| { |
| "_name_or_path": "microsoft/mdeberta-v3-base", |
| "architectures": [ |
| "DebertaV2ForSequenceClassification" |
| ], |
| "attention_probs_dropout_prob": 0.1, |
| "hidden_act": "gelu", |
| "hidden_dropout_prob": 0.1, |
| "hidden_size": 768, |
| "initializer_range": 0.02, |
| "intermediate_size": 3072, |
| "layer_norm_eps": 1e-07, |
| "max_position_embeddings": 512, |
| "max_relative_positions": -1, |
| "model_type": "deberta-v2", |
| "norm_rel_ebd": "layer_norm", |
| "num_attention_heads": 12, |
| "num_hidden_layers": 12, |
| "pad_token_id": 0, |
| "pooler_dropout": 0, |
| "pooler_hidden_act": "gelu", |
| "pooler_hidden_size": 768, |
| "pos_att_type": [ |
| "p2c", |
| "c2p" |
| ], |
| "position_biased_input": false, |
| "position_buckets": 256, |
| "relative_attention": true, |
| "share_att_key": true, |
| "torch_dtype": "float32", |
| "transformers_version": "4.28.1", |
| "type_vocab_size": 0, |
| "vocab_size": 250102 |
| } |
|
|
