Hugging Face's logo

SIRIS-Lab
/
citation-parser-SPAN

Token Classification
Transformers
Safetensors
deberta-v2
Model card Files
xet
Community
nicolauduran45 commited on
Commit
162a74a
·
verified ·
1 Parent(s): 4ce23b3

Upload DebertaV2ForTokenClassification

Browse files
Files changed (2) hide show
  1. config.json +50 -47
  2. model.safetensors +2 -2
config.json CHANGED
@@ -1,60 +1,63 @@
1
  {
2
  "architectures": [
3
- "ModernBertForTokenClassification"
4
  ],
5
- "attention_bias": false,
6
- "attention_dropout": 0.0,
7
- "bos_token_id": 50281,
8
- "classifier_activation": "gelu",
9
- "classifier_bias": false,
10
- "classifier_dropout": 0.0,
11
- "classifier_pooling": "mean",
12
- "cls_token_id": 50281,
13
- "decoder_bias": true,
14
- "deterministic_flash_attn": false,
15
- "embedding_dropout": 0.0,
16
- "eos_token_id": 50282,
17
- "global_attn_every_n_layers": 3,
18
- "global_rope_theta": 160000.0,
19
- "gradient_checkpointing": false,
20
- "hidden_activation": "gelu",
21
  "hidden_size": 768,
22
  "id2label": {
23
  "0": "O",
24
- "1": "B-CITATION-SPAN",
25
- "2": "B-CITATION-REFEERENCE",
26
- "3": "I-CITATION-SPAN",
27
- "4": "I-CITATION-REFEERENCE"
 
 
 
 
 
 
28
  },
29
- "initializer_cutoff_factor": 2.0,
30
  "initializer_range": 0.02,
31
- "intermediate_size": 1152,
32
  "label2id": {
33
- "B-CITATION-REFEERENCE": 2,
34
- "B-CITATION-SPAN": 1,
35
- "I-CITATION-REFEERENCE": 4,
36
- "I-CITATION-SPAN": 3,
 
 
 
 
 
 
37
  "O": 0
38
  },
39
- "layer_norm_eps": 1e-05,
40
- "local_attention": 128,
41
- "local_rope_theta": 10000.0,
42
- "max_position_embeddings": 8192,
43
- "mlp_bias": false,
44
- "mlp_dropout": 0.0,
45
- "model_type": "modernbert",
46
- "norm_bias": false,
47
- "norm_eps": 1e-05,
48
  "num_attention_heads": 12,
49
- "num_hidden_layers": 22,
50
- "pad_token_id": 50283,
51
- "position_embedding_type": "absolute",
52
- "reference_compile": true,
53
- "repad_logits_with_grad": false,
54
- "sep_token_id": 50282,
55
- "sparse_pred_ignore_index": -100,
56
- "sparse_prediction": false,
57
- "torch_dtype": "float32",
58
- "transformers_version": "4.50.3",
59
- "vocab_size": 50368
 
 
 
 
 
60
  }
 
1
  {
2
  "architectures": [
3
+ "DebertaV2ForTokenClassification"
4
  ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 1,
7
+ "dtype": "float32",
8
+ "eos_token_id": 2,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
 
 
 
 
 
 
 
 
 
 
11
  "hidden_size": 768,
12
  "id2label": {
13
  "0": "O",
14
+ "1": "B-CITATION_REF",
15
+ "2": "I-CITATION_REF",
16
+ "3": "B-AUTHOR",
17
+ "4": "I-AUTHOR",
18
+ "5": "B-YEAR",
19
+ "6": "I-YEAR",
20
+ "7": "B-CITATION_ID",
21
+ "8": "I-CITATION_ID",
22
+ "9": "B-CITATION_SPAN",
23
+ "10": "I-CITATION_SPAN"
24
  },
 
25
  "initializer_range": 0.02,
26
+ "intermediate_size": 3072,
27
  "label2id": {
28
+ "B-AUTHOR": 3,
29
+ "B-CITATION_ID": 7,
30
+ "B-CITATION_REF": 1,
31
+ "B-CITATION_SPAN": 9,
32
+ "B-YEAR": 5,
33
+ "I-AUTHOR": 4,
34
+ "I-CITATION_ID": 8,
35
+ "I-CITATION_REF": 2,
36
+ "I-CITATION_SPAN": 10,
37
+ "I-YEAR": 6,
38
  "O": 0
39
  },
40
+ "layer_norm_eps": 1e-07,
41
+ "legacy": true,
42
+ "max_position_embeddings": 512,
43
+ "max_relative_positions": -1,
44
+ "model_type": "deberta-v2",
45
+ "norm_rel_ebd": "layer_norm",
 
 
 
46
  "num_attention_heads": 12,
47
+ "num_hidden_layers": 12,
48
+ "pad_token_id": 0,
49
+ "pooler_dropout": 0,
50
+ "pooler_hidden_act": "gelu",
51
+ "pooler_hidden_size": 768,
52
+ "pos_att_type": [
53
+ "p2c",
54
+ "c2p"
55
+ ],
56
+ "position_biased_input": false,
57
+ "position_buckets": 256,
58
+ "relative_attention": true,
59
+ "share_att_key": true,
60
+ "transformers_version": "4.56.1",
61
+ "type_vocab_size": 0,
62
+ "vocab_size": 128100
63
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e363dd03ffe0982e070f7c57c29052f8d63798e129a7bca0fdcdabf213b4c89f
3
- size 598449012
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:195ec56f286b7ae8de8e718a80d8a93109f410e22f401ced3c85e79725f4a6f8
3
+ size 735384420