Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10M - 100M
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -31,3 +31,25 @@ configs:
|
|
| 31 |
- split: validation
|
| 32 |
path: data/validation-*
|
| 33 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
- split: validation
|
| 32 |
path: data/validation-*
|
| 33 |
---
|
| 34 |
+
|
| 35 |
+
<div align="center">
|
| 36 |
+
<img width="256" height="256" src="https://cdn-uploads.huggingface.co/production/uploads/63ca9515145e6c9716fbf376/LbTcGgiOr2gAJalHrqcNQ.png"/>
|
| 37 |
+
</div>
|
| 38 |
+
|
| 39 |
+
---
|
| 40 |
+
|
| 41 |
+
# Baby Python
|
| 42 |
+
## English + Python pretraining dataset.
|
| 43 |
+
|
| 44 |
+
|
| 45 |
+
This dataset contains the dataset from the [BabyLM 100M dataset](https://huggingface.co/nilq/babylm-100M), evenly matched by a truncated subset of the [small-python-stack](https://huggingface.co/nilq/small-python-stack).
|
| 46 |
+
|
| 47 |
+
**English / all of BabyLM 100M**
|
| 48 |
+
- CHILDES (child-directed speech)
|
| 49 |
+
- Subtitles (speech)
|
| 50 |
+
- BNC (speech)
|
| 51 |
+
- TED talks (speech)
|
| 52 |
+
- children's books (simple written language)
|
| 53 |
+
|
| 54 |
+
**Python**
|
| 55 |
+
- Python code (from [small-python-stack](https://huggingface.co/nilq/small-python-stack)) corresponding to 100M words of English.
|