Hugging Face's logo

Datasets:
AI-basic-class
/
shs

Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
1M - 10M
Libraries:
Datasets
Dask
Polars
Dataset card Data Studio Files
xet
 Community
Shrijanagain commited on
Commit
45ff1e9
·
verified ·
1 Parent(s): 15a8029

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +11 -9
README.md CHANGED
@@ -12,11 +12,11 @@ Zstd-compressed Parquet copy of `Shrijanagain/Speed-5B-mX`.
12
 
13
  ## Totals
14
 
15
- - Rows: 2,357,537
16
- - Exact tokens (`add_special_tokens=False`): 2,212,023,432
17
- - Source JSONL bytes: 8.80 GiB
18
- - Parquet bytes: 2.73 GiB
19
- - Compression ratio: 3.222x
20
  - Tokenizer: `Shrijanagain/SPEED2-V0-0B7-CPT-SFT/tokenizer.json`
21
 
22
  ## Schema
@@ -31,9 +31,10 @@ Zstd-compressed Parquet copy of `Shrijanagain/Speed-5B-mX`.
31
 
32
  | source | rows | tokens | share |
33
  | --- | --- | --- | --- |
34
- | code_stack | 599,671 | 828,986,469 | 37.48% |
35
- | fineweb_edu | 709,285 | 695,213,798 | 31.43% |
36
- | fineweb | 1,048,581 | 687,823,165 | 31.09% |
 
37
 
38
  ## Shards
39
 
@@ -51,10 +52,11 @@ Zstd-compressed Parquet copy of `Shrijanagain/Speed-5B-mX`.
51
  | `chunks_6b/mixed_6b_chunk_000009.jsonl` | 149,047 | 146,289,311 | 0.230 GiB |
52
  | `chunks_6b/mixed_6b_chunk_000010.jsonl` | 148,761 | 146,330,597 | 0.230 GiB |
53
  | `chunks_6b/mixed_6b_chunk_000011.jsonl` | 148,315 | 146,337,489 | 0.230 GiB |
 
54
 
55
  ## Audit
56
 
57
- - Record fields seen: `{"source,text": 2357538}`
58
  - Skipped rows: `{"too_long": 1}`
59
 
60
  Use `streaming=True` or read row groups directly; each file is
 
12
 
13
  ## Totals
14
 
15
+ - Rows: 2,486,559
16
+ - Exact tokens (`add_special_tokens=False`): 2,402,391,917
17
+ - Source JSONL bytes: 9.48 GiB
18
+ - Parquet bytes: 2.94 GiB
19
+ - Compression ratio: 3.227x
20
  - Tokenizer: `Shrijanagain/SPEED2-V0-0B7-CPT-SFT/tokenizer.json`
21
 
22
  ## Schema
 
31
 
32
  | source | rows | tokens | share |
33
  | --- | --- | --- | --- |
34
+ | code_stack | 599,671 | 828,986,469 | 34.51% |
35
+ | fineweb_edu | 726,000 | 711,089,153 | 29.60% |
36
+ | fineweb | 1,048,581 | 687,823,165 | 28.63% |
37
+ | finemath | 112,307 | 174,493,130 | 7.26% |
38
 
39
  ## Shards
40
 
 
52
  | `chunks_6b/mixed_6b_chunk_000009.jsonl` | 149,047 | 146,289,311 | 0.230 GiB |
53
  | `chunks_6b/mixed_6b_chunk_000010.jsonl` | 148,761 | 146,330,597 | 0.230 GiB |
54
  | `chunks_6b/mixed_6b_chunk_000011.jsonl` | 148,315 | 146,337,489 | 0.230 GiB |
55
+ | `chunks_6b/mixed_6b_chunk_000012.jsonl` | 129,022 | 190,368,485 | 0.206 GiB |
56
 
57
  ## Audit
58
 
59
+ - Record fields seen: `{"source,text": 2486560}`
60
  - Skipped rows: `{"too_long": 1}`
61
 
62
  Use `streaming=True` or read row groups directly; each file is