| { | |
| "activation_function": "gelu", | |
| "architectures": [ | |
| "TrajectoryTransformerModel" | |
| ], | |
| "N": 100, | |
| "action_weight": 5, | |
| "attn_pdrop": 0.1, | |
| "batch_size": 256, | |
| "discount": 0.99, | |
| "bos_token_id": 50256, | |
| "embd_pdrop": 0.1, | |
| "learning_rate": 0.0006, | |
| "lr_decay": true, | |
| "eos_token_id": 50256, | |
| "initializer_range": 0.02, | |
| "layer_norm_epsilon": 1e-05, | |
| "model_type": "gpt2", | |
| "n_ctx": 1024, | |
| "n_embd": 32, | |
| "n_epochs_ref": 50, | |
| "n_head": 4, | |
| "n_layer": 4, | |
| "n_saves": 3, | |
| "resid_pdrop": 0.1, | |
| "reward_weight": 1, | |
| "seed": 42, | |
| "summary_activation": null, | |
| "summary_first_dropout": 0.1, | |
| "summary_proj_to_labels": true, | |
| "summary_type": "cls_index", | |
| "summary_use_proj": true, | |
| "vocab_size": 100, | |
| "step": 1, | |
| "subsampled_sequence_length": 10, | |
| "termination_penalty": -100, | |
| "value_weight": 1 | |
| } |