| { | |
| "model_name_or_path": "../pretrained/InternVL3-2B", | |
| "freeze_backbone": true, | |
| "unfreeze_vit_layers": 0, | |
| "use_backbone_lora": 0, | |
| "freeze_llm": true, | |
| "unfreeze_lm_head": false, | |
| "use_llm_lora": 64, | |
| "freeze_mlp": false, | |
| "vision_select_layer": -1, | |
| "grad_checkpoint": true, | |
| "gradient_checkpointing": true, | |
| "drop_path_rate": 0.0, | |
| "ps_version": "v2", | |
| "use_fast_tokenizer": true, | |
| "max_seq_length": 24576, | |
| "force_image_size": 448, | |
| "down_sample_ratio": 0.5, | |
| "pad2square": true, | |
| "conv_style": "internvl2_5_nav", | |
| "meta_path": "", | |
| "use_data_resampling": false, | |
| "dynamic_image_size": false, | |
| "use_thumbnail": true, | |
| "min_dynamic_patch": 1, | |
| "max_dynamic_patch": 12, | |
| "normalize_type": "imagenet", | |
| "use_packed_ds": true, | |
| "num_images_expected": 1000, | |
| "max_packed_tokens": 24576, | |
| "max_buffer_size": 20, | |
| "log_freq": 1000, | |
| "strict_mode": false, | |
| "replacement": true, | |
| "allow_overflow": false, | |
| "loss_reduction": "square", | |
| "loss_reduction_all_gather": false, | |
| "seed": 42, | |
| "output_dir": "", | |
| "overwrite_output_dir": true, | |
| "report_to": "wandb", | |
| "save_strategy": "steps", | |
| "save_total_limit": 1, | |
| "save_steps": 0.5, | |
| "logging_steps": 1, | |
| "evaluation_strategy": "no", | |
| "dataloader_num_workers": 2, | |
| "group_by_length": false, | |
| "deepspeed": "zero_stage2_config_acc1.json", | |
| "remove_unused_columns": false, | |
| "do_train": true, | |
| "bf16": true, | |
| "learning_rate": 0.0001, | |
| "weight_decay": 0.01, | |
| "warmup_ratio": 0.03, | |
| "lr_scheduler_type": "cosine", | |
| "max_grad_norm": 1.0, | |
| "num_train_epochs": 1, | |
| "per_device_train_batch_size": 1, | |
| "gradient_accumulation_steps": 1, | |
| "max_steps": 8000, | |
| "template_name": "BEVftFOV_FrontierRGB_PosB__FRONTIER_PIXEL_NUMBER_ONLY", | |
| "use_nav_system_message": true, | |
| "use_position_embeddings": true, | |
| "use_pairwise_spatial_encoder": false, | |
| "use_dual_vit": true, | |
| "bev_image_size": 448, | |
| "num_image_token_bev": 256, | |
| "num_image_token_ego": 32, | |
| "vit_bev_freeze": true, | |
| "vit_bev_use_lora": true, | |
| "vit_bev_lora_rank": 64, | |
| "vit_rgb_freeze": true, | |
| "vit_rgb_use_lora": true, | |
| "vit_rgb_lora_rank": 16, | |
| "position_placeholders": [ | |
| "<s>", | |
| "<cand>", | |
| "<e_s>", | |
| "<e_cand>" | |
| ], | |
| "dual_text_pos_injection": true | |
| } |