| { | |
| "afn": "gelu", | |
| "architectures": [ | |
| "OpenAIGPTLMHeadModel" | |
| ], | |
| "attn_pdrop": 0.1, | |
| "embd_pdrop": 0.1, | |
| "initializer_range": 0.02, | |
| "layer_norm_epsilon": 1e-05, | |
| "model_type": "openai-gpt", | |
| "n_embd": 512, | |
| "n_head": 4, | |
| "n_layer": 4, | |
| "n_positions": 531, | |
| "resid_pdrop": 0.1, | |
| "summary_activation": null, | |
| "summary_first_dropout": 0.1, | |
| "summary_proj_to_labels": true, | |
| "summary_type": "cls_index", | |
| "summary_use_proj": true, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.34.0", | |
| "vocab_size": 40478 | |
| } | |