| { |
| "alora_invocation_tokens": null, |
| "alpha_pattern": {}, |
| "arrow_config": null, |
| "auto_mapping": null, |
| "base_model_name_or_path": null, |
| "bias": "none", |
| "corda_config": null, |
| "ensure_weight_tying": false, |
| "eva_config": null, |
| "exclude_modules": null, |
| "fan_in_fan_out": false, |
| "inference_mode": true, |
| "init_lora_weights": true, |
| "layer_replication": null, |
| "layers_pattern": null, |
| "layers_to_transform": null, |
| "loftq_config": {}, |
| "lora_alpha": 16, |
| "lora_bias": false, |
| "lora_dropout": 0.0, |
| "megatron_config": null, |
| "megatron_core": "megatron.core", |
| "modules_to_save": [ |
| "action_expert.model.layers.16.input_layernorm.dense", |
| "vlm.model.language_model.layers.13.input_layernorm", |
| "action_expert.model.layers.17.post_attention_layernorm.dense", |
| "action_expert.model.layers.6.input_layernorm.dense", |
| "action_expert.model.layers.5.post_attention_layernorm.dense", |
| "action_expert.model.layers.13.input_layernorm.dense", |
| "vlm.model.language_model.layers.7.input_layernorm", |
| "vlm.model.language_model.layers.0.post_attention_layernorm", |
| "action_expert.model.layers.3.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.6.input_layernorm", |
| "action_expert.model.layers.7.post_attention_layernorm.dense", |
| "action_expert.model.layers.10.post_attention_layernorm.dense", |
| "action_expert.model.layers.5.input_layernorm.dense", |
| "vlm.model.language_model.layers.14.post_attention_layernorm", |
| "suffix_embedder.time_mlp_in", |
| "vlm.model.language_model.layers.11.post_attention_layernorm", |
| "vlm.model.language_model.layers.15.post_attention_layernorm", |
| "vlm.model.language_model.layers.17.input_layernorm", |
| "action_expert.model.layers.8.post_attention_layernorm.dense", |
| "action_expert.model.layers.1.post_attention_layernorm.dense", |
| "action_expert.model.layers.14.post_attention_layernorm.dense", |
| "action_out_proj", |
| "action_expert.model.layers.14.input_layernorm.dense", |
| "vlm.model.language_model.layers.2.post_attention_layernorm", |
| "vlm.model.language_model.layers.1.post_attention_layernorm", |
| "action_expert.model.layers.4.input_layernorm.dense", |
| "action_expert.model.layers.15.input_layernorm.dense", |
| "vlm.model.vision_tower.vision_model.embeddings.position_embedding", |
| "vlm.model.language_model.layers.15.input_layernorm", |
| "action_expert.model.layers.4.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.17.post_attention_layernorm", |
| "vlm.model.language_model.layers.0.input_layernorm", |
| "action_expert.model.layers.0.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.6.post_attention_layernorm", |
| "vlm.model.language_model.layers.4.input_layernorm", |
| "action_expert.model.layers.0.input_layernorm.dense", |
| "vlm.model.language_model.layers.3.post_attention_layernorm", |
| "vlm.model.language_model.layers.5.post_attention_layernorm", |
| "action_expert.model.layers.11.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.8.input_layernorm", |
| "vlm.model.language_model.layers.12.post_attention_layernorm", |
| "suffix_embedder.state_proj", |
| "action_expert.model.layers.12.input_layernorm.dense", |
| "vlm.model.language_model.layers.9.post_attention_layernorm", |
| "action_expert.model.layers.15.post_attention_layernorm.dense", |
| "action_expert.model.layers.8.input_layernorm.dense", |
| "vlm.model.language_model.layers.12.input_layernorm", |
| "action_expert.model.layers.16.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.4.post_attention_layernorm", |
| "suffix_embedder.state_mlp_out", |
| "vlm.model.language_model.layers.8.post_attention_layernorm", |
| "action_expert.model.layers.17.input_layernorm.dense", |
| "vlm.model.language_model.layers.10.input_layernorm", |
| "action_expert.model.layers.2.input_layernorm.dense", |
| "vlm.model.language_model.layers.11.input_layernorm", |
| "action_expert.model.layers.13.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.16.input_layernorm", |
| "action_expert.model.layers.1.input_layernorm.dense", |
| "action_expert.model.layers.9.input_layernorm.dense", |
| "vlm.model.language_model.layers.13.post_attention_layernorm", |
| "action_expert.model.layers.7.input_layernorm.dense", |
| "action_expert.model.layers.6.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.2.input_layernorm", |
| "vlm.model.language_model.layers.16.post_attention_layernorm", |
| "vlm.model.language_model.layers.14.input_layernorm", |
| "vlm.model.language_model.layers.10.post_attention_layernorm", |
| "suffix_embedder.action_in_proj", |
| "vlm.model.vision_tower.vision_model.embeddings.patch_embedding", |
| "vlm.model.language_model.layers.7.post_attention_layernorm", |
| "vlm.model.language_model.layers.9.input_layernorm", |
| "action_expert.model.layers.12.post_attention_layernorm.dense", |
| "action_expert.model.layers.9.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.5.input_layernorm", |
| "suffix_embedder.time_mlp_out", |
| "action_expert.model.layers.10.input_layernorm.dense", |
| "action_expert.model.layers.3.input_layernorm.dense", |
| "action_expert.model.layers.11.input_layernorm.dense", |
| "suffix_embedder.state_mlp_in", |
| "action_expert.model.layers.2.post_attention_layernorm.dense", |
| "vlm.model.language_model.layers.3.input_layernorm", |
| "vlm.model.language_model.layers.1.input_layernorm" |
| ], |
| "peft_type": "LORA", |
| "peft_version": "0.18.0", |
| "qalora_group_size": 16, |
| "r": 16, |
| "rank_pattern": {}, |
| "revision": null, |
| "target_modules": [ |
| "v_proj", |
| "q_proj", |
| "fc1", |
| "o_proj", |
| "down_proj", |
| "out_proj", |
| "gate_proj", |
| "up_proj", |
| "k_proj", |
| "fc2" |
| ], |
| "target_parameters": null, |
| "task_type": "FEATURE_EXTRACTION", |
| "trainable_token_indices": null, |
| "use_dora": false, |
| "use_qalora": false, |
| "use_rslora": false |
| } |
