| { |
| "action_chunk_size": 60, |
| "action_config": { |
| "_attn_implementation_autoset": true, |
| "action_chunk_size": 60, |
| "action_dim": 26, |
| "architectures": [ |
| "ActionExpertModel" |
| ], |
| "attn_implementation": "eager", |
| "auto_map": { |
| "AutoConfig": "configuration_action_expert.ActionExpertConfig", |
| "AutoModel": "modeling_action_expert.ActionExpertModel" |
| }, |
| "bias": false, |
| "dtype": "bfloat16", |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 1024, |
| "initializer_range": 0.02, |
| "input_hidden_size": 2048, |
| "intermediate_size": 4096, |
| "max_position_embeddings": 32768, |
| "model_type": "action_expert", |
| "num_attention_heads": 16, |
| "num_hidden_layers": 24, |
| "num_key_value_heads": 8, |
| "pad_token_id": 2, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": { |
| "factor": 2.0, |
| "type": "dynamic" |
| }, |
| "rope_theta": 1000000, |
| "state_dim": 26, |
| "state_token_num": 3, |
| "use_bfloat16": true, |
| "use_cache": true, |
| "use_flash_attn": false |
| }, |
| "architectures": [ |
| "GO1Model" |
| ], |
| "auto_map": { |
| "AutoConfig": "configuration_go1.GO1ModelConfig", |
| "AutoModel": "modeling_go1.GO1Model", |
| "AutoModelForCausalLM": "modeling_internlm2_go1.py.InternLM2ForCausalLMGO1" |
| }, |
| "bos_token_id": 1, |
| "downsample_ratio": 0.5, |
| "dtype": "bfloat16", |
| "dynamic_image_size": false, |
| "eos_token_id": 2, |
| "force_image_size": 448, |
| "img_context_token_id": 92546, |
| "initializer_range": 0.02, |
| "latent_planner_config": { |
| "action_dim": 1, |
| "attn_implementation": "eager", |
| "bias": false, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 1024, |
| "initializer_range": 0.02, |
| "input_hidden_size": 2048, |
| "intermediate_size": 2048, |
| "max_position_embeddings": 2048, |
| "model_type": "intermidiate_action_expert", |
| "num_attention_heads": 16, |
| "num_hidden_layers": 24, |
| "num_key_value_heads": 8, |
| "rms_norm_eps": 1e-06, |
| "rope_scaling": null, |
| "rope_theta": 10000, |
| "state_token_num": 0, |
| "use_cache": true, |
| "vocab_size": 32 |
| }, |
| "latent_planning": false, |
| "llm_config": { |
| "_attn_implementation_autoset": true, |
| "architectures": [ |
| "InternLM2ForCausalLMGO1" |
| ], |
| "attn_implementation": "flash_attention_2", |
| "auto_map": { |
| "AutoConfig": "configuration_internlm2.InternLM2Config", |
| "AutoModel": "modeling_internlm2_go1.InternLM2ForCausalLMGO1", |
| "AutoModelForCausalLM": "modeling_internlm2_go1.py.InternLM2ForCausalLMGO1" |
| }, |
| "bias": false, |
| "dtype": "bfloat16", |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 8192, |
| "max_position_embeddings": 32768, |
| "model_type": "internlm2", |
| "num_attention_heads": 16, |
| "num_hidden_layers": 24, |
| "num_key_value_heads": 8, |
| "pad_token_id": 2, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": { |
| "factor": 2.0, |
| "type": "dynamic" |
| }, |
| "rope_theta": 1000000, |
| "use_bfloat16": true, |
| "use_cache": true, |
| "vocab_size": 92553 |
| }, |
| "max_dynamic_patch": 6, |
| "min_dynamic_patch": 1, |
| "model_type": "go1", |
| "noise_scheduler_config": { |
| "beta_schedule": "squaredcos_cap_v2", |
| "clip_sample": false, |
| "num_inference_timesteps": 5, |
| "num_train_timesteps": 1000, |
| "prediction_type": "sample" |
| }, |
| "norm": true, |
| "output_attentions": false, |
| "pad2square": false, |
| "pad_token_id": 2, |
| "ps_version": "v2", |
| "select_layer": -1, |
| "template": "internlm2-chat", |
| "transformers_version": null, |
| "use_backbone_lora": 0, |
| "use_llm_lora": 0, |
| "use_thumbnail": false, |
| "vision_config": { |
| "_attn_implementation_autoset": true, |
| "architectures": [ |
| "InternVisionModel" |
| ], |
| "attention_dropout": 0.0, |
| "drop_path_rate": 0.1, |
| "dropout": 0.0, |
| "dtype": "bfloat16", |
| "hidden_act": "gelu", |
| "hidden_size": 1024, |
| "image_size": 448, |
| "initializer_factor": 1.0, |
| "initializer_range": 0.02, |
| "intermediate_size": 4096, |
| "layer_norm_eps": 1e-06, |
| "model_type": "intern_vit_6b", |
| "norm_type": "layer_norm", |
| "num_attention_heads": 16, |
| "num_channels": 3, |
| "num_hidden_layers": 24, |
| "patch_size": 14, |
| "qk_normalization": false, |
| "qkv_bias": true, |
| "use_bfloat16": true, |
| "use_flash_attn": true |
| } |
| } |
|
|
