| { | |
| "architectures": [ | |
| "BertForMaskedLM" | |
| ], | |
| "attention_probs_dropout_prob": 0.1, | |
| "chunk_size": 256, | |
| "classifier_dropout": null, | |
| "d_conv": 7, | |
| "d_state": 64, | |
| "expand": 2, | |
| "gradient_checkpointing": false, | |
| "headdim": 64, | |
| "hidden_act": "gelu", | |
| "hidden_dropout_prob": 0.1, | |
| "hidden_size": 1024, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 4096, | |
| "is_data_dependent": true, | |
| "is_prenorm": false, | |
| "layer_norm_eps": 1e-12, | |
| "matrix_mixer_type": "hydra", | |
| "max_position_embedding": 256, | |
| "max_position_embeddings": 512, | |
| "model_type": "bert", | |
| "num_attention_heads": 16, | |
| "num_hidden_layers": 45, | |
| "pad_token_id": 0, | |
| "pool_all": false, | |
| "position_embedding_type": "absolute", | |
| "qk_dim": 16, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.46.3", | |
| "type_vocab_size": 2, | |
| "use_cache": true, | |
| "use_position_embeddings": false, | |
| "vocab_size": 30528 | |
| } | |