| { | |
| "_name_or_path": "answerdotai/ModernBERT-base", | |
| "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": 768, | |
| "id2label": { | |
| "0": "dictionary", | |
| "1": "django", | |
| "2": "list", | |
| "3": "matplotlib", | |
| "4": "numpy", | |
| "5": "pandas", | |
| "6": "python-2.7", | |
| "7": "python-3.x", | |
| "8": "regex", | |
| "9": "string" | |
| }, | |
| "initializer_cutoff_factor": 2.0, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 1152, | |
| "label2id": { | |
| "dictionary": 0, | |
| "django": 1, | |
| "list": 2, | |
| "matplotlib": 3, | |
| "numpy": 4, | |
| "pandas": 5, | |
| "python-2.7": 6, | |
| "python-3.x": 7, | |
| "regex": 8, | |
| "string": 9 | |
| }, | |
| "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": 50283, | |
| "position_embedding_type": "absolute", | |
| "problem_type": "multi_label_classification", | |
| "reference_compile": false, | |
| "sep_token_id": 50282, | |
| "sparse_pred_ignore_index": -100, | |
| "sparse_prediction": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.48.0.dev0", | |
| "vocab_size": 50368 | |
| } | |