hatmimoha commited on
Commit
0cdcc85
โ€ข
1 Parent(s): 59a42ba

Upload new model version 1.0.3

Browse files

Augment traning data and improve model annotation

config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/home/mohamed/Downloads/ar-ner.model/ar-ner.model",
3
  "architectures": [
4
  "BertForTokenClassification"
5
  ],
@@ -10,48 +10,48 @@
10
  "hidden_dropout_prob": 0.1,
11
  "hidden_size": 768,
12
  "id2label": {
13
- "0": "O",
14
- "1": "B-PERSON",
15
- "2": "I-PERSON",
16
- "3": "B-ORGANIZATION",
17
- "4": "I-ORGANIZATION",
18
- "5": "B-LOCATION",
19
- "6": "I-LOCATION",
20
- "7": "B-PRODUCT",
21
- "8": "I-PRODUCT",
22
- "9": "B-EVENT",
23
- "10": "I-EVENT",
24
- "11": "B-DATE",
25
- "12": "I-DATE",
26
- "13": "B-COMPETITION",
27
- "14": "I-COMPETITION",
28
- "15": "B-PRIZE",
29
- "16": "I-PRIZE",
30
- "17": "B-DISEASE",
31
- "18": "I-DISEASE"
32
  },
33
  "initializer_range": 0.02,
34
  "intermediate_size": 3072,
35
  "label2id": {
36
- "B-COMPETITION": 13,
37
- "B-DATE": 11,
38
- "B-DISEASE": 17,
39
- "B-EVENT": 9,
40
- "B-LOCATION": 5,
41
- "B-ORGANIZATION": 3,
42
- "B-PERSON": 1,
43
- "B-PRIZE": 15,
44
- "B-PRODUCT": 7,
45
- "I-COMPETITION": 14,
46
- "I-DATE": 12,
47
- "I-DISEASE": 18,
48
- "I-EVENT": 10,
49
- "I-LOCATION": 6,
50
- "I-ORGANIZATION": 4,
51
- "I-PERSON": 2,
52
- "I-PRIZE": 16,
53
- "I-PRODUCT": 8,
54
- "O": 0
55
  },
56
  "layer_norm_eps": 1e-12,
57
  "max_position_embeddings": 512,
@@ -62,7 +62,7 @@
62
  "pad_token_id": 0,
63
  "position_embedding_type": "absolute",
64
  "torch_dtype": "float32",
65
- "transformers_version": "4.19.0.dev0",
66
  "type_vocab_size": 2,
67
  "use_cache": true,
68
  "vocab_size": 32000
 
1
  {
2
+ "_name_or_path": "asafaya/bert-base-arabic",
3
  "architectures": [
4
  "BertForTokenClassification"
5
  ],
 
10
  "hidden_dropout_prob": 0.1,
11
  "hidden_size": 768,
12
  "id2label": {
13
+ "0": "B-COMPETITION",
14
+ "1": "B-DATE",
15
+ "2": "B-DISEASE",
16
+ "3": "B-EVENT",
17
+ "4": "B-LOCATION",
18
+ "5": "B-ORGANIZATION",
19
+ "6": "B-PERSON",
20
+ "7": "B-PRICE",
21
+ "8": "B-PRODUCT",
22
+ "9": "I-COMPETITION",
23
+ "10": "I-DATE",
24
+ "11": "I-DISEASE",
25
+ "12": "I-EVENT",
26
+ "13": "I-LOCATION",
27
+ "14": "I-ORGANIZATION",
28
+ "15": "I-PERSON",
29
+ "16": "I-PRICE",
30
+ "17": "I-PRODUCT",
31
+ "18": "O"
32
  },
33
  "initializer_range": 0.02,
34
  "intermediate_size": 3072,
35
  "label2id": {
36
+ "B-COMPETITION": 0,
37
+ "B-DATE": 1,
38
+ "B-DISEASE": 2,
39
+ "B-EVENT": 3,
40
+ "B-LOCATION": 4,
41
+ "B-ORGANIZATION": 5,
42
+ "B-PERSON": 6,
43
+ "B-PRICE": 7,
44
+ "B-PRODUCT": 8,
45
+ "I-COMPETITION": 9,
46
+ "I-DATE": 10,
47
+ "I-DISEASE": 11,
48
+ "I-EVENT": 12,
49
+ "I-LOCATION": 13,
50
+ "I-ORGANIZATION": 14,
51
+ "I-PERSON": 15,
52
+ "I-PRICE": 16,
53
+ "I-PRODUCT": 17,
54
+ "O": 18
55
  },
56
  "layer_norm_eps": 1e-12,
57
  "max_position_embeddings": 512,
 
62
  "pad_token_id": 0,
63
  "position_embedding_type": "absolute",
64
  "torch_dtype": "float32",
65
+ "transformers_version": "4.24.0",
66
  "type_vocab_size": 2,
67
  "use_cache": true,
68
  "vocab_size": 32000
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1c4b08fc317bff156e59d57465b473ada0766ebe0f127ec5e8c0b32cbb34caa7
3
  size 440235889
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a70dc2fa360ac2997727c20ee7a91e028595f2176238cac909ab79c9e5e3e29
3
  size 440235889
special_tokens_map.json CHANGED
@@ -1 +1,7 @@
1
- {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json CHANGED
@@ -1 +1,16 @@
1
- {"special_tokens_map_file": null, "full_tokenizer_file": null, "do_lower_case": false}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "do_basic_tokenize": true,
4
+ "do_lower_case": true,
5
+ "full_tokenizer_file": null,
6
+ "mask_token": "[MASK]",
7
+ "name_or_path": "asafaya/bert-base-arabic",
8
+ "never_split": null,
9
+ "pad_token": "[PAD]",
10
+ "sep_token": "[SEP]",
11
+ "special_tokens_map_file": null,
12
+ "strip_accents": null,
13
+ "tokenize_chinese_chars": true,
14
+ "tokenizer_class": "BertTokenizer",
15
+ "unk_token": "[UNK]"
16
+ }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1222658b502b480f9f2e2af5c50941fcd6ffe8a04d5beca3ea6083c3127015e
3
- size 3055
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b416605fdb4edc9a7a01f7e76cb8c86bf6f01e5f051c63f336895ac1adaca65
3
+ size 3375