{
    "robovlm_name": "RoboQwen25VL",
    "parent": null,
    "task_name": "bridge_finetune",
    "model": "qwen25vl",
    "model_url": "https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct",
    "seq_len": 1,
    "image_size": 224,
    "image_mean": [
        0.48145466,
        0.4578275,
        0.40821073
    ],
    "image_std": [
        0.26862954,
        0.26130258,
        0.27577711
    ],
    "window_size": 1,
    "fwd_pred_next_n": 4,
    "arm_gripper_loss_ratio": 0.01,
    "cap_loss_ratio": 0.05,
    "fwd_loss_ratio": 0,
    "seed": 123,
    "batch_size": 8,
    "num_workers": 4,
    "data_scale": 1,
    "optimizer": "adam",
    "learning_rate": 5e-05,
    "min_lr_scale": 0.01,
    "weight_decay": 0,
    "warmup_epochs": 0.25,
    "warmup_steps": 0,
    "warmup_ratio": null,
    "use_hand_rgb": false,
    "use_time_causal_attn": false,
    "use_mim_obs_loss": false,
    "use_pixel_loss": true,
    "use_obs_queries": true,
    "use_vision_resampler": false,
    "vision_masked_ratio": 0.9,
    "use_tube_mask": false,
    "cache_root": "/dev/shm/vlm4vla/cache/qwen25vl",
    "model_load_path": null,
    "model_load_source": "torch",
    "resume": null,
    "model_path": "/workspace/models/Qwen2.5-VL-3B-Instruct",
    "model_config": "/workspace/models/Qwen2.5-VL-3B-Instruct/config.json",
    "train_setup": {
        "precision": "bf16",
        "predict_action": true,
        "predict_forward": false,
        "predict_forward_hand": false,
        "predict_caption": false,
        "train_vision": false,
        "bits": -1,
        "freeze_mm_mlp_adapter": false,
        "freeze_backbone": false,
        "freeze_resampler": false,
        "tune_mm_mlp_adapter": false,
        "mm_use_im_start_end": false,
        "mm_use_im_patch_token": false,
        "gradient_checkpointing": false,
        "lora_enable": false,
        "mm_projector_lr": 0.0001,
        "lora_r": 64,
        "lora_alpha": 16,
        "lora_dropout": 0.05,
        "lora_bias": "none",
        "train_text_embedding": true
    },
    "vision_resampler": {
        "vis_dim": 1024,
        "depth": 8,
        "dim_head": 64,
        "heads": 8,
        "num_latents": 64
    },
    "act_encoder": null,
    "act_head": {
        "type": "FCDecoder",
        "hidden_size": 1024,
        "action_dim": 7,
        "down_sample": "none",
        "latent": 1,
        "fwd_pred_next_n": 1,
        "window_size": 1,
        "action_space": "continuous",
        "with_history": true,
        "history_type": "post"
    },
    "fwd_head": null,
    "tokenizer": {
        "type": "AutoProcessor",
        "pretrained_model_name_or_path": "/workspace/models/Qwen2.5-VL-3B-Instruct",
        "tokenizer_type": "qwen25vl",
        "additional_special_tokens": null
    },
    "vlm": {
        "type": "Qwen2_5_VLForConditionalGeneration",
        "pretrained_model_name_or_path": "/workspace/models/Qwen2.5-VL-3B-Instruct",
        "name": "qwen25vl"
    },
    "trainer": {
        "accelerator": "gpu",
        "strategy": "deepspeed_stage_2",
        "precision": "bf16",
        "logger": [
            "wandb"
        ],
        "gradient_clip_val": 1.0,
        "use_distributed_sampler": false,
        "log_every_n_steps": 10,
        "max_epochs": 5,
        "val_check_interval": 40000,
        "check_val_every_n_epoch": null,
        "max_steps": 50000,
        "accumulate_grad_batches": 8
    },
    "train_dataset": {
        "type": "OpenVLADataset",
        "data_root_dir": "/workspace/data",
        "model_name": "qwen25vl",
        "image_aug": true,
        "mode": "train",
        "data_mix": "bridge",
        "window_sample": "sliding",
        "organize_type": "interleave",
        "shuffle_buffer_size": 51200,
        "train": true
    },
    "val_dataset": {
        "type": "OpenVLADataset",
        "data_root_dir": "/workspace/data",
        "model_name": "qwen25vl",
        "mode": "train",
        "data_mix": "bridge",
        "window_sample": "sliding",
        "organize_type": "interleave",
        "shuffle_buffer_size": 10000,
        "train": false
    },
    "norm_action": true,
    "norm_min": -0.65,
    "norm_max": 0.65,
    "raw_config_path": "/workspace/VLM4VLA/configs/oxe_training/bridge/finetune_qwen25vl-3b_bridge_LOCAL_freezevis.json",
    "num_nodes": 1,
    "config": "/workspace/VLM4VLA/configs/oxe_training/bridge/finetune_qwen25vl-3b_bridge_LOCAL_freezevis.json",
    "gpus": 8,
    "log_dir": "/dev/shm/vlm4vla/logs/qwen25vl/bridge_finetune/2026-04-18/bridge_-bs512-lr5e-05-ws1-FCDecoder-latent1-freeze_vision",
    "output_dir": "/dev/shm/vlm4vla/ckpts/qwen25vl/bridge_finetune/2026-04-18/bridge_-bs512-lr5e-05-ws1-FCDecoder-latent1-freeze_vision",
    "data_dir": null,
    "annotation_file": null,
    "data_subfolder": null,
    "task_num": null,
    "exp_name": "21-04",
    "use_multi_modal_emb": false,
    "no_video_pretrained_model": false,
    "finetune": false,
    "llm": {
        "type": null,
        "n_embd": null,
        "n_layer": null,
        "n_head": null
    }
}