Hugging Face's logo

Datasets:
ai4bharat
/
sangraha

Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
Assamese
Bengali
Gujarati
Size:
100M - 1B
ArXiv:
Tags:
language-modeling
casual-lm
llm
Libraries:
Datasets
Dask
License:
Dataset card Viewer Files Community
8
safikhan commited on
Commit
4b1d822
·
verified ·
1 Parent(s): ac797c6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +148 -135
README.md CHANGED
@@ -21,144 +21,145 @@ language:
21
  - te
22
  - ur
23
  tags:
24
- - language-modeling
25
- - casual-lm
26
- - llm
27
  pretty_name: sangraha
28
  dataset_info:
29
- - config_name: verified
30
- features:
31
- - name: doc_id
32
- dtype: string
33
- - name: type
34
- dtype: string
35
- - name: text
36
- dtype: string
37
- splits:
38
- - name: asm
39
- - name: ben
40
- - name: brx
41
- - name: doi
42
- - name: eng
43
- - name: gom
44
- - name: guj
45
- - name: hin
46
- - name: kan
47
- - name: kas
48
- - name: mai
49
- - name: mal
50
- - name: mar
51
- - name: mni
52
- - name: nep
53
- - name: ori
54
- - name: pan
55
- - name: san
56
- - name: sat
57
- - name: snd
58
- - name: tam
59
- - name: tel
60
- - name: urd
61
- - config_name: unverified
62
- features:
63
- - name: doc_id
64
- dtype: string
65
- - name: text
66
- dtype: string
67
- splits:
68
- - name: asm
69
- - name: ben
70
- - name: guj
71
- - name: hin
72
- - name: kan
73
- - name: mal
74
- - name: mar
75
- - name: nep
76
- - name: ori
77
- - name: pan
78
- - name: san
79
- - name: tam
80
- - name: tel
81
- - name: urd
82
-
83
  configs:
84
- - config_name: verified
85
- data_files:
86
- - split: asm
87
- path: verified/asm.*
88
- - split: ben
89
- path: verified/ben.*
90
- - split: brx
91
- path: verified/brx.*
92
- - split: doi
93
- path: verified/doi.*
94
- - split: eng
95
- path: verified/eng.*
96
- - split: gom
97
- path: verified/gom.*
98
- - split: guj
99
- path: verified/guj.*
100
- - split: hin
101
- path: verified/hin.*
102
- - split: kan
103
- path: verified/kan.*
104
- - split: kas
105
- path: verified/kas.*
106
- - split: mai
107
- path: verified/mai.*
108
- - split: mal
109
- path: verified/mal.*
110
- - split: mar
111
- path: verified/mar.*
112
- - split: mni
113
- path: verified/mni.*
114
- - split: nep
115
- path: verified/nep.*
116
- - split: ori
117
- path: verified/ori.*
118
- - split: pan
119
- path: verified/pan.*
120
- - split: san
121
- path: verified/san.*
122
- - split: sat
123
- path: verified/sat.*
124
- - split: snd
125
- path: verified/snd.*
126
- - split: tam
127
- path: verified/tam.*
128
- - split: tel
129
- path: verified/tel.*
130
- - split: urd
131
- path: verified/urd.*
132
- - config_name: unverified
133
- data_files:
134
- - split: asm
135
- path: unverified/asm.*
136
- - split: ben
137
- path: unverified/ben.*
138
- - split: guj
139
- path: unverified/guj.*
140
- - split: hin
141
- path: unverified/hin.*
142
- - split: kan
143
- path: unverified/kan.*
144
- - split: mal
145
- path: unverified/mal.*
146
- - split: mar
147
- path: unverified/mar.*
148
- - split: nep
149
- path: unverified/nep.*
150
- - split: ori
151
- path: unverified/ori.*
152
- - split: pan
153
- path: unverified/pan.*
154
- - split: san
155
- path: unverified/san.*
156
- - split: tam
157
- path: unverified/tam.*
158
- - split: tel
159
- path: unverified/tel.*
160
- - split: urd
161
- path: unverified/urd.*
 
 
162
  ---
163
 
164
  # Sangraha
@@ -221,3 +222,15 @@ Sangraha contains three broad components:
221
  | **Total** | **64,306.1** | **162,707.9** | **24,307.7** | **251,321.0** |
222
 
223
 
 
 
 
 
 
 
 
 
 
 
 
 
 
21
  - te
22
  - ur
23
  tags:
24
+ - language-modeling
25
+ - casual-lm
26
+ - llm
27
  pretty_name: sangraha
28
  dataset_info:
29
+ - config_name: verified
30
+ features:
31
+ - name: doc_id
32
+ dtype: string
33
+ - name: type
34
+ dtype: string
35
+ - name: text
36
+ dtype: string
37
+ splits:
38
+ - name: asm
39
+ - name: ben
40
+ - name: brx
41
+ - name: doi
42
+ - name: eng
43
+ - name: gom
44
+ - name: guj
45
+ - name: hin
46
+ - name: kan
47
+ - name: kas
48
+ - name: mai
49
+ - name: mal
50
+ - name: mar
51
+ - name: mni
52
+ - name: nep
53
+ - name: ori
54
+ - name: pan
55
+ - name: san
56
+ - name: sat
57
+ - name: snd
58
+ - name: tam
59
+ - name: tel
60
+ - name: urd
61
+ - config_name: unverified
62
+ features:
63
+ - name: doc_id
64
+ dtype: string
65
+ - name: text
66
+ dtype: string
67
+ splits:
68
+ - name: asm
69
+ - name: ben
70
+ - name: guj
71
+ - name: hin
72
+ - name: kan
73
+ - name: mal
74
+ - name: mar
75
+ - name: nep
76
+ - name: ori
77
+ - name: pan
78
+ - name: san
79
+ - name: tam
80
+ - name: tel
81
+ - name: urd
 
82
  configs:
83
+ - config_name: verified
84
+ data_files:
85
+ - split: asm
86
+ path: verified/asm.*
87
+ - split: ben
88
+ path: verified/ben.*
89
+ - split: brx
90
+ path: verified/brx.*
91
+ - split: doi
92
+ path: verified/doi.*
93
+ - split: eng
94
+ path: verified/eng.*
95
+ - split: gom
96
+ path: verified/gom.*
97
+ - split: guj
98
+ path: verified/guj.*
99
+ - split: hin
100
+ path: verified/hin.*
101
+ - split: kan
102
+ path: verified/kan.*
103
+ - split: kas
104
+ path: verified/kas.*
105
+ - split: mai
106
+ path: verified/mai.*
107
+ - split: mal
108
+ path: verified/mal.*
109
+ - split: mar
110
+ path: verified/mar.*
111
+ - split: mni
112
+ path: verified/mni.*
113
+ - split: nep
114
+ path: verified/nep.*
115
+ - split: ori
116
+ path: verified/ori.*
117
+ - split: pan
118
+ path: verified/pan.*
119
+ - split: san
120
+ path: verified/san.*
121
+ - split: sat
122
+ path: verified/sat.*
123
+ - split: snd
124
+ path: verified/snd.*
125
+ - split: tam
126
+ path: verified/tam.*
127
+ - split: tel
128
+ path: verified/tel.*
129
+ - split: urd
130
+ path: verified/urd.*
131
+ - config_name: unverified
132
+ data_files:
133
+ - split: asm
134
+ path: unverified/asm.*
135
+ - split: ben
136
+ path: unverified/ben.*
137
+ - split: guj
138
+ path: unverified/guj.*
139
+ - split: hin
140
+ path: unverified/hin.*
141
+ - split: kan
142
+ path: unverified/kan.*
143
+ - split: mal
144
+ path: unverified/mal.*
145
+ - split: mar
146
+ path: unverified/mar.*
147
+ - split: nep
148
+ path: unverified/nep.*
149
+ - split: ori
150
+ path: unverified/ori.*
151
+ - split: pan
152
+ path: unverified/pan.*
153
+ - split: san
154
+ path: unverified/san.*
155
+ - split: tam
156
+ path: unverified/tam.*
157
+ - split: tel
158
+ path: unverified/tel.*
159
+ - split: urd
160
+ path: unverified/urd.*
161
+ size_categories:
162
+ - 100B<n<1T
163
  ---
164
 
165
  # Sangraha
 
222
  | **Total** | **64,306.1** | **162,707.9** | **24,307.7** | **251,321.0** |
223
 
224
 
225
+ To cite Sangraha, please use:
226
+
227
+ ```
228
+ @misc{cerebras2023slimpajama,
229
+ author = {Soboleva, Daria and Al-Khateeb, Faisal and Myers, Robert and Steeves, Jacob R and Hestness, Joel and Dey, Nolan},
230
+ title = {{SlimPajama: A 627B token cleaned and deduplicated version of RedPajama}},
231
+ month = June,
232
+ year = 2023,
233
+ howpublished = {\url{https://www.cerebras.net/blog/slimpajama-a-627b-token-cleaned-and-deduplicated-version-of-redpajama}},
234
+ url = {https://huggingface.co/datasets/cerebras/SlimPajama-627B},
235
+ }
236
+ ```