quancute commited on
Commit
61ba82d
·
verified ·
1 Parent(s): 0ecf523

Training in progress, epoch 1

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
config.json ADDED
@@ -0,0 +1,270 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "FacebookAI/xlm-roberta-large",
3
+ "architectures": [
4
+ "XLMRobertaForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "O",
15
+ "1": "RAM",
16
+ "2": "M\u1ea1ng kh\u00f4ng d\u00e2y",
17
+ "3": "V\u00f9ng c\u01a1 th\u1ec3",
18
+ "4": "\u0110\u1ed9 ph\u00e2n gi\u1ea3i",
19
+ "5": "Tr\u1ecdng l\u01b0\u1ee3ng gi\u1ea5y",
20
+ "6": "C\u1ed5 gi\u00e0y",
21
+ "7": "T\u1ed1c \u0111\u1ed9 x\u1eed l\u00fd",
22
+ "8": "\u0110\u1ed9 d\u00e0y",
23
+ "9": "Dung l\u01b0\u1ee3ng",
24
+ "10": "M\u00f9a",
25
+ "11": "K\u00edch th\u01b0\u1edbc m\u00e0n h\u00ecnh",
26
+ "12": "H\u1ec7 \u0111i\u1ec1u h\u00e0nh",
27
+ "13": "Ki\u1ec3u th\u1ea7n t\u01b0\u1ee3ng",
28
+ "14": "Lo\u1ea1i Kh\u00f3a",
29
+ "15": "M\u1ef1c v\u00e0 t\u00f4ng m\u00e0u",
30
+ "16": "C\u00f4ng su\u1ea5t",
31
+ "17": "B\u1ed9 x\u1eed l\u00fd",
32
+ "18": "Lo\u1ea1i m\u0169i gi\u00e0y",
33
+ "19": "Lo\u1ea1i h\u00ecnh th\u1ec3 thao",
34
+ "20": "Ch\u1ebf \u0111\u1ed9 \u0103n u\u1ed1ng",
35
+ "21": "D\u1ecbp",
36
+ "22": "Chi\u1ec1u r\u1ed9ng",
37
+ "23": "Lo\u1ea1i \u0110\u1ed3 L\u00f3t",
38
+ "24": "K\u00edch th\u01b0\u1edbc gi\u1ea5y",
39
+ "25": "Lo\u1ea1i \u0111\u0129a CD, DVD & Bluray",
40
+ "26": "H\u1ea1n s\u1eed d\u1ee5ng",
41
+ "27": "N\u00e9t b\u00fat",
42
+ "28": "M\u00e0n h\u00ecnh",
43
+ "29": "M\u1eb7t \u0111\u1ed3ng h\u1ed3",
44
+ "30": "Lo\u1ea1i k\u1ebft n\u1ed1i",
45
+ "31": "Ki\u1ec3u khuy",
46
+ "32": "Ki\u1ec3u k\u1ebft n\u1ed1i",
47
+ "33": "Handmade",
48
+ "34": "Ki\u1ec3u d\u00e1ng v\u00e1y",
49
+ "35": "Ch\u1ec9 s\u1ed1 ch\u1ed1ng n\u1eafng",
50
+ "36": "Ch\u1ea5t li\u1ec7u d\u00e2y \u0111eo",
51
+ "37": "H\u1ea1n b\u1ea3o h\u00e0nh",
52
+ "38": "\u0110\u01b0\u1eddng k\u00ednh",
53
+ "39": "K\u1ebft c\u1ea5u gi\u1ea5y",
54
+ "40": "\u0110\u1ed9 tu\u1ed5i khuy\u1ebfn ngh\u1ecb",
55
+ "41": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u00e2m nh\u1ea1c",
56
+ "42": "Ki\u1ec3u Bra",
57
+ "43": "Lo\u1ea1i da",
58
+ "44": "D\u00e1ng ki\u1ec3u \u00e1o",
59
+ "45": "Lo\u1ea1i m\u00f4n h\u1ecdc",
60
+ "46": "Tay \u00e1o",
61
+ "47": "Lo\u1ea1i \u0111\u1ebf gi\u00e0y",
62
+ "48": "Lo\u1ea1i phi\u00ean b\u1ea3n",
63
+ "49": "Ph\u01b0\u01a1ng th\u1ee9c ho\u1ea1t \u0111\u1ed9ng",
64
+ "50": "Lo\u1ea1i thi\u1ebft b\u1ecb an to\u00e0n",
65
+ "51": "Lo\u1ea1i \u0111\u00e8n",
66
+ "52": "Lo\u1ea1i \u0111\u1ed3 ch\u0103m s\u00f3c s\u1ee9c kh\u1ecfe m\u1eb9",
67
+ "53": "Lo\u1ea1i hoa",
68
+ "54": "Nh\u00f3m tu\u1ed5i",
69
+ "55": "Lo\u1ea1i trang thi\u1ebft b\u1ecb th\u1ec3 thao",
70
+ "56": "Lo\u1ea1i v\u1eadt d\u1ee5ng h\u1ed3 c\u00e1",
71
+ "57": "Lo\u1ea1i voucher & d\u1ecbch v\u1ee5",
72
+ "58": "Chi\u1ec1u cao",
73
+ "59": "Lo\u1ea1i tr\u00f2 ch\u01a1i",
74
+ "60": "Lo\u1ea1i thi\u1ebft b\u1ecb \u00e2m thanh",
75
+ "61": "Lo\u1ea1i b\u1ed9 \u0111\u1ed3 b\u00e0n \u0103n",
76
+ "62": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng kh\u00e1ch",
77
+ "63": "Lo\u1ea1i thi\u1ebft b\u1ecb \u0111i\u1ec7n gia d\u1ee5ng",
78
+ "64": "Chi\u1ec1u d\u00e0i",
79
+ "65": "Lo\u1ea1i gi\u00e0y d\u00e9p",
80
+ "66": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng t\u1eafm",
81
+ "67": "Th\u1ec3 t\u00edch",
82
+ "68": "Lo\u1ea1i thi\u1ebft b\u1ecb y t\u1ebf",
83
+ "69": "Gi\u1edbi t\u00ednh",
84
+ "70": "Phong c\u00e1ch",
85
+ "71": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u0111i\u1ec7n t\u1eed",
86
+ "72": "Lo\u1ea1i \u0111\u1ed3 v\u1eadt s\u01b0u t\u1ea7m",
87
+ "73": "M\u00f9i h\u01b0\u01a1ng",
88
+ "74": "Ch\u1ee9c n\u0103ng h\u1ed7 tr\u1ee3 s\u1ee9c kh\u1ecfe",
89
+ "75": "T\u00ecnh tr\u1ea1ng",
90
+ "76": "C\u1ed5 \u00e1o",
91
+ "77": "C\u00e2n n\u1eb7ng",
92
+ "78": "M\u1eabu xe t\u01b0\u01a1ng th\u00edch",
93
+ "79": "Lo\u1ea1i \u0111\u1ed3 gia d\u1ee5ng b\u1ebfp",
94
+ "80": "Lo\u1ea1i \u1ed1p",
95
+ "81": "H\u01b0\u01a1ng v\u1ecb",
96
+ "82": "Lo\u1ea1i tr\u00e1i c\u00e2y",
97
+ "83": "Th\u00e0nh Ph\u1ea7n Ho\u1ea1t T\u00ednh",
98
+ "84": "Lo\u1ea1i \u0111\u1ed3 ch\u01a1i",
99
+ "85": "Dung t\u00edch",
100
+ "86": "Xu\u1ea5t x\u1ee9",
101
+ "87": "T\u00edch h\u1ee3p ph\u1ee5 ki\u1ec7n",
102
+ "88": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng ng\u1ee7",
103
+ "89": "Lo\u1ea1i d\u1ee5ng c\u1ee5 & thi\u1ebft b\u1ecb s\u1eeda ch\u1eefa",
104
+ "90": "Lo\u1ea1i v\u1eadt d\u1ee5ng l\u00e0m v\u01b0\u1eddn",
105
+ "91": "Ch\u1ee9c n\u0103ng l\u00e0m \u0111\u1eb9p",
106
+ "92": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ch\u0103m s\u00f3c b\u00e9",
107
+ "93": "Tr\u1ecdng l\u01b0\u1ee3ng",
108
+ "94": "Thi\u1ebft b\u1ecb t\u01b0\u01a1ng th\u00edch",
109
+ "95": "H\u00ecnh th\u1ee9c s\u1ea3n ph\u1ea9m",
110
+ "96": "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 d\u1ee5ng c\u1ee5 l\u00e0m \u0111\u1eb9p",
111
+ "97": "Lo\u1ea1i trang ph\u1ee5c",
112
+ "98": "Th\u00e0nh ph\u1ea7n",
113
+ "99": "Ch\u1ea5t li\u1ec7u",
114
+ "100": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng v\u0103n ph\u00f2ng ph\u1ea9m",
115
+ "101": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n nh\u00e0 b\u1ebfp",
116
+ "102": "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 linh ki\u1ec7n \u0111i\u1ec7n t\u1eed",
117
+ "103": "Lo\u1ea1i trang s\u1ee9c / ph\u1ee5 ki\u1ec7n",
118
+ "104": "Size",
119
+ "105": "Lo\u1ea1i th\u1ee9c \u0103n ch\u1ebf bi\u1ebfn s\u1eb5n",
120
+ "106": "Th\u1ec3 lo\u1ea1i s\u00e1ch truy\u1ec7n",
121
+ "107": "Lo\u1ea1i v\u1eadt d\u1ee5ng ch\u0103m s\u00f3c c\u00e1 nh\u00e2n",
122
+ "108": "T\u00ednh n\u0103ng",
123
+ "109": "Th\u01b0\u01a1ng hi\u1ec7u",
124
+ "110": "Ki\u1ec3u \u0111\u00f3ng g\u00f3i",
125
+ "111": "Lo\u1ea1i thu\u1ed1c & th\u1ef1c ph\u1ea9m ch\u1ee9c n\u0103ng",
126
+ "112": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n / linh ki\u1ec7n xe",
127
+ "113": "Lo\u1ea1i v\u1eadt d\u1ee5ng trong nh\u00e0 kh\u00e1c",
128
+ "114": "M\u00e0u s\u1eafc",
129
+ "115": "Lo\u1ea1i th\u1ef1c ph\u1ea9m",
130
+ "116": "Lo\u1ea1i gi\u1ed1ng",
131
+ "117": "M\u1eabu",
132
+ "118": "Lo\u1ea1i s\u1ea3n ph\u1ea9m l\u00e0m \u0111\u1eb9p"
133
+ },
134
+ "initializer_range": 0.02,
135
+ "intermediate_size": 4096,
136
+ "label2id": {
137
+ "B\u1ed9 x\u1eed l\u00fd": 17,
138
+ "Chi\u1ec1u cao": 58,
139
+ "Chi\u1ec1u d\u00e0i": 64,
140
+ "Chi\u1ec1u r\u1ed9ng": 22,
141
+ "Ch\u1ea5t li\u1ec7u": 99,
142
+ "Ch\u1ea5t li\u1ec7u d\u00e2y \u0111eo": 36,
143
+ "Ch\u1ebf \u0111\u1ed9 \u0103n u\u1ed1ng": 20,
144
+ "Ch\u1ec9 s\u1ed1 ch\u1ed1ng n\u1eafng": 35,
145
+ "Ch\u1ee9c n\u0103ng h\u1ed7 tr\u1ee3 s\u1ee9c kh\u1ecfe": 74,
146
+ "Ch\u1ee9c n\u0103ng l\u00e0m \u0111\u1eb9p": 91,
147
+ "C\u00e2n n\u1eb7ng": 77,
148
+ "C\u00f4ng su\u1ea5t": 16,
149
+ "C\u1ed5 gi\u00e0y": 6,
150
+ "C\u1ed5 \u00e1o": 76,
151
+ "Dung l\u01b0\u1ee3ng": 9,
152
+ "Dung t\u00edch": 85,
153
+ "D\u00e1ng ki\u1ec3u \u00e1o": 44,
154
+ "D\u1ecbp": 21,
155
+ "Gi\u1edbi t\u00ednh": 69,
156
+ "Handmade": 33,
157
+ "H\u00ecnh th\u1ee9c s\u1ea3n ph\u1ea9m": 95,
158
+ "H\u01b0\u01a1ng v\u1ecb": 81,
159
+ "H\u1ea1n b\u1ea3o h\u00e0nh": 37,
160
+ "H\u1ea1n s\u1eed d\u1ee5ng": 26,
161
+ "H\u1ec7 \u0111i\u1ec1u h\u00e0nh": 12,
162
+ "Ki\u1ec3u Bra": 42,
163
+ "Ki\u1ec3u d\u00e1ng v\u00e1y": 34,
164
+ "Ki\u1ec3u khuy": 31,
165
+ "Ki\u1ec3u k\u1ebft n\u1ed1i": 32,
166
+ "Ki\u1ec3u th\u1ea7n t\u01b0\u1ee3ng": 13,
167
+ "Ki\u1ec3u \u0111\u00f3ng g\u00f3i": 110,
168
+ "K\u00edch th\u01b0\u1edbc gi\u1ea5y": 24,
169
+ "K\u00edch th\u01b0\u1edbc m\u00e0n h\u00ecnh": 11,
170
+ "K\u1ebft c\u1ea5u gi\u1ea5y": 39,
171
+ "Lo\u1ea1i Kh\u00f3a": 14,
172
+ "Lo\u1ea1i b\u1ed9 \u0111\u1ed3 b\u00e0n \u0103n": 61,
173
+ "Lo\u1ea1i da": 43,
174
+ "Lo\u1ea1i d\u1ee5ng c\u1ee5 & thi\u1ebft b\u1ecb s\u1eeda ch\u1eefa": 89,
175
+ "Lo\u1ea1i gi\u00e0y d\u00e9p": 65,
176
+ "Lo\u1ea1i gi\u1ed1ng": 116,
177
+ "Lo\u1ea1i hoa": 53,
178
+ "Lo\u1ea1i h\u00ecnh th\u1ec3 thao": 19,
179
+ "Lo\u1ea1i k\u1ebft n\u1ed1i": 30,
180
+ "Lo\u1ea1i m\u00f4n h\u1ecdc": 45,
181
+ "Lo\u1ea1i m\u0169i gi\u00e0y": 18,
182
+ "Lo\u1ea1i phi\u00ean b\u1ea3n": 48,
183
+ "Lo\u1ea1i ph\u1ee5 ki\u1ec7n / linh ki\u1ec7n xe": 112,
184
+ "Lo\u1ea1i ph\u1ee5 ki\u1ec7n nh\u00e0 b\u1ebfp": 101,
185
+ "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u00e2m nh\u1ea1c": 41,
186
+ "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u0111i\u1ec7n t\u1eed": 71,
187
+ "Lo\u1ea1i s\u1ea3n ph\u1ea9m l\u00e0m \u0111\u1eb9p": 118,
188
+ "Lo\u1ea1i thi\u1ebft b\u1ecb an to\u00e0n": 50,
189
+ "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 d\u1ee5ng c\u1ee5 l\u00e0m \u0111\u1eb9p": 96,
190
+ "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 linh ki\u1ec7n \u0111i\u1ec7n t\u1eed": 102,
191
+ "Lo\u1ea1i thi\u1ebft b\u1ecb y t\u1ebf": 68,
192
+ "Lo\u1ea1i thi\u1ebft b\u1ecb \u00e2m thanh": 60,
193
+ "Lo\u1ea1i thi\u1ebft b\u1ecb \u0111i\u1ec7n gia d\u1ee5ng": 63,
194
+ "Lo\u1ea1i thu\u1ed1c & th\u1ef1c ph\u1ea9m ch\u1ee9c n\u0103ng": 111,
195
+ "Lo\u1ea1i th\u1ee9c \u0103n ch\u1ebf bi\u1ebfn s\u1eb5n": 105,
196
+ "Lo\u1ea1i th\u1ef1c ph\u1ea9m": 115,
197
+ "Lo\u1ea1i trang ph\u1ee5c": 97,
198
+ "Lo\u1ea1i trang s\u1ee9c / ph\u1ee5 ki\u1ec7n": 103,
199
+ "Lo\u1ea1i trang thi\u1ebft b\u1ecb th\u1ec3 thao": 55,
200
+ "Lo\u1ea1i tr\u00e1i c\u00e2y": 82,
201
+ "Lo\u1ea1i tr\u00f2 ch\u01a1i": 59,
202
+ "Lo\u1ea1i voucher & d\u1ecbch v\u1ee5": 57,
203
+ "Lo\u1ea1i v\u1eadt d\u1ee5ng ch\u0103m s\u00f3c c\u00e1 nh\u00e2n": 107,
204
+ "Lo\u1ea1i v\u1eadt d\u1ee5ng h\u1ed3 c\u00e1": 56,
205
+ "Lo\u1ea1i v\u1eadt d\u1ee5ng l\u00e0m v\u01b0\u1eddn": 90,
206
+ "Lo\u1ea1i v\u1eadt d\u1ee5ng trong nh\u00e0 kh\u00e1c": 113,
207
+ "Lo\u1ea1i \u0110\u1ed3 L\u00f3t": 23,
208
+ "Lo\u1ea1i \u0111\u00e8n": 51,
209
+ "Lo\u1ea1i \u0111\u0129a CD, DVD & Bluray": 25,
210
+ "Lo\u1ea1i \u0111\u1ebf gi\u00e0y": 47,
211
+ "Lo\u1ea1i \u0111\u1ed3 ch\u0103m s\u00f3c s\u1ee9c kh\u1ecfe m\u1eb9": 52,
212
+ "Lo\u1ea1i \u0111\u1ed3 ch\u01a1i": 84,
213
+ "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ch\u0103m s\u00f3c b\u00e9": 92,
214
+ "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng kh\u00e1ch": 62,
215
+ "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng ng\u1ee7": 88,
216
+ "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng t\u1eafm": 66,
217
+ "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng v\u0103n ph\u00f2ng ph\u1ea9m": 100,
218
+ "Lo\u1ea1i \u0111\u1ed3 gia d\u1ee5ng b\u1ebfp": 79,
219
+ "Lo\u1ea1i \u0111\u1ed3 v\u1eadt s\u01b0u t\u1ea7m": 72,
220
+ "Lo\u1ea1i \u1ed1p": 80,
221
+ "M\u00e0n h\u00ecnh": 28,
222
+ "M\u00e0u s\u1eafc": 114,
223
+ "M\u00f9a": 10,
224
+ "M\u00f9i h\u01b0\u01a1ng": 73,
225
+ "M\u1ea1ng kh\u00f4ng d\u00e2y": 2,
226
+ "M\u1eabu": 117,
227
+ "M\u1eabu xe t\u01b0\u01a1ng th\u00edch": 78,
228
+ "M\u1eb7t \u0111\u1ed3ng h\u1ed3": 29,
229
+ "M\u1ef1c v\u00e0 t\u00f4ng m\u00e0u": 15,
230
+ "Nh\u00f3m tu\u1ed5i": 54,
231
+ "N\u00e9t b\u00fat": 27,
232
+ "O": 0,
233
+ "Phong c\u00e1ch": 70,
234
+ "Ph\u01b0\u01a1ng th\u1ee9c ho\u1ea1t \u0111\u1ed9ng": 49,
235
+ "RAM": 1,
236
+ "Size": 104,
237
+ "Tay \u00e1o": 46,
238
+ "Thi\u1ebft b\u1ecb t\u01b0\u01a1ng th\u00edch": 94,
239
+ "Th\u00e0nh Ph\u1ea7n Ho\u1ea1t T\u00ednh": 83,
240
+ "Th\u00e0nh ph\u1ea7n": 98,
241
+ "Th\u01b0\u01a1ng hi\u1ec7u": 109,
242
+ "Th\u1ec3 lo\u1ea1i s\u00e1ch truy\u1ec7n": 106,
243
+ "Th\u1ec3 t\u00edch": 67,
244
+ "Tr\u1ecdng l\u01b0\u1ee3ng": 93,
245
+ "Tr\u1ecdng l\u01b0\u1ee3ng gi\u1ea5y": 5,
246
+ "T\u00ecnh tr\u1ea1ng": 75,
247
+ "T\u00edch h\u1ee3p ph\u1ee5 ki\u1ec7n": 87,
248
+ "T\u00ednh n\u0103ng": 108,
249
+ "T\u1ed1c \u0111\u1ed9 x\u1eed l\u00fd": 7,
250
+ "V\u00f9ng c\u01a1 th\u1ec3": 3,
251
+ "Xu\u1ea5t x\u1ee9": 86,
252
+ "\u0110\u01b0\u1eddng k\u00ednh": 38,
253
+ "\u0110\u1ed9 d\u00e0y": 8,
254
+ "\u0110\u1ed9 ph\u00e2n gi\u1ea3i": 4,
255
+ "\u0110\u1ed9 tu\u1ed5i khuy\u1ebfn ngh\u1ecb": 40
256
+ },
257
+ "layer_norm_eps": 1e-05,
258
+ "max_position_embeddings": 514,
259
+ "model_type": "xlm-roberta",
260
+ "num_attention_heads": 16,
261
+ "num_hidden_layers": 24,
262
+ "output_past": true,
263
+ "pad_token_id": 1,
264
+ "position_embedding_type": "absolute",
265
+ "torch_dtype": "float32",
266
+ "transformers_version": "4.44.2",
267
+ "type_vocab_size": 1,
268
+ "use_cache": true,
269
+ "vocab_size": 250002
270
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f18ac4862381ecb12062e8ce67fc5ac2f99f068118b114bdb57a298e32e9ca8a
3
+ size 2235899764
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8373f9cd3d27591e1924426bcc1c8799bc5a9affc4fc857982c5d66668dd1f41
3
+ size 17082832
tokenizer_config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "mask_token": "<mask>",
49
+ "model_max_length": 512,
50
+ "pad_token": "<pad>",
51
+ "sep_token": "</s>",
52
+ "tokenizer_class": "XLMRobertaTokenizer",
53
+ "unk_token": "<unk>"
54
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:575dbd9c3ce56b9adb9eaaa21cbc6676cbbb0f342776956eccee51f1de66f75f
3
+ size 5240