Hugging Face's logo

Datasets:
HuggingFaceFW
/
fineweb

Tasks:
Text Generation
Languages:
English
Size Categories:
n>1T
ArXiv:
Tags:
DOI:
License:
Dataset card Viewer Files Community
25
guipenedo HF staff lvwerra HF staff commited on
Commit
dcedf0f
•
0 Parent(s):

Super-squash branch 'main' using huggingface_hub

Browse files

Co-authored-by: lvwerra <lvwerra@users.noreply.huggingface.co>

This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +55 -0
  2. README.md +791 -0
  3. data/CC-MAIN-2013-20/000_00000.parquet +3 -0
  4. data/CC-MAIN-2013-20/000_00001.parquet +3 -0
  5. data/CC-MAIN-2013-20/000_00002.parquet +3 -0
  6. data/CC-MAIN-2013-20/000_00003.parquet +3 -0
  7. data/CC-MAIN-2013-20/000_00004.parquet +3 -0
  8. data/CC-MAIN-2013-20/000_00005.parquet +3 -0
  9. data/CC-MAIN-2013-20/000_00006.parquet +3 -0
  10. data/CC-MAIN-2013-20/000_00007.parquet +3 -0
  11. data/CC-MAIN-2013-20/000_00008.parquet +3 -0
  12. data/CC-MAIN-2013-20/000_00009.parquet +3 -0
  13. data/CC-MAIN-2013-20/000_00010.parquet +3 -0
  14. data/CC-MAIN-2013-20/000_00011.parquet +3 -0
  15. data/CC-MAIN-2013-20/000_00012.parquet +3 -0
  16. data/CC-MAIN-2013-20/000_00013.parquet +3 -0
  17. data/CC-MAIN-2013-20/000_00014.parquet +3 -0
  18. data/CC-MAIN-2013-20/000_00015.parquet +3 -0
  19. data/CC-MAIN-2013-20/000_00016.parquet +3 -0
  20. data/CC-MAIN-2013-20/000_00017.parquet +3 -0
  21. data/CC-MAIN-2013-20/000_00018.parquet +3 -0
  22. data/CC-MAIN-2013-20/000_00019.parquet +3 -0
  23. data/CC-MAIN-2013-20/000_00020.parquet +3 -0
  24. data/CC-MAIN-2013-20/000_00021.parquet +3 -0
  25. data/CC-MAIN-2013-20/000_00022.parquet +3 -0
  26. data/CC-MAIN-2013-20/000_00023.parquet +3 -0
  27. data/CC-MAIN-2013-20/000_00024.parquet +3 -0
  28. data/CC-MAIN-2013-20/000_00025.parquet +3 -0
  29. data/CC-MAIN-2013-20/000_00026.parquet +3 -0
  30. data/CC-MAIN-2013-20/000_00027.parquet +3 -0
  31. data/CC-MAIN-2013-20/000_00028.parquet +3 -0
  32. data/CC-MAIN-2013-20/000_00029.parquet +3 -0
  33. data/CC-MAIN-2013-20/000_00030.parquet +3 -0
  34. data/CC-MAIN-2013-20/000_00031.parquet +3 -0
  35. data/CC-MAIN-2013-20/000_00032.parquet +3 -0
  36. data/CC-MAIN-2013-20/000_00033.parquet +3 -0
  37. data/CC-MAIN-2013-20/000_00034.parquet +3 -0
  38. data/CC-MAIN-2013-20/000_00035.parquet +3 -0
  39. data/CC-MAIN-2013-20/000_00036.parquet +3 -0
  40. data/CC-MAIN-2013-20/000_00037.parquet +3 -0
  41. data/CC-MAIN-2013-20/000_00038.parquet +3 -0
  42. data/CC-MAIN-2013-20/000_00039.parquet +3 -0
  43. data/CC-MAIN-2013-20/000_00040.parquet +3 -0
  44. data/CC-MAIN-2013-20/000_00041.parquet +3 -0
  45. data/CC-MAIN-2013-20/000_00042.parquet +3 -0
  46. data/CC-MAIN-2013-20/000_00043.parquet +3 -0
  47. data/CC-MAIN-2013-20/000_00044.parquet +3 -0
  48. data/CC-MAIN-2013-20/000_00045.parquet +3 -0
  49. data/CC-MAIN-2013-20/000_00046.parquet +3 -0
  50. data/CC-MAIN-2013-20/000_00047.parquet +3 -0
.gitattributes ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.lz4 filter=lfs diff=lfs merge=lfs -text
12
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
13
+ *.model filter=lfs diff=lfs merge=lfs -text
14
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
15
+ *.npy filter=lfs diff=lfs merge=lfs -text
16
+ *.npz filter=lfs diff=lfs merge=lfs -text
17
+ *.onnx filter=lfs diff=lfs merge=lfs -text
18
+ *.ot filter=lfs diff=lfs merge=lfs -text
19
+ *.parquet filter=lfs diff=lfs merge=lfs -text
20
+ *.pb filter=lfs diff=lfs merge=lfs -text
21
+ *.pickle filter=lfs diff=lfs merge=lfs -text
22
+ *.pkl filter=lfs diff=lfs merge=lfs -text
23
+ *.pt filter=lfs diff=lfs merge=lfs -text
24
+ *.pth filter=lfs diff=lfs merge=lfs -text
25
+ *.rar filter=lfs diff=lfs merge=lfs -text
26
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
27
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
29
+ *.tar filter=lfs diff=lfs merge=lfs -text
30
+ *.tflite filter=lfs diff=lfs merge=lfs -text
31
+ *.tgz filter=lfs diff=lfs merge=lfs -text
32
+ *.wasm filter=lfs diff=lfs merge=lfs -text
33
+ *.xz filter=lfs diff=lfs merge=lfs -text
34
+ *.zip filter=lfs diff=lfs merge=lfs -text
35
+ *.zst filter=lfs diff=lfs merge=lfs -text
36
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
37
+ # Audio files - uncompressed
38
+ *.pcm filter=lfs diff=lfs merge=lfs -text
39
+ *.sam filter=lfs diff=lfs merge=lfs -text
40
+ *.raw filter=lfs diff=lfs merge=lfs -text
41
+ # Audio files - compressed
42
+ *.aac filter=lfs diff=lfs merge=lfs -text
43
+ *.flac filter=lfs diff=lfs merge=lfs -text
44
+ *.mp3 filter=lfs diff=lfs merge=lfs -text
45
+ *.ogg filter=lfs diff=lfs merge=lfs -text
46
+ *.wav filter=lfs diff=lfs merge=lfs -text
47
+ # Image files - uncompressed
48
+ *.bmp filter=lfs diff=lfs merge=lfs -text
49
+ *.gif filter=lfs diff=lfs merge=lfs -text
50
+ *.png filter=lfs diff=lfs merge=lfs -text
51
+ *.tiff filter=lfs diff=lfs merge=lfs -text
52
+ # Image files - compressed
53
+ *.jpg filter=lfs diff=lfs merge=lfs -text
54
+ *.jpeg filter=lfs diff=lfs merge=lfs -text
55
+ *.webp filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,791 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: odc-by
3
+ task_categories:
4
+ - text-generation
5
+ language:
6
+ - en
7
+ pretty_name: FineWeb
8
+ size_categories:
9
+ - n>1T
10
+ configs:
11
+ - config_name: default
12
+ data_files:
13
+ - split: train
14
+ path: data/*/*
15
+ - config_name: CC-MAIN-2024-10
16
+ data_files:
17
+ - split: train
18
+ path: data/CC-MAIN-2024-10/*
19
+ - config_name: CC-MAIN-2023-50
20
+ data_files:
21
+ - split: train
22
+ path: data/CC-MAIN-2023-50/*
23
+ - config_name: CC-MAIN-2023-40
24
+ data_files:
25
+ - split: train
26
+ path: data/CC-MAIN-2023-40/*
27
+ - config_name: CC-MAIN-2023-23
28
+ data_files:
29
+ - split: train
30
+ path: data/CC-MAIN-2023-23/*
31
+ - config_name: CC-MAIN-2023-14
32
+ data_files:
33
+ - split: train
34
+ path: data/CC-MAIN-2023-14/*
35
+ - config_name: CC-MAIN-2023-06
36
+ data_files:
37
+ - split: train
38
+ path: data/CC-MAIN-2023-06/*
39
+ - config_name: CC-MAIN-2022-49
40
+ data_files:
41
+ - split: train
42
+ path: data/CC-MAIN-2022-49/*
43
+ - config_name: CC-MAIN-2022-40
44
+ data_files:
45
+ - split: train
46
+ path: data/CC-MAIN-2022-40/*
47
+ - config_name: CC-MAIN-2022-33
48
+ data_files:
49
+ - split: train
50
+ path: data/CC-MAIN-2022-33/*
51
+ - config_name: CC-MAIN-2022-27
52
+ data_files:
53
+ - split: train
54
+ path: data/CC-MAIN-2022-27/*
55
+ - config_name: CC-MAIN-2022-21
56
+ data_files:
57
+ - split: train
58
+ path: data/CC-MAIN-2022-21/*
59
+ - config_name: CC-MAIN-2022-05
60
+ data_files:
61
+ - split: train
62
+ path: data/CC-MAIN-2022-05/*
63
+ - config_name: CC-MAIN-2021-49
64
+ data_files:
65
+ - split: train
66
+ path: data/CC-MAIN-2021-49/*
67
+ - config_name: CC-MAIN-2021-43
68
+ data_files:
69
+ - split: train
70
+ path: data/CC-MAIN-2021-43/*
71
+ - config_name: CC-MAIN-2021-39
72
+ data_files:
73
+ - split: train
74
+ path: data/CC-MAIN-2021-39/*
75
+ - config_name: CC-MAIN-2021-31
76
+ data_files:
77
+ - split: train
78
+ path: data/CC-MAIN-2021-31/*
79
+ - config_name: CC-MAIN-2021-25
80
+ data_files:
81
+ - split: train
82
+ path: data/CC-MAIN-2021-25/*
83
+ - config_name: CC-MAIN-2021-21
84
+ data_files:
85
+ - split: train
86
+ path: data/CC-MAIN-2021-21/*
87
+ - config_name: CC-MAIN-2021-17
88
+ data_files:
89
+ - split: train
90
+ path: data/CC-MAIN-2021-17/*
91
+ - config_name: CC-MAIN-2021-10
92
+ data_files:
93
+ - split: train
94
+ path: data/CC-MAIN-2021-10/*
95
+ - config_name: CC-MAIN-2021-04
96
+ data_files:
97
+ - split: train
98
+ path: data/CC-MAIN-2021-04/*
99
+ - config_name: CC-MAIN-2020-50
100
+ data_files:
101
+ - split: train
102
+ path: data/CC-MAIN-2020-50/*
103
+ - config_name: CC-MAIN-2020-45
104
+ data_files:
105
+ - split: train
106
+ path: data/CC-MAIN-2020-45/*
107
+ - config_name: CC-MAIN-2020-40
108
+ data_files:
109
+ - split: train
110
+ path: data/CC-MAIN-2020-40/*
111
+ - config_name: CC-MAIN-2020-34
112
+ data_files:
113
+ - split: train
114
+ path: data/CC-MAIN-2020-34/*
115
+ - config_name: CC-MAIN-2020-29
116
+ data_files:
117
+ - split: train
118
+ path: data/CC-MAIN-2020-29/*
119
+ - config_name: CC-MAIN-2020-24
120
+ data_files:
121
+ - split: train
122
+ path: data/CC-MAIN-2020-24/*
123
+ - config_name: CC-MAIN-2020-16
124
+ data_files:
125
+ - split: train
126
+ path: data/CC-MAIN-2020-16/*
127
+ - config_name: CC-MAIN-2020-10
128
+ data_files:
129
+ - split: train
130
+ path: data/CC-MAIN-2020-10/*
131
+ - config_name: CC-MAIN-2020-05
132
+ data_files:
133
+ - split: train
134
+ path: data/CC-MAIN-2020-05/*
135
+ - config_name: CC-MAIN-2019-51
136
+ data_files:
137
+ - split: train
138
+ path: data/CC-MAIN-2019-51/*
139
+ - config_name: CC-MAIN-2019-47
140
+ data_files:
141
+ - split: train
142
+ path: data/CC-MAIN-2019-47/*
143
+ - config_name: CC-MAIN-2019-43
144
+ data_files:
145
+ - split: train
146
+ path: data/CC-MAIN-2019-43/*
147
+ - config_name: CC-MAIN-2019-39
148
+ data_files:
149
+ - split: train
150
+ path: data/CC-MAIN-2019-39/*
151
+ - config_name: CC-MAIN-2019-35
152
+ data_files:
153
+ - split: train
154
+ path: data/CC-MAIN-2019-35/*
155
+ - config_name: CC-MAIN-2019-30
156
+ data_files:
157
+ - split: train
158
+ path: data/CC-MAIN-2019-30/*
159
+ - config_name: CC-MAIN-2019-26
160
+ data_files:
161
+ - split: train
162
+ path: data/CC-MAIN-2019-26/*
163
+ - config_name: CC-MAIN-2019-22
164
+ data_files:
165
+ - split: train
166
+ path: data/CC-MAIN-2019-22/*
167
+ - config_name: CC-MAIN-2019-18
168
+ data_files:
169
+ - split: train
170
+ path: data/CC-MAIN-2019-18/*
171
+ - config_name: CC-MAIN-2019-13
172
+ data_files:
173
+ - split: train
174
+ path: data/CC-MAIN-2019-13/*
175
+ - config_name: CC-MAIN-2019-09
176
+ data_files:
177
+ - split: train
178
+ path: data/CC-MAIN-2019-09/*
179
+ - config_name: CC-MAIN-2019-04
180
+ data_files:
181
+ - split: train
182
+ path: data/CC-MAIN-2019-04/*
183
+ - config_name: CC-MAIN-2018-51
184
+ data_files:
185
+ - split: train
186
+ path: data/CC-MAIN-2018-51/*
187
+ - config_name: CC-MAIN-2018-47
188
+ data_files:
189
+ - split: train
190
+ path: data/CC-MAIN-2018-47/*
191
+ - config_name: CC-MAIN-2018-43
192
+ data_files:
193
+ - split: train
194
+ path: data/CC-MAIN-2018-43/*
195
+ - config_name: CC-MAIN-2018-39
196
+ data_files:
197
+ - split: train
198
+ path: data/CC-MAIN-2018-39/*
199
+ - config_name: CC-MAIN-2018-34
200
+ data_files:
201
+ - split: train
202
+ path: data/CC-MAIN-2018-34/*
203
+ - config_name: CC-MAIN-2018-30
204
+ data_files:
205
+ - split: train
206
+ path: data/CC-MAIN-2018-30/*
207
+ - config_name: CC-MAIN-2018-26
208
+ data_files:
209
+ - split: train
210
+ path: data/CC-MAIN-2018-26/*
211
+ - config_name: CC-MAIN-2018-22
212
+ data_files:
213
+ - split: train
214
+ path: data/CC-MAIN-2018-22/*
215
+ - config_name: CC-MAIN-2018-17
216
+ data_files:
217
+ - split: train
218
+ path: data/CC-MAIN-2018-17/*
219
+ - config_name: CC-MAIN-2018-13
220
+ data_files:
221
+ - split: train
222
+ path: data/CC-MAIN-2018-13/*
223
+ - config_name: CC-MAIN-2018-09
224
+ data_files:
225
+ - split: train
226
+ path: data/CC-MAIN-2018-09/*
227
+ - config_name: CC-MAIN-2018-05
228
+ data_files:
229
+ - split: train
230
+ path: data/CC-MAIN-2018-05/*
231
+ - config_name: CC-MAIN-2017-51
232
+ data_files:
233
+ - split: train
234
+ path: data/CC-MAIN-2017-51/*
235
+ - config_name: CC-MAIN-2017-47
236
+ data_files:
237
+ - split: train
238
+ path: data/CC-MAIN-2017-47/*
239
+ - config_name: CC-MAIN-2017-43
240
+ data_files:
241
+ - split: train
242
+ path: data/CC-MAIN-2017-43/*
243
+ - config_name: CC-MAIN-2017-39
244
+ data_files:
245
+ - split: train
246
+ path: data/CC-MAIN-2017-39/*
247
+ - config_name: CC-MAIN-2017-34
248
+ data_files:
249
+ - split: train
250
+ path: data/CC-MAIN-2017-34/*
251
+ - config_name: CC-MAIN-2017-30
252
+ data_files:
253
+ - split: train
254
+ path: data/CC-MAIN-2017-30/*
255
+ - config_name: CC-MAIN-2017-26
256
+ data_files:
257
+ - split: train
258
+ path: data/CC-MAIN-2017-26/*
259
+ - config_name: CC-MAIN-2017-22
260
+ data_files:
261
+ - split: train
262
+ path: data/CC-MAIN-2017-22/*
263
+ - config_name: CC-MAIN-2017-17
264
+ data_files:
265
+ - split: train
266
+ path: data/CC-MAIN-2017-17/*
267
+ - config_name: CC-MAIN-2017-13
268
+ data_files:
269
+ - split: train
270
+ path: data/CC-MAIN-2017-13/*
271
+ - config_name: CC-MAIN-2017-09
272
+ data_files:
273
+ - split: train
274
+ path: data/CC-MAIN-2017-09/*
275
+ - config_name: CC-MAIN-2017-04
276
+ data_files:
277
+ - split: train
278
+ path: data/CC-MAIN-2017-04/*
279
+ - config_name: CC-MAIN-2016-50
280
+ data_files:
281
+ - split: train
282
+ path: data/CC-MAIN-2016-50/*
283
+ - config_name: CC-MAIN-2016-44
284
+ data_files:
285
+ - split: train
286
+ path: data/CC-MAIN-2016-44/*
287
+ - config_name: CC-MAIN-2016-40
288
+ data_files:
289
+ - split: train
290
+ path: data/CC-MAIN-2016-40/*
291
+ - config_name: CC-MAIN-2016-36
292
+ data_files:
293
+ - split: train
294
+ path: data/CC-MAIN-2016-36/*
295
+ - config_name: CC-MAIN-2016-30
296
+ data_files:
297
+ - split: train
298
+ path: data/CC-MAIN-2016-30/*
299
+ - config_name: CC-MAIN-2016-26
300
+ data_files:
301
+ - split: train
302
+ path: data/CC-MAIN-2016-26/*
303
+ - config_name: CC-MAIN-2016-22
304
+ data_files:
305
+ - split: train
306
+ path: data/CC-MAIN-2016-22/*
307
+ - config_name: CC-MAIN-2016-18
308
+ data_files:
309
+ - split: train
310
+ path: data/CC-MAIN-2016-18/*
311
+ - config_name: CC-MAIN-2016-07
312
+ data_files:
313
+ - split: train
314
+ path: data/CC-MAIN-2016-07/*
315
+ - config_name: CC-MAIN-2015-48
316
+ data_files:
317
+ - split: train
318
+ path: data/CC-MAIN-2015-48/*
319
+ - config_name: CC-MAIN-2015-40
320
+ data_files:
321
+ - split: train
322
+ path: data/CC-MAIN-2015-40/*
323
+ - config_name: CC-MAIN-2015-35
324
+ data_files:
325
+ - split: train
326
+ path: data/CC-MAIN-2015-35/*
327
+ - config_name: CC-MAIN-2015-32
328
+ data_files:
329
+ - split: train
330
+ path: data/CC-MAIN-2015-32/*
331
+ - config_name: CC-MAIN-2015-27
332
+ data_files:
333
+ - split: train
334
+ path: data/CC-MAIN-2015-27/*
335
+ - config_name: CC-MAIN-2015-22
336
+ data_files:
337
+ - split: train
338
+ path: data/CC-MAIN-2015-22/*
339
+ - config_name: CC-MAIN-2015-18
340
+ data_files:
341
+ - split: train
342
+ path: data/CC-MAIN-2015-18/*
343
+ - config_name: CC-MAIN-2015-14
344
+ data_files:
345
+ - split: train
346
+ path: data/CC-MAIN-2015-14/*
347
+ - config_name: CC-MAIN-2015-11
348
+ data_files:
349
+ - split: train
350
+ path: data/CC-MAIN-2015-11/*
351
+ - config_name: CC-MAIN-2015-06
352
+ data_files:
353
+ - split: train
354
+ path: data/CC-MAIN-2015-06/*
355
+ - config_name: CC-MAIN-2014-52
356
+ data_files:
357
+ - split: train
358
+ path: data/CC-MAIN-2014-52/*
359
+ - config_name: CC-MAIN-2014-49
360
+ data_files:
361
+ - split: train
362
+ path: data/CC-MAIN-2014-49/*
363
+ - config_name: CC-MAIN-2014-42
364
+ data_files:
365
+ - split: train
366
+ path: data/CC-MAIN-2014-42/*
367
+ - config_name: CC-MAIN-2014-41
368
+ data_files:
369
+ - split: train
370
+ path: data/CC-MAIN-2014-41/*
371
+ - config_name: CC-MAIN-2014-35
372
+ data_files:
373
+ - split: train
374
+ path: data/CC-MAIN-2014-35/*
375
+ - config_name: CC-MAIN-2014-23
376
+ data_files:
377
+ - split: train
378
+ path: data/CC-MAIN-2014-23/*
379
+ - config_name: CC-MAIN-2014-15
380
+ data_files:
381
+ - split: train
382
+ path: data/CC-MAIN-2014-15/*
383
+ - config_name: CC-MAIN-2014-10
384
+ data_files:
385
+ - split: train
386
+ path: data/CC-MAIN-2014-10/*
387
+ - config_name: CC-MAIN-2013-48
388
+ data_files:
389
+ - split: train
390
+ path: data/CC-MAIN-2013-48/*
391
+ - config_name: CC-MAIN-2013-20
392
+ data_files:
393
+ - split: train
394
+ path: data/CC-MAIN-2013-20/*
395
+ ---
396
+ # 🍷 FineWeb
397
+ <center>
398
+ <img src="https://huggingface.co/datasets/HuggingFaceFW/admin/resolve/main/fineweb-logo.png" alt="FineWeb: The finest collection of data the web has to offer">
399
+ </center>
400
+
401
+ > 15 trillion tokens of the finest data the 🌐 web has to offer
402
+
403
+ # Table of Contents
404
+ - [🍷 FineWeb](#-fineweb)
405
+ * [What is it?](#what-is-it)
406
+ * [What is being released?](#what-is-being-released)
407
+ * [How to download and use 🍷 FineWeb](#how-to-download-and-use-🍷-fineweb)
408
+ + [Using 🏭 `datatrove`](#using-datatrove)
409
+ + [Using `huggingface_hub`](#using-huggingface_hub)
410
+ + [Using `datasets`](#using-datasets)
411
+ * [Breakdown by dump/crawl](#breakdown-by-dumpcrawl)
412
+ * [Dataset performance evaluation and ablations](#dataset-performance-evaluation-and-ablations)
413
+ + [Hyper-parameters for ablation models](#hyper-parameters-for-ablation-models)
414
+ + [Ablation evaluation benchmarks](#ablation-evaluation-benchmarks)
415
+ + [Comparison with other datasets](#comparison-with-other-datasets)
416
+ - [Dataset card for 🍷 FineWeb](#dataset-card-for-🍷-fineweb)
417
+ * [Dataset Summary](#dataset-summary)
418
+ * [Dataset Structure](#dataset-structure)
419
+ + [Data Instances](#data-instances)
420
+ + [Data Fields](#data-fields)
421
+ + [Data Splits](#data-splits)
422
+ * [Dataset Creation](#dataset-creation)
423
+ + [Curation Rationale](#curation-rationale)
424
+ + [Source Data](#source-data)
425
+ + [Data processing steps](#data-processing-steps)
426
+ + [Annotations](#annotations)
427
+ + [Personal and Sensitive Information](#personal-and-sensitive-information)
428
+ * [Considerations for Using the Data](#considerations-for-using-the-data)
429
+ + [Social Impact of Dataset](#social-impact-of-dataset)
430
+ + [Discussion of Biases](#discussion-of-biases)
431
+ + [Other Known Limitations](#other-known-limitations)
432
+ * [Additional Information](#additional-information)
433
+ + [Licensing Information](#licensing-information)
434
+ + [Future work](#future-work)
435
+ + [Citation Information](#citation-information)
436
+
437
+ ## What is it?
438
+
439
+ The 🍷 FineWeb dataset consists of more than **15T tokens** of cleaned and deduplicated english web data from CommonCrawl. The data processing pipeline is optimized for LLM performance and ran on the 🏭 [`datatrove`](https://github.com/huggingface/datatrove/) library, our large scale data processing library.
440
+
441
+ 🍷 FineWeb was originally meant to be a fully open replication of 🦅 [RefinedWeb](https://huggingface.co/papers/2306.01116), with a release of the **full dataset** under the **ODC-By 1.0 license**. However, by carefully adding additional filtering steps, we managed to push the performance of 🍷 FineWeb well above that of the original 🦅 RefinedWeb, and models trained on our dataset also outperform models trained on other commonly used high quality web datasets (like C4, Dolma-v1.6, The Pile, SlimPajama, RedPajam2) on our aggregate group of [benchmark tasks](https://huggingface.co/datasets/HuggingFaceFW/fineweb/blob/main/lighteval_tasks.py).
442
+
443
+ That said, we think there is still room for additional filtering and improvement and intend to continue exploring how to improve the dataset quality in coming versions of 🍷 FineWeb.
444
+
445
+ ## What is being released?
446
+
447
+ Along with the dataset, which includes all CommonCrawl dumps since 2013, we also share all the code needed to fully reproduce our processing setup using the 🏭 [`datatrove`](https://github.com/huggingface/datatrove/) library [here](https://github.com/huggingface/datatrove/blob/main/examples/fineweb.py). To enable full replication of our results, we have also published the small ablation models we have trained using [`nanotron`](https://github.com/huggingface/nanotron/) to validate the dataset and compare it with other reference datasets. You will find them [here](https://huggingface.co/collections/HuggingFaceFW/ablation-models-662457b0d213e8c14fe47f32), with checkpoints every 1000 steps. We have also published our evaluation results [here](https://huggingface.co/datasets/HuggingFaceFW/fineweb/blob/main/eval_results.csv). Our evaluation setup is available [here](https://huggingface.co/datasets/HuggingFaceFW/fineweb/blob/main/lighteval_tasks.py).
448
+
449
+ You will find details on the different processing decisions we took and some interesting explorations of deduplication methods and differences between CommonCrawl dumps in our technical report to be published in the coming days.
450
+
451
+ ## How to download and use 🍷 FineWeb
452
+
453
+ You can load the full dataset or a specific crawl/dump (see table below). Dumps have the format `CC-MAIN-(year)-(week number)`.
454
+
455
+ ### Using 🏭 [`datatrove`](https://github.com/huggingface/datatrove/)
456
+
457
+ ```python
458
+ from datatrove.pipeline.readers import ParquetReader
459
+
460
+ # limit determines how many documents will be streamed (remove for all)
461
+ # to fetch a specific dump: hf://datasets/HuggingFaceFW/fineweb/data/CC-MAIN-2024-10
462
+ data_reader = ParquetReader("hf://datasets/HuggingFaceFW/fineweb/data", limit=1000)
463
+ for document in data_reader():
464
+ # do something with document
465
+ print(document)
466
+
467
+ ###############################
468
+ # OR for a processing pipeline:
469
+ ###############################
470
+
471
+ from datatrove.executor import LocalPipelineExecutor
472
+ from datatrove.pipeline.readers import ParquetReader
473
+ from datatrove.pipeline.filters import LambdaFilter
474
+ from datatrove.pipeline.writers import JsonlWriter
475
+
476
+ pipeline_exec = LocalPipelineExecutor(
477
+ pipeline=[
478
+ ParquetReader("hf://datasets/HuggingFaceFW/fineweb/data/CC-MAIN-2024-10", limit=1000),
479
+ LambdaFilter(lambda doc: "hugging" in doc.text),
480
+ JsonlWriter("some-output-path")
481
+ ],
482
+ tasks=10
483
+ )
484
+ pipeline_exec.run()
485
+ ```
486
+
487
+ ### Using `huggingface_hub`
488
+
489
+ ```python
490
+ from huggingface_hub import snapshot_download
491
+ folder = snapshot_download(
492
+ "HuggingFaceFW/fineweb",
493
+ repo_type="dataset",
494
+ local_dir="./fineweb/",
495
+ allow_patterns="data/CC-MAIN-2023-50/*")
496
+ ```
497
+
498
+ For faster downloads, make sure to install `pip install huggingface_hub[hf_transfer]` and set the environment variable `HF_HUB_ENABLE_HF_TRANSFER=1`.
499
+
500
+ ### Using `datasets`
501
+
502
+ ```python
503
+ from datasets import load_dataset
504
+ fw = load_dataset("HuggingFaceFW/fineweb", name="CC-MAIN-2024-10", split="train", streaming=True)
505
+ ```
506
+
507
+ ## Breakdown by dump/crawl
508
+
509
+ | Dump | Time period | Disk size (GB) | gpt2 tokens (billions) |
510
+ | --- | --- | --- | --- |
511
+ | CC-MAIN-2024-10 | February/March 2024 | 432.0 | 157.2 |
512
+ | CC-MAIN-2023-50 | November/December 2023 | 650.0 | 239.7 |
513
+ | CC-MAIN-2023-40 | September/October 2023 | 578.6 | 213.0 |
514
+ | CC-MAIN-2023-23 | May/June 2023 | 585.5 | 218.5 |
515
+ | CC-MAIN-2023-14 | March/April 2023 | 543.3 | 202.4 |
516
+ | CC-MAIN-2023-06 | January/February 2023 | 540.3 | 198.6 |
517
+ | CC-MAIN-2022-49 | November/December 2022 | 557.9 | 205.6 |
518
+ | CC-MAIN-2022-40 | September/October 2022 | 535.0 | 198.2 |
519
+ | CC-MAIN-2022-33 | August 2022 | 381.4 | 141.1 |
520
+ | CC-MAIN-2022-27 | June/July 2022 | 506.4 | 187.1 |
521
+ | CC-MAIN-2022-21 | May 2022 | 573.7 | 211.7 |
522
+ | CC-MAIN-2022-05 | January 2022 | 464.7 | 171.3 |
523
+ | CC-MAIN-2021-49 | November/December 2021 | 368.2 | 135.8 |
524
+ | CC-MAIN-2021-43 | October 2021 | 601.5 | 221.0 |
525
+ | CC-MAIN-2021-39 | September 2021 | 518.9 | 190.6 |
526
+ | CC-MAIN-2021-31 | July/August 2021 | 593.9 | 217.7 |
527
+ | CC-MAIN-2021-25 | June 2021 | 424.4 | 155.7 |
528
+ | CC-MAIN-2021-21 | May 2021 | 455.9 | 167.4 |
529
+ | CC-MAIN-2021-17 | April 2021 | 556.0 | 204.1 |
530
+ | CC-MAIN-2021-10 | February/March 2021 | 463.2 | 169.6 |
531
+ | CC-MAIN-2021-04 | January 2021 | 562.4 | 205.4 |
532
+ | CC-MAIN-2020-50 | November/December 2020 | 422.8 | 154.3 |
533
+ | CC-MAIN-2020-45 | October 2020 | 426.9 | 155.8 |
534
+ | CC-MAIN-2020-40 | September 2020 | 555.5 | 202.4 |
535
+ | CC-MAIN-2020-34 | August 2020 | 379.6 | 138.7 |
536
+ | CC-MAIN-2020-29 | July 2020 | 489.6 | 178.7 |
537
+ | CC-MAIN-2020-24 | May/June 2020 | 398.7 | 145.1 |
538
+ | CC-MAIN-2020-16 | March/April 2020 | 454.0 | 165.6 |
539
+ | CC-MAIN-2020-10 | February 2020 | 369.6 | 134.7 |
540
+ | CC-MAIN-2020-05 | January 2020 | 483.3 | 176.4 |
541
+ | CC-MAIN-2019-51 | December 2019 | 359.3 | 130.9 |
542
+ | CC-MAIN-2019-47 | November 2019 | 395.4 | 144.0 |
543
+ | CC-MAIN-2019-43 | October 2019 | 422.3 | 153.9 |
544
+ | CC-MAIN-2019-39 | September 2019 | 394.4 | 143.7 |
545
+ | CC-MAIN-2019-35 | August 2019 | 454.2 | 165.4 |
546
+ | CC-MAIN-2019-30 | July 2019 | 416.6 | 151.5 |
547
+ | CC-MAIN-2019-26 | June 2019 | 412.9 | 150.1 |
548
+ | CC-MAIN-2019-22 | May 2019 | 432.8 | 157.4 |
549
+ | CC-MAIN-2019-18 | April 2019 | 426.7 | 155.3 |
550
+ | CC-MAIN-2019-13 | March 2019 | 417.8 | 152.1 |
551
+ | CC-MAIN-2019-09 | February 2019 | 467.2 | 169.9 |
552
+ | CC-MAIN-2019-04 | January 2019 | 438.1 | 158.7 |
553
+ | CC-MAIN-2018-51 | December 2018 | 498.6 | 180.8 |
554
+ | CC-MAIN-2018-47 | November 2018 | 437.7 | 158.9 |
555
+ | CC-MAIN-2018-43 | October 2018 | 468.8 | 169.9 |
556
+ | CC-MAIN-2018-39 | September 2018 | 429.2 | 155.2 |
557
+ | CC-MAIN-2018-34 | August 2018 | 408.2 | 148.0 |
558
+ | CC-MAIN-2018-30 | July 2018 | 501.5 | 181.4 |
559
+ | CC-MAIN-2018-26 | June 2018 | 467.5 | 170.0 |
560
+ | CC-MAIN-2018-22 | May 2018 | 398.6 | 144.2 |
561
+ | CC-MAIN-2018-17 | April 2018 | 435.1 | 158.1 |
562
+ | CC-MAIN-2018-13 | March 2018 | 471.5 | 171.5 |
563
+ | CC-MAIN-2018-09 | February 2018 | 490.2 | 178.0 |
564
+ | CC-MAIN-2018-05 | January 2018 | 493.5 | 180.7 |
565
+ | CC-MAIN-2017-51 | December 2017 | 442.6 | 161.5 |
566
+ | CC-MAIN-2017-47 | November 2017 | 457.9 | 167.1 |
567
+ | CC-MAIN-2017-43 | October 2017 | 535.6 | 194.9 |
568
+ | CC-MAIN-2017-39 | September 2017 | 444.5 | 162.3 |
569
+ | CC-MAIN-2017-34 | August 2017 | 503.2 | 183.4 |
570
+ | CC-MAIN-2017-30 | July 2017 | 439.2 | 161.2 |
571
+ | CC-MAIN-2017-26 | June 2017 | 491.5 | 179.8 |
572
+ | CC-MAIN-2017-22 | May 2017 | 441.0 | 161.5 |
573
+ | CC-MAIN-2017-17 | April 2017 | 596.8 | 218.6 |
574
+ | CC-MAIN-2017-13 | March 2017 | 579.8 | 212.1 |
575
+ | CC-MAIN-2017-09 | February 2017 | 492.2 | 180.2 |
576
+ | CC-MAIN-2017-04 | January 2017 | 474.3 | 174.4 |
577
+ | CC-MAIN-2016-50 | December 2016 | 448.9 | 165.4 |
578
+ | CC-MAIN-2016-44 | October 2016 | 467.8 | 172.0 |
579
+ | CC-MAIN-2016-40 | September 2016 | 386.1 | 142.8 |
580
+ | CC-MAIN-2016-36 | August 2016 | 339.6 | 126.3 |
581
+ | CC-MAIN-2016-30 | July 2016 | 346.0 | 128.4 |
582
+ | CC-MAIN-2016-26 | June 2016 | 256.5 | 95.5 |
583
+ | CC-MAIN-2016-22 | May 2016 | 310.9 | 115.4 |
584
+ | CC-MAIN-2016-18 | April 2016 | 298.1 | 110.8 |
585
+ | CC-MAIN-2016-07 | February 2016 | 342.7 | 127.2 |
586
+ | CC-MAIN-2015-48 | November 2015 | 353.9 | 131.3 |
587
+ | CC-MAIN-2015-40 | September 2015 | 284.0 | 105.5 |
588
+ | CC-MAIN-2015-35 | August 2015 | 359.4 | 133.2 |
589
+ | CC-MAIN-2015-32 | July 2015 | 352.4 | 130.1 |
590
+ | CC-MAIN-2015-27 | June 2015 | 335.5 | 124.0 |
591
+ | CC-MAIN-2015-22 | May 2015 | 380.2 | 140.4 |
592
+ | CC-MAIN-2015-18 | April 2015 | 389.0 | 143.8 |
593
+ | CC-MAIN-2015-14 | March 2015 | 337.5 | 124.5 |
594
+ | CC-MAIN-2015-11 | February 2015 | 361.4 | 133.3 |
595
+ | CC-MAIN-2015-06 | January 2015 | 356.1 | 131.3 |
596
+ | CC-MAIN-2014-52 | December 2014 | 388.5 | 143.3 |
597
+ | CC-MAIN-2014-49 | November 2014 | 319.9 | 117.7 |
598
+ | CC-MAIN-2014-42 | October 2014 | 371.1 | 136.4 |
599
+ | CC-MAIN-2014-41 | September 2014 | 408.1 | 150.2 |
600
+ | CC-MAIN-2014-35 | August 2014 | 395.7 | 145.6 |
601
+ | CC-MAIN-2014-23 | July 2014 | 425.0 | 156.5 |
602
+ | CC-MAIN-2014-15 | April 2014 | 369.1 | 135.7 |
603
+ | CC-MAIN-2014-10 | March 2014 | 396.2 | 146.2 |
604
+ | CC-MAIN-2013-48 | Winter 2013 | 396.8 | 145.9 |
605
+ | CC-MAIN-2013-20 | Summer 2013 | 393.9 | 144.5 |
606
+ | Total | | 41880.2 | 15352.9 |
607
+
608
+ ## Dataset performance evaluation and ablations
609
+
610
+ We conducted our dataset performance ablations and evaluations by training a series of 1.8B parameters models on 27 billion tokens. To compare 🍷 FineWeb with other datasets, we also trained one of these 1.8B models per target dataset, on 350 billion tokens sampled from it (or the entire dataset when its size was < 350 billion tokens).
611
+
612
+ ### Hyper-parameters for ablation models
613
+
614
+ The detailed configurations for training the 1.8B parameters ablation model can be found here (link will be added soon).
615
+
616
+ ### Ablation evaluation benchmarks
617
+
618
+ To conduct the ablations for each of our dataset filtering choices, we selected a set of benchmarks which we identified as “high-signal” benchmarks. These benchmarks were selected according to the following criteria:
619
+
620
+ - small variance between runs trained on different samplings of the same dataset
621
+ - performance increasing monotically during training (or close)
622
+ - separation between runs on datasets of known quality (C4, The Pile, RedPajama) higher than the variance between runs with various modeling/data seeds
623
+
624
+ We used the following list of benchmark for our ablation runs:
625
+
626
+ - commonsense_qa (acc_norm)
627
+ - hellaswag (acc/acc_norm)
628
+ - openbookqa (acc/acc_norm)
629
+ - piqa (acc/acc_norm)
630
+ - siqa (acc/acc_norm)
631
+ - winogrande (acc/acc_norm)
632
+ - sciq (acc/acc_norm)
633
+ - arc (acc/acc_norm)
634
+ - mmlu (acc/acc_norm)
635
+
636
+ To compare runs we consider an aggregate score, the average of the scores for these tasks.
637
+
638
+ The prompts for all these benchmarks are formatted in order to compute and compare the log-likelihood of the full answers for each multiple choice question. All the implementation details for the benchmarks are available in `lighteval` [here](https://huggingface.co/datasets/HuggingFaceFW/fineweb/blob/main/lighteval_tasks.py).
639
+
640
+ ### Comparison with other datasets
641
+
642
+ We compared 🍷 FineWeb with the following datasets:
643
+
644
+ - [RefinedWeb](https://huggingface.co/datasets/tiiuae/falcon-refinedweb)
645
+ - [C4](https://huggingface.co/datasets/allenai/c4)
646
+ - [Dolma v1.6](https://huggingface.co/datasets/allenai/dolma) (the CommonCrawl part)
647
+ - [The Pile](https://huggingface.co/datasets/EleutherAI/pile)
648
+ - [SlimPajama](https://huggingface.co/datasets/cerebras/SlimPajama-627B)
649
+ - [RedPajama2](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-V2) (deduplicated)
650
+
651
+ You will find these models on [this collection](https://huggingface.co/collections/HuggingFaceFW/ablation-models-662457b0d213e8c14fe47f32). We have uploaded checkpoints at every 1000 training steps. You will also find our full [evaluation results here](https://huggingface.co/datasets/HuggingFaceFW/fineweb/blob/main/eval_results.csv).
652
+
653
+ <center>
654
+ <img src="https://huggingface.co/datasets/HuggingFaceFW/admin/resolve/main/fineweb-ablations.png" alt="ablations">
655
+ </center>
656
+
657
+ _Note:_ The plot is smoothed by averaging 5k steps in a rolling window.
658
+
659
+ # Dataset card for 🍷 FineWeb
660
+
661
+ ## Dataset Description
662
+
663
+ - **Homepage and Repository:** [https://huggingface.co/datasets/HuggingFaceFW/fineweb](https://huggingface.co/datasets/HuggingFaceFW/fineweb)
664
+ - **Point of Contact:** please create a discussion on the Community tab
665
+ - **License:** Open Data Commons Attribution License (ODC-By) v1.0
666
+
667
+ ### Dataset Summary
668
+
669
+ This dataset was created by processing 95 [CommonCrawl](https://commoncrawl.org/) dumps comprising web data crawled from the summer of 2013 to March of 2024. 🍷 FineWeb includes a variety of domains and topics in English and is primarily intended to be used as a research artifact on public data in the context of pretraining dataset for large language models. The CommonCrawl data was carefully processed, filtered and deduplicated with the 🏭 [`datatrove`](https://github.com/huggingface/datatrove/) library, resulting in the largest publicly available clean LLM pretraining dataset, counting around 15 trillion tokens (gpt2 tokenizer).
670
+
671
+ ## Dataset Structure
672
+
673
+ ### Data Instances
674
+
675
+ The following is an example sample from the dataset. It is part of the `CC-MAIN-2021-43` and was crawled on `2021-10-15T21:20:12Z`.
676
+
677
+ ```json
678
+ {
679
+ "text": "This is basically a peanut flavoured cream thickened with egg yolks and then set into a ramekin on top of some jam. Tony, one of the Wedgwood chefs, suggested sprinkling on some toasted crushed peanuts at the end to create extra crunch, which I thought was a great idea. The result is excellent.",
680
+ "id": "<urn:uuid:e5a3e79a-13d4-4147-a26e-167536fcac5d>",
681
+ "dump": "CC-MAIN-2021-43",
682
+ "url": "<http://allrecipes.co.uk/recipe/24758/peanut-butter-and-jam-creme-brulee.aspx?o_is=SimilarRecipes&o_ln=SimRecipes_Photo_7>",
683
+ "date": "2021-10-15T21:20:12Z",
684
+ "file_path": "s3://commoncrawl/crawl-data/CC-MAIN-2021-43/segments/1634323583083.92/warc/CC-MAIN-20211015192439-20211015222439-00600.warc.gz",
685
+ "language": "en",
686
+ "language_score": 0.948729,
687
+ "token_count": 69
688
+ }
689
+ ```
690
+
691
+ ### Data Fields
692
+
693
+ - `text` (string): the main text content
694
+ - `id` (string): original unique identifier for this sample from CommonCrawl
695
+ - `dump` (string): the CommonCrawl dump this sample was a part of
696
+ - `url` (string): url to the original page where `text` was present
697
+ - `date` (string): crawl date (from CommonCrawl)
698
+ - `file_path` (string): s3 path for the individual CommonCrawl warc file containing this sample
699
+ - `language` (string): `en` for all the samples in this dataset
700
+ - `language_score` (float): language prediction score (`0.01.0`) as reported by the [fastText language classifier](https://github.com/huggingface/datatrove/blob/main/src/datatrove/pipeline/filters/language_filter.py)
701
+ - `token_count` (int): number of tokens when applying the `gpt2` tokenizer to this sample
702
+
703
+ ### Data Splits
704
+
705
+ The `default` subset includes the entire dataset. If you would like to only use the data from a particular [CommonCrawl dump](https://commoncrawl.org/overview), you can use the dump name as a subset. You will find the full list of available dumps on the table above.
706
+ From experiments we have run, not all dumps give the same performance. For relatively small trainings (<400 billion tokens) we recommend using the recent `CC-MAIN-2023-50` and `CC-MAIN-2024-10`. We will share more details about these experiments on our future technical report.
707
+
708
+ ## Dataset Creation
709
+
710
+ ### Curation Rationale
711
+
712
+ While multiple open-weights models have regularly been released in recent months, these releases often do not include the model's training data. With 🍷 FineWeb we aim to provide the open source community with a very large clean pretraining dataset that can be used to push the envelope on truly open source models (open source models where data is also released).
713
+
714
+ ### Source Data
715
+
716
+ The source data consists of webpages crawled by the CommonCrawl foundation over the 2013-2024 time period.
717
+
718
+ We then extracted the main page text from the html of each webpage, carefully filtered each sample and deduplicated each individual CommonCrawl dump/crawl.
719
+
720
+ While we originally intended to deduplicate the dataset as a whole, our ablations showed that training on a sampling of individually deduplicated dumps/crawls outperformed training on a sampling of all the dumps/crawls deduplicated together. We will discuss this further in our technical report.
721
+
722
+ ### Data processing steps
723
+
724
+ We used the 🏭 `datatrove` library to process the data.
725
+ You can find a **working script** that launches the [entire processing pipeline here](https://github.com/huggingface/datatrove/blob/main/examples/fineweb.py).
726
+
727
+ The data processing pipeline consists of:
728
+
729
+ 1. [Url Filtering](https://github.com/huggingface/datatrove/blob/9a88bebc86a554f8521faa70b12ad4fa0c227537/src/datatrove/pipeline/filters/url_filter.py), removing documents originating from Malicious and NSFW websites, using both block-list as well as subwords detection
730
+ 2. [Trafilatura](https://github.com/huggingface/datatrove/blob/9a88bebc86a554f8521faa70b12ad4fa0c227537/src/datatrove/pipeline/extractors/trafilatura.py) text extraction on the raw HTML from CommonCrawl’s warc files
731
+ 3. [FastText LanguageFilter](https://github.com/huggingface/datatrove/blob/9a88bebc86a554f8521faa70b12ad4fa0c227537/src/datatrove/pipeline/filters/language_filter.py), removing any document with `en` language score lower than **0.65**
732
+ 4. Quality filtering
733
+ 1. [Gopher Repetition /](https://github.com/huggingface/datatrove/blob/9a88bebc86a554f8521faa70b12ad4fa0c227537/src/datatrove/pipeline/filters/gopher_repetition_filter.py) [Quality](https://github.com/huggingface/datatrove/blob/9a88bebc86a554f8521faa70b12ad4fa0c227537/src/datatrove/pipeline/filters/gopher_quality_filter.py)
734
+ 2. [C4 Quality filters](https://github.com/huggingface/datatrove/blob/9a88bebc86a554f8521faa70b12ad4fa0c227537/src/datatrove/pipeline/filters/c4_quality_filter.py) except `terminal_punct` rule
735
+ 3. [FineWeb custom filters](https://github.com/huggingface/datatrove/blob/05194d3960741e7d5c0bd0d6dd69d44514622549/src/datatrove/pipeline/filters/fineweb_quality_filter.py), consisting of heuristics for removing list-like documents, documents with repeated lines and documents with likely wrong line formatting.
736
+ 5. [MinHash deduplication](https://github.com/huggingface/datatrove/blob/6daa5e879e06b21e6886b37e2b1be4ae58a658b6/src/datatrove/pipeline/dedup/minhash.py) with each crawl deduplicated individually (5-grams, 14x8 hash functions)
737
+ 6. [PII Formatting](https://github.com/huggingface/datatrove/blob/main/src/datatrove/pipeline/formatters/pii.py) to anonymize email and public IP addresses
738
+
739
+ ### Annotations
740
+
741
+ We augment the original samples with the `language`, `language_score` and `token_count` annotations. The language related annotations are automatically generated by our [language filter](https://github.com/huggingface/datatrove/blob/main/src/datatrove/pipeline/filters/language_filter.py). `token_count` is generated by [applying the gpt2 tokenizer](https://github.com/huggingface/datatrove/blob/main/src/datatrove/pipeline/tokens/counter.py) to the `text` column.
742
+
743
+ ### Personal and Sensitive Information
744
+
745
+ We anonymize email addresses and public IP addresses.
746
+
747
+ For emails, we apply a regex pattern and replace any occurrence of an email address with either `email@example.com` or `firstname.lastname@example.org`. For IP addresses, we also employ a regex pattern and then further filter to only anonymize IP addresses [allocated for public networks](https://www.iana.org/assignments/iana-ipv4-special-registry/iana-ipv4-special-registry.xhtml). Matched IP addresses are then replaced with one of the following randomly generated IP addresses, which at the time of dataset creation were not responding to ping requests: `22.214.171.124`, `126.96.36.199`, `188.8.131.52`, `184.108.40.206`, `220.127.116.11`, and `18.104.22.168`. We decided against applying regex patterns for phone numbers due to the high false positive rate.
748
+
749
+ Despite our efforts, given that 🍷 FineWeb is sourced from the internet at large, it is very likely that some personable identifiable information (PII) will be present. If you find your own PII in 🍷 FineWeb and would like it removed, please fill out our [PII removal form](https://forms.gle/VyNT3ZAUPZjPuWp39).
750
+
751
+ ## Considerations for Using the Data
752
+
753
+ ### Social Impact of Dataset
754
+
755
+ With the release of this dataset we aim to make model training more accessible to the machine learning community at large.
756
+
757
+ While multiple open-weights models with strong performance have been publicly released in the past, more often than not these releases are not accompanied by the corresponding training dataset. This is unfortunate as the dataset specificities and characteristics have been demonstrated to have a very large impact and role in the performances of the models. As the creation of a high quality training dataset is a fundamental requirement to training an LLM capable of excelling at downstream tasks, with 🍷 FineWeb we (a) not only make the dataset creation process more transparent, by sharing our entire processing setup including the codebase used, we also (b) help alleviate the costs of dataset curation, both in time and in compute, for model creators by publicly releasing our dataset with the community.
758
+
759
+ ### Discussion of Biases
760
+
761
+ Efforts were made to minimize the amount of NSFW and toxic content present in the dataset by employing filtering on the URL level. However, there are still a significant number of documents present in the final dataset that could be considered toxic or contain harmful content. As 🍷 FineWeb was sourced from the web as a whole, any harmful biases typically present in it may be reproduced on our dataset.
762
+
763
+ We deliberately avoided using machine learning filtering methods that define text quality based on the similarity to a “gold” source such as wikipedia or toxicity classifiers as these methods have been known to [disproportionately remove content in specific dialects](https://aclanthology.org/D16-1120/) and [overclassify as toxic text related to specific social identities](https://arxiv.org/pdf/2109.07445.pdf), respectively.
764
+
765
+ ### Other Known Limitations
766
+
767
+ As a consequence of some of the filtering steps applied, it is likely that code content is not prevalent in our dataset. If you are training a model that should also perform code tasks, we recommend you use 🍷 FineWeb with a code dataset, such as [The Stack v2](https://huggingface.co/datasets/bigcode/the-stack-v2). You should also probably consider complementing 🍷 FineWeb with specialized curated sources (such as Wikipedia, for example) as they will likely have better formatting than the wikipedia content included in 🍷 FineWeb (we did not tailor the processing to individual websites).
768
+
769
+ ## Additional Information
770
+
771
+ ### Licensing Information
772
+
773
+ The dataset is released under the **Open Data Commons Attribution License (ODC-By) v1.0** [license](https://opendatacommons.org/licenses/by/1-0/). The use of this dataset is also subject to [CommonCrawl's Terms of Use](https://commoncrawl.org/terms-of-use).
774
+
775
+ ### Future work
776
+
777
+ We plan to not only continue but also expand our efforts to create open-source high quality training datasets and to improve 🍷 FineWeb itself in future iterations.
778
+
779
+ ### Citation Information
780
+
781
+ ```
782
+ @software{penedo2024fineweb,
783
+ author = {Penedo, Guilherme and Kydlíček, Hynek and von Werra, Leandro and Wolf, Thomas},
784
+ title = {FineWeb},
785
+ month = April,
786
+ year = 2024,
787
+ doi = { 10.57967/hf/2092 },
788
+ url = {https://huggingface.co/datasets/HuggingFaceFW/fineweb}
789
+ }
790
+
791
+ ```
data/CC-MAIN-2013-20/000_00000.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6063b85cf7fc07da2738d79548d854ae5e5b1cbfb5bc9f86db6154207cb5cf14
3
+ size 2153049317
data/CC-MAIN-2013-20/000_00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f20489329c6dc9ff991cd0b4e48319d39b8c7af82e2295b7a64dc16c441faa8
3
+ size 2152352172
data/CC-MAIN-2013-20/000_00002.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8652abce174cd18991c104a951aaa5eca0a81fafe2195cecb72a1ef29d84c17d
3
+ size 2152579413
data/CC-MAIN-2013-20/000_00003.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76aee1cd6687eb964c39753fa539be729ac5eebd124a38ecf6e9ad450c5f056d
3
+ size 2152834053
data/CC-MAIN-2013-20/000_00004.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05239e019b77598cab6bfec69d6cac77d4c2d77c54a28020f4be43dba8d97386
3
+ size 2152308872
data/CC-MAIN-2013-20/000_00005.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2e314bdc2e7970869c3904844cd0b2cc7b9d8b0539408307016a3ec2b7bc233
3
+ size 2153618905
data/CC-MAIN-2013-20/000_00006.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09a84594749e8a91e1f4d1b221ce498661d02a349b309235d968dc3b9cdb8a03
3
+ size 2152677608
data/CC-MAIN-2013-20/000_00007.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e205844b00104491a14f1ba60ba0c45896af1997d2cf8c6d915ae576a0fbb960
3
+ size 2152443532
data/CC-MAIN-2013-20/000_00008.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5deec389e0241f8f44e99cd64871458dcca4bedb29fb5bb435d459d14a72cae9
3
+ size 2152822112
data/CC-MAIN-2013-20/000_00009.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a9f6d555823f50528a8c01e811376c5bde15165b7fa76c64b7d76f67098f408
3
+ size 2152505238
data/CC-MAIN-2013-20/000_00010.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:193a9305dbce20bc681cffc8092708fd3d806a0c2a3f675402918731b97c6f87
3
+ size 2153230706
data/CC-MAIN-2013-20/000_00011.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8270060f3391f56c17c66177bedd1d279155d33ed02c102ced4e29bbc8da066d
3
+ size 2152133786
data/CC-MAIN-2013-20/000_00012.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e23ed30f6b00574263eed42eba07c7e3b63b720db5c8f3542f9c89fb0e684f9c
3
+ size 2153465123
data/CC-MAIN-2013-20/000_00013.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd332af4a35ff8fc40ee929685b44e3a7e259117de55ee0b347e8e1c895c4982
3
+ size 2153294819
data/CC-MAIN-2013-20/000_00014.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d025b9b352c1b37eed9cb687e4daccbaa32d4f9417c4edc579777ceeae7a726
3
+ size 2153395190
data/CC-MAIN-2013-20/000_00015.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1de7baaa4b3d027e17a0fe6dd6810dc7e73731d6f3b40131e3dfb57eb33c3597
3
+ size 2153114991
data/CC-MAIN-2013-20/000_00016.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d0640f955d98433ec03da6e90df74126f4a8ed458a6292c3c9dac792f42a6f8
3
+ size 2153649696
data/CC-MAIN-2013-20/000_00017.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c007f47f5a4136a50e60bb4c9e873fd17ee38e9133683e678bb600e031496970
3
+ size 2151641186
data/CC-MAIN-2013-20/000_00018.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0d178ee5e517a2eeb5a07489add8557624f367c09be9c638943711bdde73b74
3
+ size 2153046641
data/CC-MAIN-2013-20/000_00019.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24c96731c4fcefd50c0fdd1409fe50c07cf1eca913dfc133ae715e46ba8a5bba
3
+ size 2153787321
data/CC-MAIN-2013-20/000_00020.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:949d3f547ee1e7fc8e9a4363a3d51cc81c350cc00276b7966594c0e6d39509ff
3
+ size 2152240258
data/CC-MAIN-2013-20/000_00021.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:66db7657140ea0d604fcd8d77ea2a0f270f0599e64998e5a378a217131cca6cb
3
+ size 2153422963
data/CC-MAIN-2013-20/000_00022.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53b4dcfe4d12807ff27664df34616ac3e188f2132e38c0063f34e61229b95712
3
+ size 2152485625
data/CC-MAIN-2013-20/000_00023.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44fdca1d9fa68296f92597024efee7bec3a58ee411b890e8cc5869bd280ca828
3
+ size 2152118981
data/CC-MAIN-2013-20/000_00024.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c27fa1c41c3a40d4c14723117d8bc43d1c9131a9cc604eaefcba1e672b95a58e
3
+ size 2152714230
data/CC-MAIN-2013-20/000_00025.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56febb9b793aef9757941cf1c6ebfafc49876d523bf0fd2299021479e5d0f4a2
3
+ size 2153179929
data/CC-MAIN-2013-20/000_00026.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c53a78702b9c59d1dccfbf4b7967723afa787d173d814372df344348ec9fd7cf
3
+ size 2153129659
data/CC-MAIN-2013-20/000_00027.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e0510515e901ac8eca153208118ba4c0d944eb8abedf870489d68d8f6296e20
3
+ size 2153202171
data/CC-MAIN-2013-20/000_00028.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4716f2dc1a4ee378fa1b1c06f01e481d3475a3c47ce9ba3dda975e3af220ece
3
+ size 2152513020
data/CC-MAIN-2013-20/000_00029.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60eae9221f4f6a9bb6cce53d9f521997793a27ea0807b7df2a64f1e921c6ff85
3
+ size 2153361786
data/CC-MAIN-2013-20/000_00030.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96c25accc2b0b685ce9e257391d28a39d4c2af1c0abafbb05c68ce3d5fa8e376
3
+ size 2151750143
data/CC-MAIN-2013-20/000_00031.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77581f1b89cb53a003755446489baf0cd864e8884d6c610eb92a493142d764c4
3
+ size 2153118159
data/CC-MAIN-2013-20/000_00032.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22111af43cbd3005c28f956677576ae7dab0c32c38a90ea79d2e1d556e625d2c
3
+ size 2153490901
data/CC-MAIN-2013-20/000_00033.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:411606db21cc3808827f215db11b0e27250887fbf99586e1e2713c2094f905da
3
+ size 2151978349
data/CC-MAIN-2013-20/000_00034.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9d53a9e4c6867e4d459ba880a254bea18b4d4ed01d86ecd7b831193cb5c008c
3
+ size 2153590858
data/CC-MAIN-2013-20/000_00035.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f2e1da59d99f3bdab9d5306b644ff10ec607ed64dc918bcbcbcda4deeed3bba
3
+ size 2153350687
data/CC-MAIN-2013-20/000_00036.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:846ba8f84afeb5b60cba49c9d3e14fe5d63932a1ffbadf3eb2d46cba2332d7e8
3
+ size 2152816506
data/CC-MAIN-2013-20/000_00037.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:836d7bb632dc9d3f242b501cb25fd93b7f0a6b173fd72070c19ca994daf760c6
3
+ size 2153230184
data/CC-MAIN-2013-20/000_00038.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73619038de855595c64a527b747ac408066e71cdb05d79362bc784b82dee4d14
3
+ size 2153159508
data/CC-MAIN-2013-20/000_00039.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a13e92921a9b1627105fc1260d612a1cf5829e95186bb5836109211287d7d717
3
+ size 2151758624
data/CC-MAIN-2013-20/000_00040.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa79a363cfe69883b69d8b2ca67d01e91414f427d813769e5453a8d113c66847
3
+ size 2153156146
data/CC-MAIN-2013-20/000_00041.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbd87a823cc59c388bc95c0c0bbc050289a1f0e16d83211387c3434980aa66f2
3
+ size 2152264337
data/CC-MAIN-2013-20/000_00042.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc7cdba1d58e94b67722277f7e07fa342616ac863bd3293d7760f584afc90f81
3
+ size 2152518387
data/CC-MAIN-2013-20/000_00043.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:508395052f186aaeb043c41a994d42bfda91c6925b2c85d8b2a5b1ae2ceb96d1
3
+ size 2151793720
data/CC-MAIN-2013-20/000_00044.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f52e468a44464688f695d4b124c3097563a41bee7c00f0bc64de79a63c720c87
3
+ size 2151714888
data/CC-MAIN-2013-20/000_00045.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:413bd3ea084e38fce4a44c0495534db3a0d2d3dcbe693ef6156fd4c6ac224c3e
3
+ size 2152286781
data/CC-MAIN-2013-20/000_00046.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc718af8aa92a72b2289dafb7ac2ba6424df61a6b99853cc60a25fb2982fe7af
3
+ size 2152413303
data/CC-MAIN-2013-20/000_00047.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5de602b0499c73f0328fb170dd13d155982856aad66d4362f6c484112ac205d
3
+ size 2151941578