| { | |
| "architectures": [ | |
| "ModernBertForSequenceClassification" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 50281, | |
| "classifier_activation": "gelu", | |
| "classifier_bias": false, | |
| "classifier_dropout": 0.0, | |
| "classifier_pooling": "mean", | |
| "cls_token_id": 50281, | |
| "decoder_bias": true, | |
| "deterministic_flash_attn": false, | |
| "embedding_dropout": 0.0, | |
| "eos_token_id": 50282, | |
| "global_attn_every_n_layers": 3, | |
| "global_rope_theta": 160000.0, | |
| "gradient_checkpointing": false, | |
| "hidden_activation": "gelu", | |
| "hidden_size": 1024, | |
| "id2label": { | |
| "0": "law_commentary", | |
| "1": "sociology", | |
| "2": "biography", | |
| "3": "literary_criticism", | |
| "4": "theatre_play", | |
| "5": "novel", | |
| "6": "law,_decree_and_rules", | |
| "7": "economics", | |
| "8": "court_transcripts", | |
| "9": "biology_and_medicine", | |
| "10": "lists_and_bibliographies", | |
| "11": "history", | |
| "12": "philosophy", | |
| "13": "political_news", | |
| "14": "other_science", | |
| "15": "political_science", | |
| "16": "foreign_news", | |
| "17": "advertising", | |
| "18": "dictionary", | |
| "19": "crime_news", | |
| "20": "art_criticism", | |
| "21": "music_criticism" | |
| }, | |
| "initializer_cutoff_factor": 2.0, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 2624, | |
| "label2id": { | |
| "advertising": 17, | |
| "art_criticism": 20, | |
| "biography": 2, | |
| "biology_and_medicine": 9, | |
| "court_transcripts": 8, | |
| "crime_news": 19, | |
| "dictionary": 18, | |
| "economics": 7, | |
| "foreign_news": 16, | |
| "history": 11, | |
| "law,_decree_and_rules": 6, | |
| "law_commentary": 0, | |
| "lists_and_bibliographies": 10, | |
| "literary_criticism": 3, | |
| "music_criticism": 21, | |
| "novel": 5, | |
| "other_science": 14, | |
| "philosophy": 12, | |
| "political_news": 13, | |
| "political_science": 15, | |
| "sociology": 1, | |
| "theatre_play": 4 | |
| }, | |
| "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": 16, | |
| "num_hidden_layers": 28, | |
| "pad_token_id": 50283, | |
| "position_embedding_type": "absolute", | |
| "problem_type": "multi_label_classification", | |
| "reference_compile": true, | |
| "repad_logits_with_grad": false, | |
| "sep_token_id": 50282, | |
| "sparse_pred_ignore_index": -100, | |
| "sparse_prediction": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.50.0", | |
| "vocab_size": 50368 | |
| } | |