TeeA/t5-text2sql-english-unidecode
Browse files- README.md +12 -17
- config.json +1 -1
- model.safetensors +2 -2
- runs/Mar18_11-43-10_bf3d97118b80/events.out.tfevents.1710762208.bf3d97118b80.153.0 +3 -0
- runs/Mar18_11-43-10_bf3d97118b80/events.out.tfevents.1710762265.bf3d97118b80.153.1 +3 -0
- runs/Mar18_11-47-52_bf3d97118b80/events.out.tfevents.1710762486.bf3d97118b80.153.2 +3 -0
- runs/Mar18_11-49-54_bf3d97118b80/events.out.tfevents.1710762596.bf3d97118b80.153.3 +3 -0
- runs/Mar18_11-49-54_bf3d97118b80/events.out.tfevents.1710762662.bf3d97118b80.153.4 +3 -0
- runs/Mar18_11-54-40_bf3d97118b80/events.out.tfevents.1710762891.bf3d97118b80.153.5 +3 -0
- runs/Mar18_11-54-40_bf3d97118b80/events.out.tfevents.1710762980.bf3d97118b80.153.6 +3 -0
- runs/Mar18_12-01-19_bf3d97118b80/events.out.tfevents.1710763287.bf3d97118b80.153.7 +3 -0
- runs/Mar18_12-01-19_bf3d97118b80/events.out.tfevents.1710763397.bf3d97118b80.153.8 +3 -0
- runs/Mar18_12-08-22_bf3d97118b80/events.out.tfevents.1710763716.bf3d97118b80.153.9 +3 -0
- runs/Mar18_12-08-22_bf3d97118b80/events.out.tfevents.1710763808.bf3d97118b80.153.10 +3 -0
- tokenizer.json +0 -207
- tokenizer_config.json +0 -184
- training_args.bin +1 -1
README.md
CHANGED
@@ -17,11 +17,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
|
18 |
This model is a fine-tuned version of [cssupport/t5-small-awesome-text-to-sql](https://huggingface.co/cssupport/t5-small-awesome-text-to-sql) on the None dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss:
|
21 |
-
- Rouge1: 0.
|
22 |
-
- Rouge2: 0.
|
23 |
-
- Rougel: 0.
|
24 |
-
- Rougelsum: 0.
|
25 |
|
26 |
## Model description
|
27 |
|
@@ -46,28 +46,23 @@ The following hyperparameters were used during training:
|
|
46 |
- seed: 42
|
47 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
48 |
- lr_scheduler_type: linear
|
49 |
-
- num_epochs:
|
50 |
- mixed_precision_training: Native AMP
|
51 |
|
52 |
### Training results
|
53 |
|
54 |
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|
55 |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
|
56 |
-
| No log | 1.0 |
|
57 |
-
| No log | 2.0 |
|
58 |
-
| No log | 3.0 |
|
59 |
-
| No log | 4.0 |
|
60 |
-
| No log | 5.0 |
|
61 |
-
| No log | 6.0 | 90 | 0.9272 | 0.4521 | 0.3272 | 0.4424 | 0.4454 |
|
62 |
-
| No log | 7.0 | 105 | 0.9151 | 0.4499 | 0.3247 | 0.4413 | 0.4440 |
|
63 |
-
| No log | 8.0 | 120 | 0.9071 | 0.4510 | 0.3247 | 0.4424 | 0.4446 |
|
64 |
-
| No log | 9.0 | 135 | 0.9022 | 0.4529 | 0.3258 | 0.4443 | 0.4461 |
|
65 |
-
| No log | 10.0 | 150 | 0.9005 | 0.4529 | 0.3258 | 0.4443 | 0.4461 |
|
66 |
|
67 |
|
68 |
### Framework versions
|
69 |
|
70 |
- Transformers 4.38.2
|
71 |
-
- Pytorch 2.1
|
72 |
- Datasets 2.18.0
|
73 |
- Tokenizers 0.15.2
|
|
|
17 |
|
18 |
This model is a fine-tuned version of [cssupport/t5-small-awesome-text-to-sql](https://huggingface.co/cssupport/t5-small-awesome-text-to-sql) on the None dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 2.0454
|
21 |
+
- Rouge1: 0.3934
|
22 |
+
- Rouge2: 0.2246
|
23 |
+
- Rougel: 0.3769
|
24 |
+
- Rougelsum: 0.3750
|
25 |
|
26 |
## Model description
|
27 |
|
|
|
46 |
- seed: 42
|
47 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
48 |
- lr_scheduler_type: linear
|
49 |
+
- num_epochs: 5
|
50 |
- mixed_precision_training: Native AMP
|
51 |
|
52 |
### Training results
|
53 |
|
54 |
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|
55 |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
|
56 |
+
| No log | 1.0 | 21 | 2.6820 | 0.3744 | 0.1996 | 0.3567 | 0.3541 |
|
57 |
+
| No log | 2.0 | 42 | 2.3569 | 0.3835 | 0.2191 | 0.3667 | 0.3651 |
|
58 |
+
| No log | 3.0 | 63 | 2.1723 | 0.3904 | 0.2216 | 0.3746 | 0.3725 |
|
59 |
+
| No log | 4.0 | 84 | 2.0758 | 0.3899 | 0.2214 | 0.3728 | 0.3708 |
|
60 |
+
| No log | 5.0 | 105 | 2.0454 | 0.3934 | 0.2246 | 0.3769 | 0.3750 |
|
|
|
|
|
|
|
|
|
|
|
61 |
|
62 |
|
63 |
### Framework versions
|
64 |
|
65 |
- Transformers 4.38.2
|
66 |
+
- Pytorch 2.2.1+cu121
|
67 |
- Datasets 2.18.0
|
68 |
- Tokenizers 0.15.2
|
config.json
CHANGED
@@ -57,5 +57,5 @@
|
|
57 |
"torch_dtype": "float32",
|
58 |
"transformers_version": "4.38.2",
|
59 |
"use_cache": true,
|
60 |
-
"vocab_size":
|
61 |
}
|
|
|
57 |
"torch_dtype": "float32",
|
58 |
"transformers_version": "4.38.2",
|
59 |
"use_cache": true,
|
60 |
+
"vocab_size": 32128
|
61 |
}
|
model.safetensors
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 sha256:82460e941372d5af9c6df8c6f70a2c43100dbfd50a667280301bd585bf5c33d3
|
3 |
+
size 242041896
|
runs/Mar18_11-43-10_bf3d97118b80/events.out.tfevents.1710762208.bf3d97118b80.153.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:14f8b11d22143ca38ccb2e846633afbe059e3e8cffb283fc235bc039cf939c1b
|
3 |
+
size 7294
|
runs/Mar18_11-43-10_bf3d97118b80/events.out.tfevents.1710762265.bf3d97118b80.153.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:99a0c75a2695784a8430962591e3026b8e28d68356e9f1f5bec9d21b2519cb46
|
3 |
+
size 553
|
runs/Mar18_11-47-52_bf3d97118b80/events.out.tfevents.1710762486.bf3d97118b80.153.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4a8d9c22fdd2f019fe5e8e5d8b81d00fac02872f623c08ad3e89a838c9828285
|
3 |
+
size 488
|
runs/Mar18_11-49-54_bf3d97118b80/events.out.tfevents.1710762596.bf3d97118b80.153.3
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:702a99575270602aab139cc6092fa15211d649ad573f7d4ef1a2a52057567eab
|
3 |
+
size 8289
|
runs/Mar18_11-49-54_bf3d97118b80/events.out.tfevents.1710762662.bf3d97118b80.153.4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b88227b038c4906dcb6fa1fe70fefedefecd816abc10a2e057d357a31ba05c41
|
3 |
+
size 553
|
runs/Mar18_11-54-40_bf3d97118b80/events.out.tfevents.1710762891.bf3d97118b80.153.5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:183dba51fe5fcce917b552eadcab0f3b62f1d37ad6edffd98604deae6127c05d
|
3 |
+
size 8689
|
runs/Mar18_11-54-40_bf3d97118b80/events.out.tfevents.1710762980.bf3d97118b80.153.6
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:428e60d7dd0912937ab7d671e83f83eeb58de999421404437cb16d449126656a
|
3 |
+
size 8754
|
runs/Mar18_12-01-19_bf3d97118b80/events.out.tfevents.1710763287.bf3d97118b80.153.7
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b250567253ac3e4bdf21535e8785410720f0f796708f80cea431cb6b4e9def2d
|
3 |
+
size 11057
|
runs/Mar18_12-01-19_bf3d97118b80/events.out.tfevents.1710763397.bf3d97118b80.153.8
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7cceb931736aa0315500c682f2289e6697d271b8395e9dd9147c602281f68781
|
3 |
+
size 562
|
runs/Mar18_12-08-22_bf3d97118b80/events.out.tfevents.1710763716.bf3d97118b80.153.9
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:66b5f15040f0598c01cefe740d7e27a613dcda66be74725bb5f6b5f3d2acb490
|
3 |
+
size 8689
|
runs/Mar18_12-08-22_bf3d97118b80/events.out.tfevents.1710763808.bf3d97118b80.153.10
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0fe82df4617dc0f0b8537e82f7a59263068867fbd2ca9333c3a0543d9440408d
|
3 |
+
size 553
|
tokenizer.json
CHANGED
@@ -941,213 +941,6 @@
|
|
941 |
"rstrip": false,
|
942 |
"normalized": false,
|
943 |
"special": true
|
944 |
-
},
|
945 |
-
{
|
946 |
-
"id": 32100,
|
947 |
-
"content": "ữ",
|
948 |
-
"single_word": false,
|
949 |
-
"lstrip": false,
|
950 |
-
"rstrip": false,
|
951 |
-
"normalized": true,
|
952 |
-
"special": false
|
953 |
-
},
|
954 |
-
{
|
955 |
-
"id": 32101,
|
956 |
-
"content": "ử",
|
957 |
-
"single_word": false,
|
958 |
-
"lstrip": false,
|
959 |
-
"rstrip": false,
|
960 |
-
"normalized": true,
|
961 |
-
"special": false
|
962 |
-
},
|
963 |
-
{
|
964 |
-
"id": 32102,
|
965 |
-
"content": "ị",
|
966 |
-
"single_word": false,
|
967 |
-
"lstrip": false,
|
968 |
-
"rstrip": false,
|
969 |
-
"normalized": true,
|
970 |
-
"special": false
|
971 |
-
},
|
972 |
-
{
|
973 |
-
"id": 32103,
|
974 |
-
"content": "ọ",
|
975 |
-
"single_word": false,
|
976 |
-
"lstrip": false,
|
977 |
-
"rstrip": false,
|
978 |
-
"normalized": true,
|
979 |
-
"special": false
|
980 |
-
},
|
981 |
-
{
|
982 |
-
"id": 32104,
|
983 |
-
"content": "ậ",
|
984 |
-
"single_word": false,
|
985 |
-
"lstrip": false,
|
986 |
-
"rstrip": false,
|
987 |
-
"normalized": true,
|
988 |
-
"special": false
|
989 |
-
},
|
990 |
-
{
|
991 |
-
"id": 32105,
|
992 |
-
"content": "ấ",
|
993 |
-
"single_word": false,
|
994 |
-
"lstrip": false,
|
995 |
-
"rstrip": false,
|
996 |
-
"normalized": true,
|
997 |
-
"special": false
|
998 |
-
},
|
999 |
-
{
|
1000 |
-
"id": 32106,
|
1001 |
-
"content": "ế",
|
1002 |
-
"single_word": false,
|
1003 |
-
"lstrip": false,
|
1004 |
-
"rstrip": false,
|
1005 |
-
"normalized": true,
|
1006 |
-
"special": false
|
1007 |
-
},
|
1008 |
-
{
|
1009 |
-
"id": 32107,
|
1010 |
-
"content": "ớ",
|
1011 |
-
"single_word": false,
|
1012 |
-
"lstrip": false,
|
1013 |
-
"rstrip": false,
|
1014 |
-
"normalized": true,
|
1015 |
-
"special": false
|
1016 |
-
},
|
1017 |
-
{
|
1018 |
-
"id": 32108,
|
1019 |
-
"content": "ờ",
|
1020 |
-
"single_word": false,
|
1021 |
-
"lstrip": false,
|
1022 |
-
"rstrip": false,
|
1023 |
-
"normalized": true,
|
1024 |
-
"special": false
|
1025 |
-
},
|
1026 |
-
{
|
1027 |
-
"id": 32109,
|
1028 |
-
"content": "ỹ",
|
1029 |
-
"single_word": false,
|
1030 |
-
"lstrip": false,
|
1031 |
-
"rstrip": false,
|
1032 |
-
"normalized": true,
|
1033 |
-
"special": false
|
1034 |
-
},
|
1035 |
-
{
|
1036 |
-
"id": 32110,
|
1037 |
-
"content": "ạ",
|
1038 |
-
"single_word": false,
|
1039 |
-
"lstrip": false,
|
1040 |
-
"rstrip": false,
|
1041 |
-
"normalized": true,
|
1042 |
-
"special": false
|
1043 |
-
},
|
1044 |
-
{
|
1045 |
-
"id": 32111,
|
1046 |
-
"content": "ề",
|
1047 |
-
"single_word": false,
|
1048 |
-
"lstrip": false,
|
1049 |
-
"rstrip": false,
|
1050 |
-
"normalized": true,
|
1051 |
-
"special": false
|
1052 |
-
},
|
1053 |
-
{
|
1054 |
-
"id": 32112,
|
1055 |
-
"content": "ủ",
|
1056 |
-
"single_word": false,
|
1057 |
-
"lstrip": false,
|
1058 |
-
"rstrip": false,
|
1059 |
-
"normalized": true,
|
1060 |
-
"special": false
|
1061 |
-
},
|
1062 |
-
{
|
1063 |
-
"id": 32113,
|
1064 |
-
"content": "ì",
|
1065 |
-
"single_word": false,
|
1066 |
-
"lstrip": false,
|
1067 |
-
"rstrip": false,
|
1068 |
-
"normalized": true,
|
1069 |
-
"special": false
|
1070 |
-
},
|
1071 |
-
{
|
1072 |
-
"id": 32114,
|
1073 |
-
"content": "ả",
|
1074 |
-
"single_word": false,
|
1075 |
-
"lstrip": false,
|
1076 |
-
"rstrip": false,
|
1077 |
-
"normalized": true,
|
1078 |
-
"special": false
|
1079 |
-
},
|
1080 |
-
{
|
1081 |
-
"id": 32115,
|
1082 |
-
"content": "ú",
|
1083 |
-
"single_word": false,
|
1084 |
-
"lstrip": false,
|
1085 |
-
"rstrip": false,
|
1086 |
-
"normalized": true,
|
1087 |
-
"special": false
|
1088 |
-
},
|
1089 |
-
{
|
1090 |
-
"id": 32116,
|
1091 |
-
"content": "ố",
|
1092 |
-
"single_word": false,
|
1093 |
-
"lstrip": false,
|
1094 |
-
"rstrip": false,
|
1095 |
-
"normalized": true,
|
1096 |
-
"special": false
|
1097 |
-
},
|
1098 |
-
{
|
1099 |
-
"id": 32117,
|
1100 |
-
"content": "ầ",
|
1101 |
-
"single_word": false,
|
1102 |
-
"lstrip": false,
|
1103 |
-
"rstrip": false,
|
1104 |
-
"normalized": true,
|
1105 |
-
"special": false
|
1106 |
-
},
|
1107 |
-
{
|
1108 |
-
"id": 32118,
|
1109 |
-
"content": "ự",
|
1110 |
-
"single_word": false,
|
1111 |
-
"lstrip": false,
|
1112 |
-
"rstrip": false,
|
1113 |
-
"normalized": true,
|
1114 |
-
"special": false
|
1115 |
-
},
|
1116 |
-
{
|
1117 |
-
"id": 32119,
|
1118 |
-
"content": "ễ",
|
1119 |
-
"single_word": false,
|
1120 |
-
"lstrip": false,
|
1121 |
-
"rstrip": false,
|
1122 |
-
"normalized": true,
|
1123 |
-
"special": false
|
1124 |
-
},
|
1125 |
-
{
|
1126 |
-
"id": 32120,
|
1127 |
-
"content": "í",
|
1128 |
-
"single_word": false,
|
1129 |
-
"lstrip": false,
|
1130 |
-
"rstrip": false,
|
1131 |
-
"normalized": true,
|
1132 |
-
"special": false
|
1133 |
-
},
|
1134 |
-
{
|
1135 |
-
"id": 32121,
|
1136 |
-
"content": "ý",
|
1137 |
-
"single_word": false,
|
1138 |
-
"lstrip": false,
|
1139 |
-
"rstrip": false,
|
1140 |
-
"normalized": true,
|
1141 |
-
"special": false
|
1142 |
-
},
|
1143 |
-
{
|
1144 |
-
"id": 32122,
|
1145 |
-
"content": "ệ",
|
1146 |
-
"single_word": false,
|
1147 |
-
"lstrip": false,
|
1148 |
-
"rstrip": false,
|
1149 |
-
"normalized": true,
|
1150 |
-
"special": false
|
1151 |
}
|
1152 |
],
|
1153 |
"normalizer": {
|
|
|
941 |
"rstrip": false,
|
942 |
"normalized": false,
|
943 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
944 |
}
|
945 |
],
|
946 |
"normalizer": {
|
tokenizer_config.json
CHANGED
@@ -824,190 +824,6 @@
|
|
824 |
"rstrip": false,
|
825 |
"single_word": false,
|
826 |
"special": true
|
827 |
-
},
|
828 |
-
"32100": {
|
829 |
-
"content": "ữ",
|
830 |
-
"lstrip": false,
|
831 |
-
"normalized": true,
|
832 |
-
"rstrip": false,
|
833 |
-
"single_word": false,
|
834 |
-
"special": false
|
835 |
-
},
|
836 |
-
"32101": {
|
837 |
-
"content": "ử",
|
838 |
-
"lstrip": false,
|
839 |
-
"normalized": true,
|
840 |
-
"rstrip": false,
|
841 |
-
"single_word": false,
|
842 |
-
"special": false
|
843 |
-
},
|
844 |
-
"32102": {
|
845 |
-
"content": "ị",
|
846 |
-
"lstrip": false,
|
847 |
-
"normalized": true,
|
848 |
-
"rstrip": false,
|
849 |
-
"single_word": false,
|
850 |
-
"special": false
|
851 |
-
},
|
852 |
-
"32103": {
|
853 |
-
"content": "ọ",
|
854 |
-
"lstrip": false,
|
855 |
-
"normalized": true,
|
856 |
-
"rstrip": false,
|
857 |
-
"single_word": false,
|
858 |
-
"special": false
|
859 |
-
},
|
860 |
-
"32104": {
|
861 |
-
"content": "ậ",
|
862 |
-
"lstrip": false,
|
863 |
-
"normalized": true,
|
864 |
-
"rstrip": false,
|
865 |
-
"single_word": false,
|
866 |
-
"special": false
|
867 |
-
},
|
868 |
-
"32105": {
|
869 |
-
"content": "ấ",
|
870 |
-
"lstrip": false,
|
871 |
-
"normalized": true,
|
872 |
-
"rstrip": false,
|
873 |
-
"single_word": false,
|
874 |
-
"special": false
|
875 |
-
},
|
876 |
-
"32106": {
|
877 |
-
"content": "ế",
|
878 |
-
"lstrip": false,
|
879 |
-
"normalized": true,
|
880 |
-
"rstrip": false,
|
881 |
-
"single_word": false,
|
882 |
-
"special": false
|
883 |
-
},
|
884 |
-
"32107": {
|
885 |
-
"content": "ớ",
|
886 |
-
"lstrip": false,
|
887 |
-
"normalized": true,
|
888 |
-
"rstrip": false,
|
889 |
-
"single_word": false,
|
890 |
-
"special": false
|
891 |
-
},
|
892 |
-
"32108": {
|
893 |
-
"content": "ờ",
|
894 |
-
"lstrip": false,
|
895 |
-
"normalized": true,
|
896 |
-
"rstrip": false,
|
897 |
-
"single_word": false,
|
898 |
-
"special": false
|
899 |
-
},
|
900 |
-
"32109": {
|
901 |
-
"content": "ỹ",
|
902 |
-
"lstrip": false,
|
903 |
-
"normalized": true,
|
904 |
-
"rstrip": false,
|
905 |
-
"single_word": false,
|
906 |
-
"special": false
|
907 |
-
},
|
908 |
-
"32110": {
|
909 |
-
"content": "ạ",
|
910 |
-
"lstrip": false,
|
911 |
-
"normalized": true,
|
912 |
-
"rstrip": false,
|
913 |
-
"single_word": false,
|
914 |
-
"special": false
|
915 |
-
},
|
916 |
-
"32111": {
|
917 |
-
"content": "ề",
|
918 |
-
"lstrip": false,
|
919 |
-
"normalized": true,
|
920 |
-
"rstrip": false,
|
921 |
-
"single_word": false,
|
922 |
-
"special": false
|
923 |
-
},
|
924 |
-
"32112": {
|
925 |
-
"content": "ủ",
|
926 |
-
"lstrip": false,
|
927 |
-
"normalized": true,
|
928 |
-
"rstrip": false,
|
929 |
-
"single_word": false,
|
930 |
-
"special": false
|
931 |
-
},
|
932 |
-
"32113": {
|
933 |
-
"content": "ì",
|
934 |
-
"lstrip": false,
|
935 |
-
"normalized": true,
|
936 |
-
"rstrip": false,
|
937 |
-
"single_word": false,
|
938 |
-
"special": false
|
939 |
-
},
|
940 |
-
"32114": {
|
941 |
-
"content": "ả",
|
942 |
-
"lstrip": false,
|
943 |
-
"normalized": true,
|
944 |
-
"rstrip": false,
|
945 |
-
"single_word": false,
|
946 |
-
"special": false
|
947 |
-
},
|
948 |
-
"32115": {
|
949 |
-
"content": "ú",
|
950 |
-
"lstrip": false,
|
951 |
-
"normalized": true,
|
952 |
-
"rstrip": false,
|
953 |
-
"single_word": false,
|
954 |
-
"special": false
|
955 |
-
},
|
956 |
-
"32116": {
|
957 |
-
"content": "ố",
|
958 |
-
"lstrip": false,
|
959 |
-
"normalized": true,
|
960 |
-
"rstrip": false,
|
961 |
-
"single_word": false,
|
962 |
-
"special": false
|
963 |
-
},
|
964 |
-
"32117": {
|
965 |
-
"content": "ầ",
|
966 |
-
"lstrip": false,
|
967 |
-
"normalized": true,
|
968 |
-
"rstrip": false,
|
969 |
-
"single_word": false,
|
970 |
-
"special": false
|
971 |
-
},
|
972 |
-
"32118": {
|
973 |
-
"content": "ự",
|
974 |
-
"lstrip": false,
|
975 |
-
"normalized": true,
|
976 |
-
"rstrip": false,
|
977 |
-
"single_word": false,
|
978 |
-
"special": false
|
979 |
-
},
|
980 |
-
"32119": {
|
981 |
-
"content": "ễ",
|
982 |
-
"lstrip": false,
|
983 |
-
"normalized": true,
|
984 |
-
"rstrip": false,
|
985 |
-
"single_word": false,
|
986 |
-
"special": false
|
987 |
-
},
|
988 |
-
"32120": {
|
989 |
-
"content": "í",
|
990 |
-
"lstrip": false,
|
991 |
-
"normalized": true,
|
992 |
-
"rstrip": false,
|
993 |
-
"single_word": false,
|
994 |
-
"special": false
|
995 |
-
},
|
996 |
-
"32121": {
|
997 |
-
"content": "ý",
|
998 |
-
"lstrip": false,
|
999 |
-
"normalized": true,
|
1000 |
-
"rstrip": false,
|
1001 |
-
"single_word": false,
|
1002 |
-
"special": false
|
1003 |
-
},
|
1004 |
-
"32122": {
|
1005 |
-
"content": "ệ",
|
1006 |
-
"lstrip": false,
|
1007 |
-
"normalized": true,
|
1008 |
-
"rstrip": false,
|
1009 |
-
"single_word": false,
|
1010 |
-
"special": false
|
1011 |
}
|
1012 |
},
|
1013 |
"additional_special_tokens": [
|
|
|
824 |
"rstrip": false,
|
825 |
"single_word": false,
|
826 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
827 |
}
|
828 |
},
|
829 |
"additional_special_tokens": [
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5048
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:03aedf79907c5fc61fc33dbafa5e636693826a8fb5a693a1ac955ebdfdc44d9a
|
3 |
size 5048
|