config.json
992 B · 41 lines · json Raw
1 {
2 "architectures": [
3 "DebertaV2ForQuestionAnswering"
4 ],
5 "attention_probs_dropout_prob": 0.1,
6 "hidden_act": "gelu",
7 "hidden_dropout_prob": 0.1,
8 "hidden_size": 768,
9 "initializer_range": 0.02,
10 "intermediate_size": 3072,
11 "language": "english",
12 "layer_norm_eps": 1e-07,
13 "max_position_embeddings": 512,
14 "max_relative_positions": -1,
15 "model_type": "deberta-v2",
16 "name": "DebertaV2",
17 "norm_rel_ebd": "layer_norm",
18 "num_attention_heads": 12,
19 "num_hidden_layers": 12,
20 "pad_token_id": 0,
21 "pooler_dropout": 0,
22 "pooler_hidden_act": "gelu",
23 "pooler_hidden_size": 768,
24 "pos_att_type": [
25 "p2c",
26 "c2p"
27 ],
28 "position_biased_input": false,
29 "position_buckets": 256,
30 "relative_attention": true,
31 "share_att_key": true,
32 "summary_activation": "tanh",
33 "summary_last_dropout": 0,
34 "summary_type": "first",
35 "summary_use_proj": false,
36 "torch_dtype": "float32",
37 "transformers_version": "4.20.1",
38 "type_vocab_size": 0,
39 "vocab_size": 128100
40 }
41