| { | |
| "architectures": [ | |
| "FLASHForMaskedLM" | |
| ], | |
| "classifier_dropout": 0.1, | |
| "dropout": 0.0, | |
| "expansion_factor": 2, | |
| "gradient_checkpointing": false, | |
| "group_size": 256, | |
| "hidden_size": 768, | |
| "initializer_range": 0.02, | |
| "layer_norm_eps": 1e-05, | |
| "max_position_embeddings": 512, | |
| "model_type": "flash", | |
| "norm_type": "scalenorm", | |
| "num_hidden_layers": 24, | |
| "pad_token_id": 0, | |
| "query_key_dim": 128, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.17.0", | |
| "vocab_size": 12000 | |
| } |