tokenizers 0.2.2-aarch64-linux
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/CHANGELOG.md +32 -0
- data/Cargo.lock +909 -0
- data/Cargo.toml +5 -0
- data/LICENSE-THIRD-PARTY.txt +17956 -0
- data/LICENSE.txt +202 -0
- data/README.md +65 -0
- data/lib/tokenizers/2.7/tokenizers.so +0 -0
- data/lib/tokenizers/3.0/tokenizers.so +0 -0
- data/lib/tokenizers/3.1/tokenizers.so +0 -0
- data/lib/tokenizers/3.2/tokenizers.so +0 -0
- data/lib/tokenizers/char_bpe_tokenizer.rb +19 -0
- data/lib/tokenizers/from_pretrained.rb +119 -0
- data/lib/tokenizers/version.rb +3 -0
- data/lib/tokenizers.rb +17 -0
- metadata +59 -0
data/Cargo.lock
ADDED
@@ -0,0 +1,909 @@
|
|
1
|
+
# This file is automatically @generated by Cargo.
|
2
|
+
# It is not intended for manual editing.
|
3
|
+
version = 3
|
4
|
+
|
5
|
+
[[package]]
|
6
|
+
name = "aho-corasick"
|
7
|
+
version = "0.7.20"
|
8
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
9
|
+
checksum = "cc936419f96fa211c1b9166887b38e5e40b19958e5b895be7c1f93adec7071ac"
|
10
|
+
dependencies = [
|
11
|
+
"memchr",
|
12
|
+
]
|
13
|
+
|
14
|
+
[[package]]
|
15
|
+
name = "autocfg"
|
16
|
+
version = "1.1.0"
|
17
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
18
|
+
checksum = "d468802bab17cbc0cc575e9b053f41e72aa36bfa6b7f55e3529ffa43161b97fa"
|
19
|
+
|
20
|
+
[[package]]
|
21
|
+
name = "base64"
|
22
|
+
version = "0.13.1"
|
23
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
24
|
+
checksum = "9e1b586273c5702936fe7b7d6896644d8be71e6314cfe09d3167c95f712589e8"
|
25
|
+
|
26
|
+
[[package]]
|
27
|
+
name = "bindgen"
|
28
|
+
version = "0.60.1"
|
29
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
30
|
+
checksum = "062dddbc1ba4aca46de6338e2bf87771414c335f7b2f2036e8f3e9befebf88e6"
|
31
|
+
dependencies = [
|
32
|
+
"bitflags",
|
33
|
+
"cexpr",
|
34
|
+
"clang-sys",
|
35
|
+
"lazy_static",
|
36
|
+
"lazycell",
|
37
|
+
"peeking_take_while",
|
38
|
+
"proc-macro2",
|
39
|
+
"quote",
|
40
|
+
"regex",
|
41
|
+
"rustc-hash",
|
42
|
+
"shlex",
|
43
|
+
]
|
44
|
+
|
45
|
+
[[package]]
|
46
|
+
name = "bitflags"
|
47
|
+
version = "1.3.2"
|
48
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
49
|
+
checksum = "bef38d45163c2f1dde094a7dfd33ccf595c92905c8f8f4fdc18d06fb1037718a"
|
50
|
+
|
51
|
+
[[package]]
|
52
|
+
name = "cc"
|
53
|
+
version = "1.0.78"
|
54
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
55
|
+
checksum = "a20104e2335ce8a659d6dd92a51a767a0c062599c73b343fd152cb401e828c3d"
|
56
|
+
|
57
|
+
[[package]]
|
58
|
+
name = "cexpr"
|
59
|
+
version = "0.6.0"
|
60
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
61
|
+
checksum = "6fac387a98bb7c37292057cffc56d62ecb629900026402633ae9160df93a8766"
|
62
|
+
dependencies = [
|
63
|
+
"nom",
|
64
|
+
]
|
65
|
+
|
66
|
+
[[package]]
|
67
|
+
name = "cfg-if"
|
68
|
+
version = "1.0.0"
|
69
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
70
|
+
checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
|
71
|
+
|
72
|
+
[[package]]
|
73
|
+
name = "clang-sys"
|
74
|
+
version = "1.4.0"
|
75
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
76
|
+
checksum = "fa2e27ae6ab525c3d369ded447057bca5438d86dc3a68f6faafb8269ba82ebf3"
|
77
|
+
dependencies = [
|
78
|
+
"glob",
|
79
|
+
"libc",
|
80
|
+
"libloading",
|
81
|
+
]
|
82
|
+
|
83
|
+
[[package]]
|
84
|
+
name = "console"
|
85
|
+
version = "0.15.5"
|
86
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
87
|
+
checksum = "c3d79fbe8970a77e3e34151cc13d3b3e248aa0faaecb9f6091fa07ebefe5ad60"
|
88
|
+
dependencies = [
|
89
|
+
"encode_unicode",
|
90
|
+
"lazy_static",
|
91
|
+
"libc",
|
92
|
+
"unicode-width",
|
93
|
+
"windows-sys",
|
94
|
+
]
|
95
|
+
|
96
|
+
[[package]]
|
97
|
+
name = "crossbeam-channel"
|
98
|
+
version = "0.5.6"
|
99
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
100
|
+
checksum = "c2dd04ddaf88237dc3b8d8f9a3c1004b506b54b3313403944054d23c0870c521"
|
101
|
+
dependencies = [
|
102
|
+
"cfg-if",
|
103
|
+
"crossbeam-utils",
|
104
|
+
]
|
105
|
+
|
106
|
+
[[package]]
|
107
|
+
name = "crossbeam-deque"
|
108
|
+
version = "0.8.2"
|
109
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
110
|
+
checksum = "715e8152b692bba2d374b53d4875445368fdf21a94751410af607a5ac677d1fc"
|
111
|
+
dependencies = [
|
112
|
+
"cfg-if",
|
113
|
+
"crossbeam-epoch",
|
114
|
+
"crossbeam-utils",
|
115
|
+
]
|
116
|
+
|
117
|
+
[[package]]
|
118
|
+
name = "crossbeam-epoch"
|
119
|
+
version = "0.9.13"
|
120
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
121
|
+
checksum = "01a9af1f4c2ef74bb8aa1f7e19706bc72d03598c8a570bb5de72243c7a9d9d5a"
|
122
|
+
dependencies = [
|
123
|
+
"autocfg",
|
124
|
+
"cfg-if",
|
125
|
+
"crossbeam-utils",
|
126
|
+
"memoffset",
|
127
|
+
"scopeguard",
|
128
|
+
]
|
129
|
+
|
130
|
+
[[package]]
|
131
|
+
name = "crossbeam-utils"
|
132
|
+
version = "0.8.14"
|
133
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
134
|
+
checksum = "4fb766fa798726286dbbb842f174001dab8abc7b627a1dd86e0b7222a95d929f"
|
135
|
+
dependencies = [
|
136
|
+
"cfg-if",
|
137
|
+
]
|
138
|
+
|
139
|
+
[[package]]
|
140
|
+
name = "darling"
|
141
|
+
version = "0.14.2"
|
142
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
143
|
+
checksum = "b0dd3cd20dc6b5a876612a6e5accfe7f3dd883db6d07acfbf14c128f61550dfa"
|
144
|
+
dependencies = [
|
145
|
+
"darling_core",
|
146
|
+
"darling_macro",
|
147
|
+
]
|
148
|
+
|
149
|
+
[[package]]
|
150
|
+
name = "darling_core"
|
151
|
+
version = "0.14.2"
|
152
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
153
|
+
checksum = "a784d2ccaf7c98501746bf0be29b2022ba41fd62a2e622af997a03e9f972859f"
|
154
|
+
dependencies = [
|
155
|
+
"fnv",
|
156
|
+
"ident_case",
|
157
|
+
"proc-macro2",
|
158
|
+
"quote",
|
159
|
+
"strsim",
|
160
|
+
"syn",
|
161
|
+
]
|
162
|
+
|
163
|
+
[[package]]
|
164
|
+
name = "darling_macro"
|
165
|
+
version = "0.14.2"
|
166
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
167
|
+
checksum = "7618812407e9402654622dd402b0a89dff9ba93badd6540781526117b92aab7e"
|
168
|
+
dependencies = [
|
169
|
+
"darling_core",
|
170
|
+
"quote",
|
171
|
+
"syn",
|
172
|
+
]
|
173
|
+
|
174
|
+
[[package]]
|
175
|
+
name = "derive_builder"
|
176
|
+
version = "0.12.0"
|
177
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
178
|
+
checksum = "8d67778784b508018359cbc8696edb3db78160bab2c2a28ba7f56ef6932997f8"
|
179
|
+
dependencies = [
|
180
|
+
"derive_builder_macro",
|
181
|
+
]
|
182
|
+
|
183
|
+
[[package]]
|
184
|
+
name = "derive_builder_core"
|
185
|
+
version = "0.12.0"
|
186
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
187
|
+
checksum = "c11bdc11a0c47bc7d37d582b5285da6849c96681023680b906673c5707af7b0f"
|
188
|
+
dependencies = [
|
189
|
+
"darling",
|
190
|
+
"proc-macro2",
|
191
|
+
"quote",
|
192
|
+
"syn",
|
193
|
+
]
|
194
|
+
|
195
|
+
[[package]]
|
196
|
+
name = "derive_builder_macro"
|
197
|
+
version = "0.12.0"
|
198
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
199
|
+
checksum = "ebcda35c7a396850a55ffeac740804b40ffec779b98fffbb1738f4033f0ee79e"
|
200
|
+
dependencies = [
|
201
|
+
"derive_builder_core",
|
202
|
+
"syn",
|
203
|
+
]
|
204
|
+
|
205
|
+
[[package]]
|
206
|
+
name = "dirs"
|
207
|
+
version = "3.0.2"
|
208
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
209
|
+
checksum = "30baa043103c9d0c2a57cf537cc2f35623889dc0d405e6c3cccfadbc81c71309"
|
210
|
+
dependencies = [
|
211
|
+
"dirs-sys",
|
212
|
+
]
|
213
|
+
|
214
|
+
[[package]]
|
215
|
+
name = "dirs-sys"
|
216
|
+
version = "0.3.7"
|
217
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
218
|
+
checksum = "1b1d1d91c932ef41c0f2663aa8b0ca0342d444d842c06914aa0a7e352d0bada6"
|
219
|
+
dependencies = [
|
220
|
+
"libc",
|
221
|
+
"redox_users",
|
222
|
+
"winapi",
|
223
|
+
]
|
224
|
+
|
225
|
+
[[package]]
|
226
|
+
name = "either"
|
227
|
+
version = "1.8.0"
|
228
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
229
|
+
checksum = "90e5c1c8368803113bf0c9584fc495a58b86dc8a29edbf8fe877d21d9507e797"
|
230
|
+
|
231
|
+
[[package]]
|
232
|
+
name = "encode_unicode"
|
233
|
+
version = "0.3.6"
|
234
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
235
|
+
checksum = "a357d28ed41a50f9c765dbfe56cbc04a64e53e5fc58ba79fbc34c10ef3df831f"
|
236
|
+
|
237
|
+
[[package]]
|
238
|
+
name = "esaxx-rs"
|
239
|
+
version = "0.1.8"
|
240
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
241
|
+
checksum = "1f748b253ceca9fed5f42f8b5ceb3851e93102199bc25b64b65369f76e5c0a35"
|
242
|
+
dependencies = [
|
243
|
+
"cc",
|
244
|
+
]
|
245
|
+
|
246
|
+
[[package]]
|
247
|
+
name = "fnv"
|
248
|
+
version = "1.0.7"
|
249
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
250
|
+
checksum = "3f9eec918d3f24069decb9af1554cad7c880e2da24a9afd88aca000531ab82c1"
|
251
|
+
|
252
|
+
[[package]]
|
253
|
+
name = "getrandom"
|
254
|
+
version = "0.2.8"
|
255
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
256
|
+
checksum = "c05aeb6a22b8f62540c194aac980f2115af067bfe15a0734d7277a768d396b31"
|
257
|
+
dependencies = [
|
258
|
+
"cfg-if",
|
259
|
+
"libc",
|
260
|
+
"wasi",
|
261
|
+
]
|
262
|
+
|
263
|
+
[[package]]
|
264
|
+
name = "glob"
|
265
|
+
version = "0.3.1"
|
266
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
267
|
+
checksum = "d2fabcfbdc87f4758337ca535fb41a6d701b65693ce38287d856d1674551ec9b"
|
268
|
+
|
269
|
+
[[package]]
|
270
|
+
name = "hermit-abi"
|
271
|
+
version = "0.2.6"
|
272
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
273
|
+
checksum = "ee512640fe35acbfb4bb779db6f0d80704c2cacfa2e39b601ef3e3f47d1ae4c7"
|
274
|
+
dependencies = [
|
275
|
+
"libc",
|
276
|
+
]
|
277
|
+
|
278
|
+
[[package]]
|
279
|
+
name = "ident_case"
|
280
|
+
version = "1.0.1"
|
281
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
282
|
+
checksum = "b9e0384b61958566e926dc50660321d12159025e767c18e043daf26b70104c39"
|
283
|
+
|
284
|
+
[[package]]
|
285
|
+
name = "indicatif"
|
286
|
+
version = "0.15.0"
|
287
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
288
|
+
checksum = "7baab56125e25686df467fe470785512329883aab42696d661247aca2a2896e4"
|
289
|
+
dependencies = [
|
290
|
+
"console",
|
291
|
+
"lazy_static",
|
292
|
+
"number_prefix",
|
293
|
+
"regex",
|
294
|
+
]
|
295
|
+
|
296
|
+
[[package]]
|
297
|
+
name = "itertools"
|
298
|
+
version = "0.8.2"
|
299
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
300
|
+
checksum = "f56a2d0bc861f9165be4eb3442afd3c236d8a98afd426f65d92324ae1091a484"
|
301
|
+
dependencies = [
|
302
|
+
"either",
|
303
|
+
]
|
304
|
+
|
305
|
+
[[package]]
|
306
|
+
name = "itertools"
|
307
|
+
version = "0.9.0"
|
308
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
309
|
+
checksum = "284f18f85651fe11e8a991b2adb42cb078325c996ed026d994719efcfca1d54b"
|
310
|
+
dependencies = [
|
311
|
+
"either",
|
312
|
+
]
|
313
|
+
|
314
|
+
[[package]]
|
315
|
+
name = "itoa"
|
316
|
+
version = "1.0.5"
|
317
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
318
|
+
checksum = "fad582f4b9e86b6caa621cabeb0963332d92eea04729ab12892c2533951e6440"
|
319
|
+
|
320
|
+
[[package]]
|
321
|
+
name = "lazy_static"
|
322
|
+
version = "1.4.0"
|
323
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
324
|
+
checksum = "e2abad23fbc42b3700f2f279844dc832adb2b2eb069b2df918f455c4e18cc646"
|
325
|
+
|
326
|
+
[[package]]
|
327
|
+
name = "lazycell"
|
328
|
+
version = "1.3.0"
|
329
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
330
|
+
checksum = "830d08ce1d1d941e6b30645f1a0eb5643013d835ce3779a5fc208261dbe10f55"
|
331
|
+
|
332
|
+
[[package]]
|
333
|
+
name = "libc"
|
334
|
+
version = "0.2.139"
|
335
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
336
|
+
checksum = "201de327520df007757c1f0adce6e827fe8562fbc28bfd9c15571c66ca1f5f79"
|
337
|
+
|
338
|
+
[[package]]
|
339
|
+
name = "libloading"
|
340
|
+
version = "0.7.4"
|
341
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
342
|
+
checksum = "b67380fd3b2fbe7527a606e18729d21c6f3951633d0500574c4dc22d2d638b9f"
|
343
|
+
dependencies = [
|
344
|
+
"cfg-if",
|
345
|
+
"winapi",
|
346
|
+
]
|
347
|
+
|
348
|
+
[[package]]
|
349
|
+
name = "log"
|
350
|
+
version = "0.4.17"
|
351
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
352
|
+
checksum = "abb12e687cfb44aa40f41fc3978ef76448f9b6038cad6aef4259d3c095a2382e"
|
353
|
+
dependencies = [
|
354
|
+
"cfg-if",
|
355
|
+
]
|
356
|
+
|
357
|
+
[[package]]
|
358
|
+
name = "macro_rules_attribute"
|
359
|
+
version = "0.1.3"
|
360
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
361
|
+
checksum = "cf0c9b980bf4f3a37fd7b1c066941dd1b1d0152ce6ee6e8fe8c49b9f6810d862"
|
362
|
+
dependencies = [
|
363
|
+
"macro_rules_attribute-proc_macro",
|
364
|
+
"paste",
|
365
|
+
]
|
366
|
+
|
367
|
+
[[package]]
|
368
|
+
name = "macro_rules_attribute-proc_macro"
|
369
|
+
version = "0.1.3"
|
370
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
371
|
+
checksum = "58093314a45e00c77d5c508f76e77c3396afbbc0d01506e7fae47b018bac2b1d"
|
372
|
+
|
373
|
+
[[package]]
|
374
|
+
name = "magnus"
|
375
|
+
version = "0.4.4"
|
376
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
377
|
+
checksum = "fc87660cd7daa49fddbfd524c836de54d5c927d520cd163f43700c5087c57d6c"
|
378
|
+
dependencies = [
|
379
|
+
"magnus-macros",
|
380
|
+
"rb-sys",
|
381
|
+
"rb-sys-env",
|
382
|
+
]
|
383
|
+
|
384
|
+
[[package]]
|
385
|
+
name = "magnus-macros"
|
386
|
+
version = "0.3.0"
|
387
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
388
|
+
checksum = "206cb23bfeea05180c97522ef6a3e52a4eb17b0ed2f30ee3ca9c4f994d2378ae"
|
389
|
+
dependencies = [
|
390
|
+
"proc-macro2",
|
391
|
+
"quote",
|
392
|
+
"syn",
|
393
|
+
]
|
394
|
+
|
395
|
+
[[package]]
|
396
|
+
name = "memchr"
|
397
|
+
version = "2.5.0"
|
398
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
399
|
+
checksum = "2dffe52ecf27772e601905b7522cb4ef790d2cc203488bbd0e2fe85fcb74566d"
|
400
|
+
|
401
|
+
[[package]]
|
402
|
+
name = "memoffset"
|
403
|
+
version = "0.7.1"
|
404
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
405
|
+
checksum = "5de893c32cde5f383baa4c04c5d6dbdd735cfd4a794b0debdb2bb1b421da5ff4"
|
406
|
+
dependencies = [
|
407
|
+
"autocfg",
|
408
|
+
]
|
409
|
+
|
410
|
+
[[package]]
|
411
|
+
name = "minimal-lexical"
|
412
|
+
version = "0.2.1"
|
413
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
414
|
+
checksum = "68354c5c6bd36d73ff3feceb05efa59b6acb7626617f4962be322a825e61f79a"
|
415
|
+
|
416
|
+
[[package]]
|
417
|
+
name = "nom"
|
418
|
+
version = "7.1.2"
|
419
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
420
|
+
checksum = "e5507769c4919c998e69e49c839d9dc6e693ede4cc4290d6ad8b41d4f09c548c"
|
421
|
+
dependencies = [
|
422
|
+
"memchr",
|
423
|
+
"minimal-lexical",
|
424
|
+
]
|
425
|
+
|
426
|
+
[[package]]
|
427
|
+
name = "num_cpus"
|
428
|
+
version = "1.15.0"
|
429
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
430
|
+
checksum = "0fac9e2da13b5eb447a6ce3d392f23a29d8694bff781bf03a16cd9ac8697593b"
|
431
|
+
dependencies = [
|
432
|
+
"hermit-abi",
|
433
|
+
"libc",
|
434
|
+
]
|
435
|
+
|
436
|
+
[[package]]
|
437
|
+
name = "number_prefix"
|
438
|
+
version = "0.3.0"
|
439
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
440
|
+
checksum = "17b02fc0ff9a9e4b35b3342880f48e896ebf69f2967921fe8646bf5b7125956a"
|
441
|
+
|
442
|
+
[[package]]
|
443
|
+
name = "once_cell"
|
444
|
+
version = "1.17.0"
|
445
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
446
|
+
checksum = "6f61fba1741ea2b3d6a1e3178721804bb716a68a6aeba1149b5d52e3d464ea66"
|
447
|
+
|
448
|
+
[[package]]
|
449
|
+
name = "onig"
|
450
|
+
version = "6.4.0"
|
451
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
452
|
+
checksum = "8c4b31c8722ad9171c6d77d3557db078cab2bd50afcc9d09c8b315c59df8ca4f"
|
453
|
+
dependencies = [
|
454
|
+
"bitflags",
|
455
|
+
"libc",
|
456
|
+
"once_cell",
|
457
|
+
"onig_sys",
|
458
|
+
]
|
459
|
+
|
460
|
+
[[package]]
|
461
|
+
name = "onig_sys"
|
462
|
+
version = "69.8.1"
|
463
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
464
|
+
checksum = "7b829e3d7e9cc74c7e315ee8edb185bf4190da5acde74afd7fc59c35b1f086e7"
|
465
|
+
dependencies = [
|
466
|
+
"cc",
|
467
|
+
"pkg-config",
|
468
|
+
]
|
469
|
+
|
470
|
+
[[package]]
|
471
|
+
name = "paste"
|
472
|
+
version = "1.0.11"
|
473
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
474
|
+
checksum = "d01a5bd0424d00070b0098dd17ebca6f961a959dead1dbcbbbc1d1cd8d3deeba"
|
475
|
+
|
476
|
+
[[package]]
|
477
|
+
name = "peeking_take_while"
|
478
|
+
version = "0.1.2"
|
479
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
480
|
+
checksum = "19b17cddbe7ec3f8bc800887bab5e717348c95ea2ca0b1bf0837fb964dc67099"
|
481
|
+
|
482
|
+
[[package]]
|
483
|
+
name = "pkg-config"
|
484
|
+
version = "0.3.26"
|
485
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
486
|
+
checksum = "6ac9a59f73473f1b8d852421e59e64809f025994837ef743615c6d0c5b305160"
|
487
|
+
|
488
|
+
[[package]]
|
489
|
+
name = "ppv-lite86"
|
490
|
+
version = "0.2.17"
|
491
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
492
|
+
checksum = "5b40af805b3121feab8a3c29f04d8ad262fa8e0561883e7653e024ae4479e6de"
|
493
|
+
|
494
|
+
[[package]]
|
495
|
+
name = "proc-macro2"
|
496
|
+
version = "1.0.49"
|
497
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
498
|
+
checksum = "57a8eca9f9c4ffde41714334dee777596264c7825420f521abc92b5b5deb63a5"
|
499
|
+
dependencies = [
|
500
|
+
"unicode-ident",
|
501
|
+
]
|
502
|
+
|
503
|
+
[[package]]
|
504
|
+
name = "quote"
|
505
|
+
version = "1.0.23"
|
506
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
507
|
+
checksum = "8856d8364d252a14d474036ea1358d63c9e6965c8e5c1885c18f73d70bff9c7b"
|
508
|
+
dependencies = [
|
509
|
+
"proc-macro2",
|
510
|
+
]
|
511
|
+
|
512
|
+
[[package]]
|
513
|
+
name = "rand"
|
514
|
+
version = "0.8.5"
|
515
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
516
|
+
checksum = "34af8d1a0e25924bc5b7c43c079c942339d8f0a8b57c39049bef581b46327404"
|
517
|
+
dependencies = [
|
518
|
+
"libc",
|
519
|
+
"rand_chacha",
|
520
|
+
"rand_core",
|
521
|
+
]
|
522
|
+
|
523
|
+
[[package]]
|
524
|
+
name = "rand_chacha"
|
525
|
+
version = "0.3.1"
|
526
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
527
|
+
checksum = "e6c10a63a0fa32252be49d21e7709d4d4baf8d231c2dbce1eaa8141b9b127d88"
|
528
|
+
dependencies = [
|
529
|
+
"ppv-lite86",
|
530
|
+
"rand_core",
|
531
|
+
]
|
532
|
+
|
533
|
+
[[package]]
|
534
|
+
name = "rand_core"
|
535
|
+
version = "0.6.4"
|
536
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
537
|
+
checksum = "ec0be4795e2f6a28069bec0b5ff3e2ac9bafc99e6a9a7dc3547996c5c816922c"
|
538
|
+
dependencies = [
|
539
|
+
"getrandom",
|
540
|
+
]
|
541
|
+
|
542
|
+
[[package]]
|
543
|
+
name = "rayon"
|
544
|
+
version = "1.6.1"
|
545
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
546
|
+
checksum = "6db3a213adf02b3bcfd2d3846bb41cb22857d131789e01df434fb7e7bc0759b7"
|
547
|
+
dependencies = [
|
548
|
+
"either",
|
549
|
+
"rayon-core",
|
550
|
+
]
|
551
|
+
|
552
|
+
[[package]]
|
553
|
+
name = "rayon-cond"
|
554
|
+
version = "0.1.0"
|
555
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
556
|
+
checksum = "fd1259362c9065e5ea39a789ef40b1e3fd934c94beb7b5ab3ac6629d3b5e7cb7"
|
557
|
+
dependencies = [
|
558
|
+
"either",
|
559
|
+
"itertools 0.8.2",
|
560
|
+
"rayon",
|
561
|
+
]
|
562
|
+
|
563
|
+
[[package]]
|
564
|
+
name = "rayon-core"
|
565
|
+
version = "1.10.1"
|
566
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
567
|
+
checksum = "cac410af5d00ab6884528b4ab69d1e8e146e8d471201800fa1b4524126de6ad3"
|
568
|
+
dependencies = [
|
569
|
+
"crossbeam-channel",
|
570
|
+
"crossbeam-deque",
|
571
|
+
"crossbeam-utils",
|
572
|
+
"num_cpus",
|
573
|
+
]
|
574
|
+
|
575
|
+
[[package]]
|
576
|
+
name = "rb-sys"
|
577
|
+
version = "0.9.56"
|
578
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
579
|
+
checksum = "ef82428221475c6f9e7893fe30b88d45ac86bdb12e58e7c92055ba4bceb78a69"
|
580
|
+
dependencies = [
|
581
|
+
"rb-sys-build",
|
582
|
+
]
|
583
|
+
|
584
|
+
[[package]]
|
585
|
+
name = "rb-sys-build"
|
586
|
+
version = "0.9.56"
|
587
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
588
|
+
checksum = "950bfc239d2e7704576abe4d37b008876bbfd70a99196a188c5caeae2ba7344a"
|
589
|
+
dependencies = [
|
590
|
+
"bindgen",
|
591
|
+
"regex",
|
592
|
+
"shell-words",
|
593
|
+
]
|
594
|
+
|
595
|
+
[[package]]
|
596
|
+
name = "rb-sys-env"
|
597
|
+
version = "0.1.1"
|
598
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
599
|
+
checksum = "74c38752410925faeb82c400c06ba2fd9ee6aa8f719dd33994c9e53f5242d25f"
|
600
|
+
|
601
|
+
[[package]]
|
602
|
+
name = "redox_syscall"
|
603
|
+
version = "0.2.16"
|
604
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
605
|
+
checksum = "fb5a58c1855b4b6819d59012155603f0b22ad30cad752600aadfcb695265519a"
|
606
|
+
dependencies = [
|
607
|
+
"bitflags",
|
608
|
+
]
|
609
|
+
|
610
|
+
[[package]]
|
611
|
+
name = "redox_users"
|
612
|
+
version = "0.4.3"
|
613
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
614
|
+
checksum = "b033d837a7cf162d7993aded9304e30a83213c648b6e389db233191f891e5c2b"
|
615
|
+
dependencies = [
|
616
|
+
"getrandom",
|
617
|
+
"redox_syscall",
|
618
|
+
"thiserror",
|
619
|
+
]
|
620
|
+
|
621
|
+
[[package]]
|
622
|
+
name = "regex"
|
623
|
+
version = "1.7.1"
|
624
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
625
|
+
checksum = "48aaa5748ba571fb95cd2c85c09f629215d3a6ece942baa100950af03a34f733"
|
626
|
+
dependencies = [
|
627
|
+
"aho-corasick",
|
628
|
+
"memchr",
|
629
|
+
"regex-syntax",
|
630
|
+
]
|
631
|
+
|
632
|
+
[[package]]
|
633
|
+
name = "regex-syntax"
|
634
|
+
version = "0.6.28"
|
635
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
636
|
+
checksum = "456c603be3e8d448b072f410900c09faf164fbce2d480456f50eea6e25f9c848"
|
637
|
+
|
638
|
+
[[package]]
|
639
|
+
name = "rustc-hash"
|
640
|
+
version = "1.1.0"
|
641
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
642
|
+
checksum = "08d43f7aa6b08d49f382cde6a7982047c3426db949b1424bc4b7ec9ae12c6ce2"
|
643
|
+
|
644
|
+
[[package]]
|
645
|
+
name = "ryu"
|
646
|
+
version = "1.0.12"
|
647
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
648
|
+
checksum = "7b4b9743ed687d4b4bcedf9ff5eaa7398495ae14e61cba0a295704edbc7decde"
|
649
|
+
|
650
|
+
[[package]]
|
651
|
+
name = "scopeguard"
|
652
|
+
version = "1.1.0"
|
653
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
654
|
+
checksum = "d29ab0c6d3fc0ee92fe66e2d99f700eab17a8d57d1c1d3b748380fb20baa78cd"
|
655
|
+
|
656
|
+
[[package]]
|
657
|
+
name = "serde"
|
658
|
+
version = "1.0.152"
|
659
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
660
|
+
checksum = "bb7d1f0d3021d347a83e556fc4683dea2ea09d87bccdf88ff5c12545d89d5efb"
|
661
|
+
dependencies = [
|
662
|
+
"serde_derive",
|
663
|
+
]
|
664
|
+
|
665
|
+
[[package]]
|
666
|
+
name = "serde_derive"
|
667
|
+
version = "1.0.152"
|
668
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
669
|
+
checksum = "af487d118eecd09402d70a5d72551860e788df87b464af30e5ea6a38c75c541e"
|
670
|
+
dependencies = [
|
671
|
+
"proc-macro2",
|
672
|
+
"quote",
|
673
|
+
"syn",
|
674
|
+
]
|
675
|
+
|
676
|
+
[[package]]
|
677
|
+
name = "serde_json"
|
678
|
+
version = "1.0.91"
|
679
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
680
|
+
checksum = "877c235533714907a8c2464236f5c4b2a17262ef1bd71f38f35ea592c8da6883"
|
681
|
+
dependencies = [
|
682
|
+
"itoa",
|
683
|
+
"ryu",
|
684
|
+
"serde",
|
685
|
+
]
|
686
|
+
|
687
|
+
[[package]]
|
688
|
+
name = "shell-words"
|
689
|
+
version = "1.1.0"
|
690
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
691
|
+
checksum = "24188a676b6ae68c3b2cb3a01be17fbf7240ce009799bb56d5b1409051e78fde"
|
692
|
+
|
693
|
+
[[package]]
|
694
|
+
name = "shlex"
|
695
|
+
version = "1.1.0"
|
696
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
697
|
+
checksum = "43b2853a4d09f215c24cc5489c992ce46052d359b5109343cbafbf26bc62f8a3"
|
698
|
+
|
699
|
+
[[package]]
|
700
|
+
name = "smallvec"
|
701
|
+
version = "1.10.0"
|
702
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
703
|
+
checksum = "a507befe795404456341dfab10cef66ead4c041f62b8b11bbb92bffe5d0953e0"
|
704
|
+
|
705
|
+
[[package]]
|
706
|
+
name = "spm_precompiled"
|
707
|
+
version = "0.1.4"
|
708
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
709
|
+
checksum = "5851699c4033c63636f7ea4cf7b7c1f1bf06d0cc03cfb42e711de5a5c46cf326"
|
710
|
+
dependencies = [
|
711
|
+
"base64",
|
712
|
+
"nom",
|
713
|
+
"serde",
|
714
|
+
"unicode-segmentation",
|
715
|
+
]
|
716
|
+
|
717
|
+
[[package]]
|
718
|
+
name = "strsim"
|
719
|
+
version = "0.10.0"
|
720
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
721
|
+
checksum = "73473c0e59e6d5812c5dfe2a064a6444949f089e20eec9a2e5506596494e4623"
|
722
|
+
|
723
|
+
[[package]]
|
724
|
+
name = "syn"
|
725
|
+
version = "1.0.107"
|
726
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
727
|
+
checksum = "1f4064b5b16e03ae50984a5a8ed5d4f8803e6bc1fd170a3cda91a1be4b18e3f5"
|
728
|
+
dependencies = [
|
729
|
+
"proc-macro2",
|
730
|
+
"quote",
|
731
|
+
"unicode-ident",
|
732
|
+
]
|
733
|
+
|
734
|
+
[[package]]
|
735
|
+
name = "thiserror"
|
736
|
+
version = "1.0.38"
|
737
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
738
|
+
checksum = "6a9cd18aa97d5c45c6603caea1da6628790b37f7a34b6ca89522331c5180fed0"
|
739
|
+
dependencies = [
|
740
|
+
"thiserror-impl",
|
741
|
+
]
|
742
|
+
|
743
|
+
[[package]]
|
744
|
+
name = "thiserror-impl"
|
745
|
+
version = "1.0.38"
|
746
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
747
|
+
checksum = "1fb327af4685e4d03fa8cbcf1716380da910eeb2bb8be417e7f9fd3fb164f36f"
|
748
|
+
dependencies = [
|
749
|
+
"proc-macro2",
|
750
|
+
"quote",
|
751
|
+
"syn",
|
752
|
+
]
|
753
|
+
|
754
|
+
[[package]]
|
755
|
+
name = "tokenizers"
|
756
|
+
version = "0.2.2"
|
757
|
+
dependencies = [
|
758
|
+
"magnus",
|
759
|
+
"tokenizers 0.13.2",
|
760
|
+
]
|
761
|
+
|
762
|
+
[[package]]
|
763
|
+
name = "tokenizers"
|
764
|
+
version = "0.13.2"
|
765
|
+
source = "git+https://github.com/huggingface/tokenizers#fe4ae7dc38be11a5c93ae703816c869f993c21ab"
|
766
|
+
dependencies = [
|
767
|
+
"aho-corasick",
|
768
|
+
"derive_builder",
|
769
|
+
"dirs",
|
770
|
+
"esaxx-rs",
|
771
|
+
"getrandom",
|
772
|
+
"indicatif",
|
773
|
+
"itertools 0.9.0",
|
774
|
+
"lazy_static",
|
775
|
+
"log",
|
776
|
+
"macro_rules_attribute",
|
777
|
+
"onig",
|
778
|
+
"paste",
|
779
|
+
"rand",
|
780
|
+
"rayon",
|
781
|
+
"rayon-cond",
|
782
|
+
"regex",
|
783
|
+
"regex-syntax",
|
784
|
+
"serde",
|
785
|
+
"serde_json",
|
786
|
+
"spm_precompiled",
|
787
|
+
"thiserror",
|
788
|
+
"unicode-normalization-alignments",
|
789
|
+
"unicode-segmentation",
|
790
|
+
"unicode_categories",
|
791
|
+
]
|
792
|
+
|
793
|
+
[[package]]
|
794
|
+
name = "unicode-ident"
|
795
|
+
version = "1.0.6"
|
796
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
797
|
+
checksum = "84a22b9f218b40614adcb3f4ff08b703773ad44fa9423e4e0d346d5db86e4ebc"
|
798
|
+
|
799
|
+
[[package]]
|
800
|
+
name = "unicode-normalization-alignments"
|
801
|
+
version = "0.1.12"
|
802
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
803
|
+
checksum = "43f613e4fa046e69818dd287fdc4bc78175ff20331479dab6e1b0f98d57062de"
|
804
|
+
dependencies = [
|
805
|
+
"smallvec",
|
806
|
+
]
|
807
|
+
|
808
|
+
[[package]]
|
809
|
+
name = "unicode-segmentation"
|
810
|
+
version = "1.10.0"
|
811
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
812
|
+
checksum = "0fdbf052a0783de01e944a6ce7a8cb939e295b1e7be835a1112c3b9a7f047a5a"
|
813
|
+
|
814
|
+
[[package]]
|
815
|
+
name = "unicode-width"
|
816
|
+
version = "0.1.10"
|
817
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
818
|
+
checksum = "c0edd1e5b14653f783770bce4a4dabb4a5108a5370a5f5d8cfe8710c361f6c8b"
|
819
|
+
|
820
|
+
[[package]]
|
821
|
+
name = "unicode_categories"
|
822
|
+
version = "0.1.1"
|
823
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
824
|
+
checksum = "39ec24b3121d976906ece63c9daad25b85969647682eee313cb5779fdd69e14e"
|
825
|
+
|
826
|
+
[[package]]
|
827
|
+
name = "wasi"
|
828
|
+
version = "0.11.0+wasi-snapshot-preview1"
|
829
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
830
|
+
checksum = "9c8d87e72b64a3b4db28d11ce29237c246188f4f51057d65a7eab63b7987e423"
|
831
|
+
|
832
|
+
[[package]]
|
833
|
+
name = "winapi"
|
834
|
+
version = "0.3.9"
|
835
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
836
|
+
checksum = "5c839a674fcd7a98952e593242ea400abe93992746761e38641405d28b00f419"
|
837
|
+
dependencies = [
|
838
|
+
"winapi-i686-pc-windows-gnu",
|
839
|
+
"winapi-x86_64-pc-windows-gnu",
|
840
|
+
]
|
841
|
+
|
842
|
+
[[package]]
|
843
|
+
name = "winapi-i686-pc-windows-gnu"
|
844
|
+
version = "0.4.0"
|
845
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
846
|
+
checksum = "ac3b87c63620426dd9b991e5ce0329eff545bccbbb34f3be09ff6fb6ab51b7b6"
|
847
|
+
|
848
|
+
[[package]]
|
849
|
+
name = "winapi-x86_64-pc-windows-gnu"
|
850
|
+
version = "0.4.0"
|
851
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
852
|
+
checksum = "712e227841d057c1ee1cd2fb22fa7e5a5461ae8e48fa2ca79ec42cfc1931183f"
|
853
|
+
|
854
|
+
[[package]]
|
855
|
+
name = "windows-sys"
|
856
|
+
version = "0.42.0"
|
857
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
858
|
+
checksum = "5a3e1820f08b8513f676f7ab6c1f99ff312fb97b553d30ff4dd86f9f15728aa7"
|
859
|
+
dependencies = [
|
860
|
+
"windows_aarch64_gnullvm",
|
861
|
+
"windows_aarch64_msvc",
|
862
|
+
"windows_i686_gnu",
|
863
|
+
"windows_i686_msvc",
|
864
|
+
"windows_x86_64_gnu",
|
865
|
+
"windows_x86_64_gnullvm",
|
866
|
+
"windows_x86_64_msvc",
|
867
|
+
]
|
868
|
+
|
869
|
+
[[package]]
|
870
|
+
name = "windows_aarch64_gnullvm"
|
871
|
+
version = "0.42.1"
|
872
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
873
|
+
checksum = "8c9864e83243fdec7fc9c5444389dcbbfd258f745e7853198f365e3c4968a608"
|
874
|
+
|
875
|
+
[[package]]
|
876
|
+
name = "windows_aarch64_msvc"
|
877
|
+
version = "0.42.1"
|
878
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
879
|
+
checksum = "4c8b1b673ffc16c47a9ff48570a9d85e25d265735c503681332589af6253c6c7"
|
880
|
+
|
881
|
+
[[package]]
|
882
|
+
name = "windows_i686_gnu"
|
883
|
+
version = "0.42.1"
|
884
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
885
|
+
checksum = "de3887528ad530ba7bdbb1faa8275ec7a1155a45ffa57c37993960277145d640"
|
886
|
+
|
887
|
+
[[package]]
|
888
|
+
name = "windows_i686_msvc"
|
889
|
+
version = "0.42.1"
|
890
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
891
|
+
checksum = "bf4d1122317eddd6ff351aa852118a2418ad4214e6613a50e0191f7004372605"
|
892
|
+
|
893
|
+
[[package]]
|
894
|
+
name = "windows_x86_64_gnu"
|
895
|
+
version = "0.42.1"
|
896
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
897
|
+
checksum = "c1040f221285e17ebccbc2591ffdc2d44ee1f9186324dd3e84e99ac68d699c45"
|
898
|
+
|
899
|
+
[[package]]
|
900
|
+
name = "windows_x86_64_gnullvm"
|
901
|
+
version = "0.42.1"
|
902
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
903
|
+
checksum = "628bfdf232daa22b0d64fdb62b09fcc36bb01f05a3939e20ab73aaf9470d0463"
|
904
|
+
|
905
|
+
[[package]]
|
906
|
+
name = "windows_x86_64_msvc"
|
907
|
+
version = "0.42.1"
|
908
|
+
source = "registry+https://github.com/rust-lang/crates.io-index"
|
909
|
+
checksum = "447660ad36a13288b1db4d4248e857b510e8c3a225c822ba4fb748c0aafecffd"
|