| { |
| "architectures": [ |
| "ModernBertSmallForMaskedLM" |
| ], |
| "attention_dropout": 0.0, |
| "attention_pattern": "alternating", |
| "auto_map": { |
| "AutoConfig": "configuration_modernbert_small.ModernBertSmallConfig", |
| "AutoModelForMaskedLM": "modeling_modernbert_small.ModernBertSmallForMaskedLM", |
| "AutoModel": "modeling_modernbert_small.ModernBertSmallModel" |
| }, |
| "dtype": "float32", |
| "embedding_dropout": 0.0, |
| "embedding_type": "factorized_linear", |
| "factorized_linear_bottleneck_dim": 128, |
| "factorized_mlp_bottleneck_dim": 128, |
| "global_attn_every_n_layers": 3, |
| "global_rope_theta": 160000.0, |
| "hidden_activation": "gelu", |
| "hidden_size": 384, |
| "initializer_range": 0.02, |
| "intermediate_size": 576, |
| "local_attention_window": 128, |
| "local_rope_theta": 10000.0, |
| "mlp_dropout": 0.0, |
| "model_type": "modernbert_small", |
| "norm_eps": 1e-05, |
| "num_attention_heads": 6, |
| "num_hidden_layers": 16, |
| "pad_token_id": 0, |
| "tie_word_embeddings": false, |
| "transformers_version": "4.57.6", |
| "vocab_size": 30522 |
| } |
