| { |
| "model_type": "modernbert", |
| "architectures": [ |
| "ModernBertForSequenceClassification" |
| ], |
| "num_labels": 2, |
| "id2label": { |
| "0": "SAFE", |
| "1": "VIOLATION" |
| }, |
| "label2id": { |
| "SAFE": 0, |
| "VIOLATION": 1 |
| }, |
| "problem_type": "single_label_classification", |
| "hidden_size": 768, |
| "num_hidden_layers": 22, |
| "num_attention_heads": 12, |
| "intermediate_size": 3072, |
| "hidden_act": "gelu", |
| "hidden_dropout_prob": 0.0, |
| "attention_probs_dropout_prob": 0.0, |
| "max_position_embeddings": 8192, |
| "type_vocab_size": 1, |
| "initializer_range": 0.02, |
| "layer_norm_eps": 1e-05, |
| "pad_token_id": 50283, |
| "vocab_size": 50368, |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "classifier_pooling": "cls", |
| "classifier_dropout": 0.0, |
| "classifier_bias": false, |
| "decoder_bias": true, |
| "global_rope_theta": 160000.0, |
| "local_attention": 128, |
| "local_rope_theta": 10000.0, |
| "reference_compile": "3.13", |
| "norm_bias": false, |
| "norm_eps": 1e-05, |
| "torch_dtype": "bfloat16", |
| "transformers_version": "4.48.0.dev0" |
| } |
|
|
|
|
