{ "architectures": [ "RobertaForSequenceClassification" ], "attention_probs_dropout_prob": 0.1, "bos_token_id": 0, "eos_token_id": 2, "finetuning_task": "ours", "gradient_checkpointing": false, "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 1024, "id2label": { "0": "motivation", "1": "background", "2": "uses", "3": "extends", "4": "similarities", "5": "differences", "6": "future_work" }, "initializer_range": 0.02, "intermediate_size": 4096, "label2id": { "LABEL_0": 0, "LABEL_1": 1, "LABEL_2": 2, "LABEL_3": 3, "LABEL_4": 4, "LABEL_5": 5, "LABEL_6": 6 }, "layer_norm_eps": 1e-05, "max_position_embeddings": 514, "model_type": "roberta", "num_attention_heads": 16, "num_hidden_layers": 24, "pad_token_id": 1, "type_vocab_size": 1, "vocab_size": 50265 }