{ "_name_or_path": "./NEWV2T_distilled_e1_seg_double", "architectures": [ "BertForMultiTask" ], "attention_probs_dropout_prob": 0.0, "classifier_dropout": null, "hidden_act": "gelu", "hidden_dropout_prob": 0.0, "hidden_size": 192, "initializer_range": 0.02, "intermediate_size": 768, "layer_norm_eps": 1e-12, "max_position_embeddings": 32, "model_type": "bert", "num_attention_heads": 8, "num_hidden_layers": 8, "pad_token_id": 0, "position_embedding_type": "absolute", "torch_dtype": "float32", "transformers_version": "4.44.2", "type_vocab_size": 2, "use_cache": true, "vocab_size": 39 }