{
  "action_act": "linear",
  "attn_implementation": null,
  "chunk_size": 16,
  "device": "cpu",
  "dtype": "bfloat16",
  "flow_loss_weight": 1.0,
  "force_fp32_autocast": true,
  "gradient_checkpointing": false,
  "image_max_pixels": 100352,
  "image_min_pixels": 50176,
  "input_features": {},
  "license": "mit",
  "max_action_dim": 32,
  "max_period": 4.0,
  "max_state_dim": 32,
  "min_period": 0.004,
  "n_action_steps": 16,
  "n_obs_steps": 1,
  "normalization_mapping": {
    "ACTION": "MEAN_STD",
    "STATE": "MEAN_STD",
    "VISUAL": "IDENTITY"
  },
  "num_action_layers": 2,
  "num_denoise_steps": 10,
  "optimizer_betas": [
    0.9,
    0.999
  ],
  "optimizer_eps": 1e-08,
  "optimizer_grad_clip_norm": 1.0,
  "optimizer_lr": 0.0001,
  "optimizer_weight_decay": 0.1,
  "output_features": {},
  "pretrained_path": "lerobot/eo1-base",
  "pretrained_revision": null,
  "private": null,
  "push_to_hub": true,
  "recipe_path": null,
  "repo_id": "lerobot/eo1-base",
  "scheduler_decay_lr": 0.0,
  "scheduler_decay_steps": 30000,
  "scheduler_warmup_steps": 900,
  "supervise_padding_action_dims": true,
  "supervise_padding_actions": true,
  "tags": [
    "eo1",
    "robotics",
    "vision-language-action",
    "lerobot"
  ],
  "text_loss_weight": 0.01,
  "time_sampling_beta_alpha": 1.5,
  "time_sampling_beta_beta": 1.0,
  "time_sampling_offset": 0.001,
  "time_sampling_scale": 0.999,
  "tokenizer_max_length": 1000,
  "type": "eo1",
  "use_amp": false,
  "use_fast_processor": false,
  "use_peft": false,
  "vlm_base": "lerobot/eo1-base",
  "vlm_config": {
    "_name_or_path": "",
    "action_pass_id": 151672,
    "action_token_id": 151666,
    "architectures": [
      "Qwen2_5_VLForConditionalGeneration"
    ],
    "chunk_size_feed_forward": 0,
    "dtype": "bfloat16",
    "eos_token_id": 151645,
    "id2label": {
      "0": "LABEL_0",
      "1": "LABEL_1"
    },
    "image_token_id": 151655,
    "is_encoder_decoder": false,
    "label2id": {
      "LABEL_0": 0,
      "LABEL_1": 1
    },
    "model_type": "qwen2_5_vl",
    "output_attentions": false,
    "output_hidden_states": false,
    "pad_token_id": 151643,
    "problem_type": null,
    "return_dict": true,
    "state_token_id": 151669,
    "text_config": {
      "_name_or_path": "",
      "architectures": null,
      "attention_dropout": 0.0,
      "bos_token_id": 151643,
      "chunk_size_feed_forward": 0,
      "dtype": "bfloat16",
      "eos_token_id": 151645,
      "hidden_act": "silu",
      "hidden_size": 2048,
      "id2label": {
        "0": "LABEL_0",
        "1": "LABEL_1"
      },
      "initializer_range": 0.02,
      "intermediate_size": 11008,
      "is_encoder_decoder": false,
      "label2id": {
        "LABEL_0": 0,
        "LABEL_1": 1
      },
      "layer_types": [
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention",
        "full_attention"
      ],
      "max_position_embeddings": 128000,
      "max_window_layers": 70,
      "model_type": "qwen2_5_vl_text",
      "num_attention_heads": 16,
      "num_hidden_layers": 36,
      "num_key_value_heads": 2,
      "output_attentions": false,
      "output_hidden_states": false,
      "pad_token_id": null,
      "problem_type": null,
      "return_dict": true,
      "rms_norm_eps": 1e-06,
      "rope_parameters": {
        "mrope_section": [
          16,
          24,
          24
        ],
        "rope_theta": 1000000.0,
        "rope_type": "default",
        "type": "mrope"
      },
      "sliding_window": null,
      "use_cache": true,
      "use_sliding_window": false,
      "vocab_size": 151936
    },
    "tie_word_embeddings": true,
    "transformers_version": "5.5.4",
    "use_cache": false,
    "video_token_id": 151656,
    "vision_config": {
      "_name_or_path": "",
      "architectures": null,
      "chunk_size_feed_forward": 0,
      "depth": 32,
      "dtype": null,
      "fullatt_block_indexes": [
        7,
        15,
        23,
        31
      ],
      "hidden_act": "silu",
      "hidden_size": 1280,
      "id2label": {
        "0": "LABEL_0",
        "1": "LABEL_1"
      },
      "in_channels": 3,
      "in_chans": 3,
      "initializer_range": 0.02,
      "intermediate_size": 3420,
      "is_encoder_decoder": false,
      "label2id": {
        "LABEL_0": 0,
        "LABEL_1": 1
      },
      "model_type": "qwen2_5_vl",
      "num_heads": 16,
      "out_hidden_size": 2048,
      "output_attentions": false,
      "output_hidden_states": false,
      "patch_size": 14,
      "problem_type": null,
      "return_dict": true,
      "spatial_merge_size": 2,
      "spatial_patch_size": 14,
      "temporal_patch_size": 2,
      "tokens_per_second": 2,
      "window_size": 112
    },
    "vision_end_token_id": 151653,
    "vision_start_token_id": 151652,
    "vision_token_id": 151654,
    "vocab_size": 151936
  }
}