| { | |
| "architectures": [ | |
| "MegatronBertForNextSentencePrediction" | |
| ], | |
| "attention_probs_dropout_prob": 0.1, | |
| "embedding_size": 32, | |
| "hidden_act": "gelu", | |
| "hidden_dropout_prob": 0.1, | |
| "hidden_size": 64, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 37, | |
| "layer_norm_eps": 1e-12, | |
| "max_position_embeddings": 512, | |
| "model_type": "megatron-bert", | |
| "num_attention_heads": 4, | |
| "num_hidden_layers": 5, | |
| "pad_token_id": 0, | |
| "position_embedding_type": "absolute", | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.28.0.dev0", | |
| "type_vocab_size": 16, | |
| "use_cache": true, | |
| "vocab_size": 1124 | |
| } | |