Upload new model version 1.0.3
Browse filesAugment traning data and improve model annotation
- config.json +40 -40
- pytorch_model.bin +1 -1
- special_tokens_map.json +7 -1
- tokenizer.json +0 -0
- tokenizer_config.json +16 -1
- training_args.bin +2 -2
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "/
|
| 3 |
"architectures": [
|
| 4 |
"BertForTokenClassification"
|
| 5 |
],
|
|
@@ -10,48 +10,48 @@
|
|
| 10 |
"hidden_dropout_prob": 0.1,
|
| 11 |
"hidden_size": 768,
|
| 12 |
"id2label": {
|
| 13 |
-
"0": "
|
| 14 |
-
"1": "B-
|
| 15 |
-
"2": "
|
| 16 |
-
"3": "B-
|
| 17 |
-
"4": "
|
| 18 |
-
"5": "B-
|
| 19 |
-
"6": "
|
| 20 |
-
"7": "B-
|
| 21 |
-
"8": "
|
| 22 |
-
"9": "
|
| 23 |
-
"10": "I-
|
| 24 |
-
"11": "
|
| 25 |
-
"12": "I-
|
| 26 |
-
"13": "
|
| 27 |
-
"14": "I-
|
| 28 |
-
"15": "
|
| 29 |
-
"16": "I-
|
| 30 |
-
"17": "
|
| 31 |
-
"18": "
|
| 32 |
},
|
| 33 |
"initializer_range": 0.02,
|
| 34 |
"intermediate_size": 3072,
|
| 35 |
"label2id": {
|
| 36 |
-
"B-COMPETITION":
|
| 37 |
-
"B-DATE":
|
| 38 |
-
"B-DISEASE":
|
| 39 |
-
"B-EVENT":
|
| 40 |
-
"B-LOCATION":
|
| 41 |
-
"B-ORGANIZATION":
|
| 42 |
-
"B-PERSON":
|
| 43 |
-
"B-
|
| 44 |
-
"B-PRODUCT":
|
| 45 |
-
"I-COMPETITION":
|
| 46 |
-
"I-DATE":
|
| 47 |
-
"I-DISEASE":
|
| 48 |
-
"I-EVENT":
|
| 49 |
-
"I-LOCATION":
|
| 50 |
-
"I-ORGANIZATION":
|
| 51 |
-
"I-PERSON":
|
| 52 |
-
"I-
|
| 53 |
-
"I-PRODUCT":
|
| 54 |
-
"O":
|
| 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.
|
| 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:
|
| 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 |
-
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
-
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b416605fdb4edc9a7a01f7e76cb8c86bf6f01e5f051c63f336895ac1adaca65
|
| 3 |
+
size 3375
|