Update README.md
Browse files
README.md
CHANGED
|
@@ -28,3 +28,15 @@ This dataset has also been decontaminated with respect to the following benchmar
|
|
| 28 |
Total number of samples: 13,071,217.
|
| 29 |
|
| 30 |
Size of downloaded parquet files: 34G.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
Total number of samples: 13,071,217.
|
| 29 |
|
| 30 |
Size of downloaded parquet files: 34G.
|
| 31 |
+
|
| 32 |
+
### Filtered Version
|
| 33 |
+
|
| 34 |
+
There is a model-filtered version in the [filtered branch](https://huggingface.co/datasets/Geralt-Targaryen/openwebtext2/tree/filtered), including 12,804,779 samples .
|
| 35 |
+
|
| 36 |
+
Qwen2.5-32B-Instruct is used to generate language quality annotation (on a scale of 1-5) for 250K C4 samples. A RoBERT-large classifier is trained with regression on these annotations. Any document receiving a score of 1 or 2 from the classifier is removed. The remaining documents are also accompanied by their scores.
|
| 37 |
+
|
| 38 |
+
You can download this version by specifying the `--revision` argument:
|
| 39 |
+
|
| 40 |
+
```
|
| 41 |
+
huggingface-cli download --repo-type dataset Geralt-Targaryen/openwebtext2 --revision filtered --local-dir .
|
| 42 |
+
```
|