{ "alora_invocation_tokens": null, "alpha_pattern": {}, "arrow_config": null, "auto_mapping": null, "base_model_name_or_path": "./checkpoints/DM05", "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, "lora_ga_config": null, "megatron_config": null, "megatron_core": "megatron.core", "modules_to_save": [ "action_in_proj", "action_out_proj", "time_mlp_in", "time_mlp_out", "input_time_modulators.0", "input_time_modulators.1", "input_time_modulators.2", "input_time_modulators.3", "input_time_modulators.4", "input_time_modulators.5", "input_time_modulators.6", "input_time_modulators.7", "input_time_modulators.8", "input_time_modulators.9", "input_time_modulators.10", "input_time_modulators.11", "input_time_modulators.12", "input_time_modulators.13", "input_time_modulators.14", "input_time_modulators.15", "input_time_modulators.16", "input_time_modulators.17", "input_time_modulators.18", "input_time_modulators.19", "input_time_modulators.20", "input_time_modulators.21", "input_time_modulators.22", "input_time_modulators.23", "input_time_modulators.24", "input_time_modulators.25", "input_time_modulators.26", "input_time_modulators.27", "input_time_modulators.28", "input_time_modulators.29", "input_time_modulators.30", "input_time_modulators.31", "input_time_modulators.32", "input_time_modulators.33", "mlp_time_modulators.0", "mlp_time_modulators.1", "mlp_time_modulators.2", "mlp_time_modulators.3", "mlp_time_modulators.4", "mlp_time_modulators.5", "mlp_time_modulators.6", "mlp_time_modulators.7", "mlp_time_modulators.8", "mlp_time_modulators.9", "mlp_time_modulators.10", "mlp_time_modulators.11", "mlp_time_modulators.12", "mlp_time_modulators.13", "mlp_time_modulators.14", "mlp_time_modulators.15", "mlp_time_modulators.16", "mlp_time_modulators.17", "mlp_time_modulators.18", "mlp_time_modulators.19", "mlp_time_modulators.20", "mlp_time_modulators.21", "mlp_time_modulators.22", "mlp_time_modulators.23", "mlp_time_modulators.24", "mlp_time_modulators.25", "mlp_time_modulators.26", "mlp_time_modulators.27", "mlp_time_modulators.28", "mlp_time_modulators.29", "mlp_time_modulators.30", "mlp_time_modulators.31", "mlp_time_modulators.32", "mlp_time_modulators.33", "final_time_modulator" ], "peft_type": "LORA", "peft_version": "0.19.1", "qalora_group_size": 16, "r": 32, "rank_pattern": {}, "revision": null, "target_modules": [ "input_time_modulators.17", "mlp_time_modulators.16", "fc2", "input_time_modulators.16", "input_time_modulators.32", "mlp_time_modulators.10", "mlp_time_modulators.32", "action_out_proj", "q_proj", "input_time_modulators.26", "up_proj", "fc1", "mlp_time_modulators.2", "input_time_modulators.23", "time_mlp_in", "mlp_time_modulators.6", "time_mlp_out", "mlp_time_modulators.3", "input_time_modulators.25", "gate_proj", "mlp_time_modulators.8", "final_time_modulator", "out_proj", "mlp_time_modulators.19", "mlp_time_modulators.12", "mlp_time_modulators.13", "mlp_time_modulators.7", "input_time_modulators.3", "input_time_modulators.13", "input_time_modulators.15", "mlp_time_modulators.15", "input_time_modulators.21", "mlp_time_modulators.9", "mlp_time_modulators.22", "input_time_modulators.29", "input_time_modulators.31", "mlp_time_modulators.0", "input_time_modulators.1", "mlp_time_modulators.11", "mlp_time_modulators.14", "mlp_time_modulators.18", "mlp_time_modulators.25", "mlp_time_modulators.4", "action_in_proj", "input_time_modulators.11", "input_time_modulators.22", "input_time_modulators.5", "mlp_time_modulators.23", "input_time_modulators.24", "mlp_time_modulators.26", "mlp_time_modulators.30", "o_proj", "mlp_time_modulators.29", "input_time_modulators.20", "down_proj", "input_time_modulators.8", "mlp_time_modulators.21", "input_time_modulators.19", "input_time_modulators.2", "k_proj", "input_time_modulators.28", "mlp_time_modulators.31", "mlp_time_modulators.5", "mlp_time_modulators.24", "input_time_modulators.33", "input_time_modulators.14", "input_time_modulators.9", "mlp_time_modulators.27", "input_time_modulators.7", "input_time_modulators.27", "mlp_time_modulators.28", "input_time_modulators.0", "mlp_time_modulators.17", "input_time_modulators.10", "input_time_modulators.12", "v_proj", "input_time_modulators.18", "mlp_time_modulators.33", "mlp_time_modulators.20", "input_time_modulators.4", "input_time_modulators.30", "input_time_modulators.6", "mlp_time_modulators.1" ], "target_parameters": null, "task_type": "CAUSAL_LM", "trainable_token_indices": null, "use_bdlora": null, "use_dora": false, "use_qalora": false, "use_rslora": false }