talos-10m / tokenizer.json
PrometheanStudio's picture
Release Talos 10M — 100k steps
d2d10c8 verified
Raw History Blame Contribute Delete
4.93 kB
{
"version": 1,
"type": "ByteLevelBPE",
"config": {
"vocab_size": 512,
"bos_token": "<|beginoftext|>",
"eos_token": "<|endoftext|>",
"pad_token": "<|pad|>",
"unk_token": "<|unk|>",
"extra_special_tokens": [],
"reserved_tokens": [],
"pre_tokenize": null,
"split_special": false
},
"merges": [
[
100,
101
],
[
104,
105
],
[
256,
102
],
[
116,
117
],
[
114,
115
],
[
111,
112
],
[
97,
98
],
[
262,
99
],
[
263,
258
],
[
264,
103
],
[
265,
257
],
[
266,
106
],
[
267,
107
],
[
268,
108
],
[
269,
109
],
[
270,
110
],
[
271,
261
],
[
272,
113
],
[
273,
260
],
[
274,
259
],
[
275,
118
],
[
276,
119
],
[
277,
120
],
[
278,
121
],
[
279,
122
],
[
280,
280
],
[
281,
281
],
[
282,
282
],
[
283,
283
],
[
224,
164
],
[
101,
32
],
[
284,
284
],
[
105,
110
],
[
110,
32
],
[
32,
226
],
[
114,
101
],
[
32,
32
],
[
224,
165
],
[
217,
132
],
[
102,
111
],
[
101,
114
],
[
101,
110
],
[
32,
98
],
[
297,
116
],
[
290,
136
],
[
208,
184
],
[
116,
104
],
[
115,
32
],
[
114,
97
],
[
108,
32
],
[
103,
32
],
[
95,
95
],
[
302,
286
],
[
296,
32
],
[
294,
216
],
[
290,
137
],
[
287,
287
],
[
217,
138
],
[
216,
169
],
[
314,
32
],
[
216,
167
],
[
208,
187
],
[
121,
116
],
[
121,
32
],
[
120,
32
],
[
117,
105
],
[
116,
304
],
[
115,
111
],
[
113,
321
],
[
111,
289
],
[
110,
101
],
[
109,
111
],
[
108,
97
],
[
105,
325
],
[
97,
108
],
[
32,
194
],
[
328,
122
],
[
332,
319
],
[
333,
100
],
[
334,
111
],
[
335,
306
],
[
327,
256
],
[
337,
305
],
[
324,
99
],
[
339,
107
],
[
340,
298
],
[
341,
114
],
[
342,
111
],
[
343,
119
],
[
344,
289
],
[
345,
295
],
[
346,
320
],
[
347,
106
],
[
348,
117
],
[
349,
109
],
[
350,
112
],
[
351,
303
],
[
352,
111
],
[
353,
118
],
[
354,
309
],
[
322,
288
],
[
318,
101
],
[
316,
310
],
[
358,
185
],
[
359,
216
],
[
360,
177
],
[
361,
216
],
[
362,
168
],
[
363,
313
],
[
364,
315
],
[
303,
330
],
[
301,
209
],
[
299,
329
],
[
293,
141
],
[
369,
285
],
[
292,
292
],
[
288,
306
],
[
285,
168
],
[
231,
187
],
[
209,
129
],
[
208,
186
],
[
178,
32
],
[
120,
41
],
[
116,
368
],
[
116,
379
],
[
115,
101
],
[
115,
99
],
[
114,
99
],
[
110,
100
],
[
109,
97
],
[
105,
366
],
[
103,
286
],
[
58,
10
],
[
388,
371
],
[
40,
378
],
[
32,
61
]
]
}