Datasets:

HuggingFaceTB
/

dclm-edu

Modalities:

Formats:

Languages:

Size:

ArXiv:

Libraries:

License:

Dataset card Data Studio Files Files and versions Community

loubnabnl HF staff commited on 2 days ago

Commit

dbad8ad

·

verified ·

1 Parent(s): 32054a8

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -56,7 +56,7 @@ pipeline_exec.run()
 We train a 360M model (using [SmolLM2](https://huggingface.co/HuggingFaceTB/SmolLM2-360M) setup) on 200B tokens from DCLM, FineWeb-Edu and DCLM-Edu and evaluate on different benchmarks. DCLM-Edu denotes DCLM samples with an educational score higher than 3.
 We find that the model trained on DCLM-Edu performs better on knowledge and reasoning tasks (MMLU & ARC):
-<img src="https://cdn-uploads.huggingface.co/production/uploads/61c141342aac764ce1654e43/hOFJRusg6fEEtCpN-RJaP.png)" width="700" alt="image">
 We invite users to experiment with different data mixing depending on their model size.

 We train a 360M model (using [SmolLM2](https://huggingface.co/HuggingFaceTB/SmolLM2-360M) setup) on 200B tokens from DCLM, FineWeb-Edu and DCLM-Edu and evaluate on different benchmarks. DCLM-Edu denotes DCLM samples with an educational score higher than 3.
 We find that the model trained on DCLM-Edu performs better on knowledge and reasoning tasks (MMLU & ARC):
+<img src="https://cdn-uploads.huggingface.co/production/uploads/61c141342aac764ce1654e43/hOFJRusg6fEEtCpN-RJaP.png" width="700" alt="image">
 We invite users to experiment with different data mixing depending on their model size.