min_num_params": 0, |
| "fsdp_config": { |
| "min_num_params": 0, |
| "xla": false, |
| "xla_fsdp_v2": false, |
| "xla_fsdp_grad_ckpt": false |
| }, |
| "fsdp_transformer_layer_cls_to_wrap": null, |
| "accelerator_config": { |
| "split_batches": false, |
| "dispatch_batches": null, |
| "even_batches": true, |
| "use_seedable_sampler": true, |
| "non_blocking": false, |
| "gradient_accumulation_kwargs": null |
| }, |
| "parallelism_config": null, |
| "deepspeed": "scripts/zero1.json", |
| "label_smoothing_factor": 0.0, |
| "optim": "adamw_torch", |
| "optim_args": null, |
| "adafactor": false, |
| "group_by_length": false, |
| "length_column_name": "length", |
| "report_to": [ |
| "wandb" |
| ], |
| "project": "huggingface", |
| "trackio_space_id": "trackio", |
| "ddp_find_unused_parameters": false, |
| "ddp_bucket_cap_mb": null, |
| "ddp_broadcast_buffers": null, |
| "dataloader_pin_memory": true, |
| "dataloader_persistent_workers": false, |
| "skip_memory_metrics": true, |
| "use_legacy_prediction_loop": false, |
| "push_to_hub": false, |
| "resume_from_checkpoint": "False", |
| "hub_model_id": null, |
| "hub_strategy": "every_save", |
| "hub_token": "<HUB_TOKEN>", |
| "hub_private_repo": null, |
| "hub_always_push": false, |
| "hub_revision": null, |
| "gradient_checkpointing": true, |
| "gradient_checkpointing_kwargs": null, |
| "include_inputs_for_metrics": false, |
| "include_for_metrics": [], |
| "eval_do_concat_batches": true, |
| "fp16_backend": "auto", |
| "push_to_hub_model_id": null, |
| "push_to_hub_organization": null, |
| "push_to_hub_token": "<PUSH_TO_HUB_TOKEN>", |
| "mp_parameters": "", |
| "auto_find_batch_size": false, |
| "full_determinism": false, |
| "torchdynamo": null, |
| "ray_scope": "last", |
| "ddp_timeout": 1800, |
| "torch_compile": false, |
| "torch_compile_backend": null, |
| "torch_compile_mode": null, |
| "include_tokens_per_second": false, |
| "include_num_input_tokens_seen": "no", |
| "neftune_noise_alpha": null, |
| "optim_target_modules": null, |
| "batch_eval_metrics": false, |
| "eval_on_start": false, |
| "use_liger_kernel": true, |
| "liger_kernel_config": null, |
| "eval_use_gather_object": false, |
| "average_tokens_across_devices": true, |
| "cache_dir": null, |
| "double_quant": true, |
| "quant_type": "nf4", |
| "bits": 16, |
| "trainable_layers": 0, |
| "frozen_text_encoder": false, |
| "lora_enable": false, |
| "lora_rank": 64, |
| "lora_alpha": 16, |
| "lora_dropout": 0.05, |
| "lora_weight_path": "", |
| "lora_bias": "none", |
| "use_lora": false, |
| "tune_llm": false, |
| "tune_connector": false, |
| "tune_vision_temporal": true, |
| "tune_vision_other": false, |
| "tune_vision_spatial_pair_t": false, |
| "tune_pooling_head": false, |
| "max_train_samples": 20000000, |
| "llm_lr": 0.0, |
| "connector_lr": 0.0, |
| "vision_temporal_lr": 0.0005, |
| "vision_other_lr": 0.0, |
| "vision_spatial_lora_lr": 0.0005, |
| "vision_spatial_pair_t_lr": null, |
| "pooling_head_lr": 5e-06, |
| "vision_spatial_lora_enable": false, |
| "vision_spatial_lora_rank": 32, |
| "vision_spatial_lora_alpha": 32, |
| "vision_spatial_lora_dropout": 0.0, |
| "llm_sequence_packing": false, |
| "max_packed_tokens": 0, |
| "packing_buffer_size": 512, |
| "packing_prefetch_workers": 4, |
| "cross_rank_packing_balance": true, |
| "grad_norm_skip_threshold": 0.0, |
| "drop_token": "<DROP_TOKEN>", |
| "downsample": false, |
| "debug_cuda_sync_timing": false, |
| "log_rank_balance": true |
| } |
| } |
