dexvla_robotwin / config.json
cuichaowei's picture
Upload folder using huggingface_hub
defaaa3 verified
{
"Using_EMA_Pretrain_DiT": false,
"_name_or_path": "/data/public/wrt/m2/ccw/ckpt/dexvla/qwen2_vl_3_cameras_1_17_all_data_pretrain_6w_DiT_H_Non_EMA_full_param_stage_1_50/checkpoint-60000",
"architectures": [
"DexVLA"
],
"attention_dropout": 0.0,
"bos_token_id": 151643,
"camera_names": [
"observation.images.front",
"observation.images.left",
"observation.images.right"
],
"concat": "token_cat",
"eos_token_id": 151645,
"external_vision_encoder": "None",
"freeze_policy_head": false,
"freeze_vision_tower": false,
"from_scratch": false,
"ground_truth_reasoning": false,
"hidden_act": "silu",
"hidden_size": 1536,
"image_token_id": 151655,
"initializer_range": 0.02,
"intermediate_size": 8960,
"is_tinyvla": false,
"llm_loss_weight": 1.0,
"load_pretrain_dit": false,
"max_position_embeddings": 32768,
"max_window_layers": 28,
"mm_use_im_patch_token": false,
"mm_use_im_start_end": false,
"model_name_or_path": "/data/public/wrt/m2/ccw/ckpt/dexvla/qwen2_vl_3_cameras_1_17_all_data_pretrain_6w_DiT_H_Non_EMA_full_param_stage_1_50/checkpoint-60000",
"model_pretrain": "/data/public/wrt/m2/ccw/ckpt/dexvla/qwen2_vl_3_cameras_1_17_all_data_pretrain_6w_DiT_H_Non_EMA_full_param_stage_1_50/checkpoint-60000",
"model_type": "dex_vla",
"non_lora_lr": null,
"num_attention_heads": 12,
"num_hidden_layers": 28,
"num_key_value_heads": 2,
"only_using_input_embeddings": false,
"policy_class": "dit_diffusion_policy",
"policy_head_config": {
"_name_or_path": "",
"add_cross_attention": false,
"architectures": null,
"bad_words_ids": null,
"begin_suppress_tokens": null,
"bos_token_id": null,
"chunk_size_feed_forward": 0,
"cond_dim": 1536,
"cross_attention_hidden_size": null,
"decoder_start_token_id": null,
"depth": 32,
"diversity_penalty": 0.0,
"do_sample": false,
"early_stopping": false,
"encoder_no_repeat_ngram_size": 0,
"eos_token_id": null,
"eval": false,
"exponential_decay_length_penalty": null,
"external_vision_encoder": "None",
"finetuning_task": null,
"forced_bos_token_id": null,
"forced_eos_token_id": null,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"input_dim": 14,
"is_decoder": false,
"is_encoder_decoder": false,
"is_tinyvla": false,
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"learn_sigma": false,
"length_penalty": 1.0,
"max_length": 20,
"min_length": 0,
"mlp_ratio": 4.0,
"model_type": "dit_diffusion_policy",
"n_emb": 1280,
"n_obs_steps": 2,
"no_repeat_ngram_size": 0,
"noise_samples": 1,
"num_beam_groups": 1,
"num_beams": 1,
"num_heads": 16,
"num_inference_timesteps": 10,
"num_queries": 50,
"num_return_sequences": 1,
"num_train_timesteps": 100,
"obs_as_cond": true,
"output_attentions": false,
"output_dim": 14,
"output_hidden_states": false,
"output_scores": false,
"pad_token_id": null,
"prediction_horizon": 50,
"prefix": null,
"problem_type": null,
"pruned_heads": {},
"remove_invalid_values": false,
"repetition_penalty": 1.0,
"return_dict": true,
"return_dict_in_generate": false,
"sep_token_id": null,
"state_dim": 14,
"suppress_tokens": null,
"task_specific_params": null,
"temperature": 1.0,
"tf_legacy_loss": false,
"tie_encoder_decoder": false,
"tie_word_embeddings": true,
"time_as_cond": true,
"tokenizer_class": null,
"top_k": 50,
"top_p": 1.0,
"torch_dtype": null,
"torchscript": false,
"typical_p": 1.0,
"use_bfloat16": false
},
"policy_head_size": "DiT_H",
"policy_head_type": "dit_diffusion_policy",
"pretrain_dit_path": "None",
"rms_norm_eps": 1e-06,
"rope_scaling": {
"mrope_section": [
16,
24,
24
],
"rope_type": "default",
"type": "default"
},
"rope_theta": 1000000.0,
"sliding_window": 32768,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.45.2",
"tune_mm_mlp_adapter": true,
"use_cache": true,
"use_sliding_window": false,
"using_all_reasoning_hidden": false,
"using_channel_cat": false,
"using_film": true,
"using_state": false,
"using_xattn": false,
"version": "v0",
"video_token_id": 151656,
"vision_config": {
"hidden_size": 1536,
"in_chans": 3,
"model_type": "dex_vla",
"spatial_patch_size": 14
},
"vision_end_token_id": 151653,
"vision_start_token_id": 151652,
"vision_token_id": 151654,
"vocab_size": 151936,
"with_llm_head": true,
"with_text_fcs": false
}