{ "architectures": [ "ModernBertForSequenceClassification" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 1, "classifier_activation": "silu", "classifier_bias": false, "classifier_dropout": 0.0, "classifier_pooling": "mean", "cls_token_id": 1, "decoder_bias": true, "deterministic_flash_attn": false, "embedding_dropout": 0.0, "eos_token_id": 2, "finetuning_task": "text-classification", "global_attn_every_n_layers": 3, "global_rope_theta": 160000.0, "gradient_checkpointing": false, "hidden_activation": "gelu", "hidden_size": 768, "id2label": { "0": "Biomedical & mechanistic science", "1": "Clinical cases & vignettes", "2": "Clinical guidelines & pathways", "3": "Commercial & promotional", "4": "Drugs, trials & regulation", "5": "Health IT, telemedicine & operations", "6": "Health services & facilities", "7": "Health workforce education & training", "8": "Medical devices, diagnostics & imaging", "9": "Occupational health & safety", "10": "Other health", "11": "Others", "12": "Patient education & lifestyle", "13": "Public health, policy & programs", "14": "Wellness, supplements & CAM" }, "initializer_cutoff_factor": 2.0, "initializer_range": 0.02, "intermediate_size": 1152, "label2id": { "Biomedical & mechanistic science": 0, "Clinical cases & vignettes": 1, "Clinical guidelines & pathways": 2, "Commercial & promotional": 3, "Drugs, trials & regulation": 4, "Health IT, telemedicine & operations": 5, "Health services & facilities": 6, "Health workforce education & training": 7, "Medical devices, diagnostics & imaging": 8, "Occupational health & safety": 9, "Other health": 10, "Others": 11, "Patient education & lifestyle": 12, "Public health, policy & programs": 13, "Wellness, supplements & CAM": 14 }, "layer_norm_eps": 1e-05, "local_attention": 128, "local_rope_theta": 10000.0, "max_position_embeddings": 8192, "mlp_bias": false, "mlp_dropout": 0.0, "model_type": "modernbert", "norm_bias": false, "norm_eps": 1e-05, "num_attention_heads": 12, "num_hidden_layers": 22, "pad_token_id": 0, "position_embedding_type": "absolute", "problem_type": "single_label_classification", "repad_logits_with_grad": false, "sep_token_id": 2, "sparse_pred_ignore_index": -100, "sparse_prediction": false, "torch_dtype": "float32", "transformers_version": "4.55.4", "vocab_size": 32768 }