Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
pretty_name: Speed-5B-mX-parquet
|
| 3 |
+
task_categories:
|
| 4 |
+
- text-generation
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# Speed-5B-mX (compressed)
|
| 10 |
+
|
| 11 |
+
Zstd-compressed Parquet copy of `Shrijanagain/Speed-5B-mX`.
|
| 12 |
+
|
| 13 |
+
## Totals
|
| 14 |
+
|
| 15 |
+
- Rows: 222,987
|
| 16 |
+
- Exact tokens (`add_special_tokens=False`): 147,439,498
|
| 17 |
+
- Source JSONL bytes: 669.93 MiB
|
| 18 |
+
- Parquet bytes: 241.63 MiB
|
| 19 |
+
- Compression ratio: 2.773x
|
| 20 |
+
- Tokenizer: `Shrijanagain/SPEED2-V0-0B7-CPT-SFT/tokenizer.json`
|
| 21 |
+
|
| 22 |
+
## Schema
|
| 23 |
+
|
| 24 |
+
| column | type | notes |
|
| 25 |
+
| --- | --- | --- |
|
| 26 |
+
| `text` | large_string | document text |
|
| 27 |
+
| `source` | string | mixture source id |
|
| 28 |
+
| `tokens` | int32 | exact token count of `text` |
|
| 29 |
+
|
| 30 |
+
## Tokens per source
|
| 31 |
+
|
| 32 |
+
| source | rows | tokens | share |
|
| 33 |
+
| --- | --- | --- | --- |
|
| 34 |
+
| fineweb | 222,987 | 147,439,498 | 100.00% |
|
| 35 |
+
|
| 36 |
+
## Shards
|
| 37 |
+
|
| 38 |
+
| source shard | rows | tokens | parquet size |
|
| 39 |
+
| --- | --- | --- | --- |
|
| 40 |
+
| `chunks_6b/mixed_6b_chunk_000000.jsonl` | 222,987 | 147,439,498 | 0.236 GiB |
|
| 41 |
+
|
| 42 |
+
## Audit
|
| 43 |
+
|
| 44 |
+
- Record fields seen: `{"source,text": 222987}`
|
| 45 |
+
- Skipped rows: `{}`
|
| 46 |
+
|
| 47 |
+
Use `streaming=True` or read row groups directly; each file is
|
| 48 |
+
independently readable.
|