Update README.md
Browse files
README.md
CHANGED
|
@@ -17,12 +17,10 @@ language:
|
|
| 17 |
- ar
|
| 18 |
---
|
| 19 |
|
| 20 |
-
|
| 21 |
|
| 22 |
AraMix (https://arxiv.org/abs/2512.18834v1) is an Arabic pretraining corpus containing 178 billion tokens across 179 million documents (in the minhash subset). Rather than scraping the web again, AraMix combines seven publicly available Arabic datasets, applies Arabic-specific quality filtering, and performs cross-dataset deduplication.
|
| 23 |
|
| 24 |
-
<img src="https://huggingface.co/datasets/AdaMLLab/AraMix/resolve/main/finetasks_arabic_results.png" width="900" alt="Finetasks benchmark scores, showing AraMix-Consensus as SOTA.">
|
| 25 |
-
|
| 26 |
We train a 1.4B parameter language model through nanotron on 30 billion tokens to show that the `consensus` subset of AraMix outperforms the previous state-of-the-art, [arabicweb24](https://huggingface.co/datasets/lightonai/ArabicWeb24) (see [Appendix A9 in the Fineweb-2 paper](https://arxiv.org/pdf/2506.20920)) while having more total tokens. Furthermore, the `minhash_deduped` subset performs on-par with nearly 5 times the total number of tokens.
|
| 27 |
|
| 28 |
In this setup, we remove all samples in `consensus` with more than 5 duplicates.
|
|
|
|
| 17 |
- ar
|
| 18 |
---
|
| 19 |
|
| 20 |
+
<img src="https://huggingface.co/datasets/AdaMLLab/AraMix/resolve/main/finetasks_arabic_results.png" width="900" alt="Finetasks benchmark scores, showing AraMix-Consensus as SOTA.">
|
| 21 |
|
| 22 |
AraMix (https://arxiv.org/abs/2512.18834v1) is an Arabic pretraining corpus containing 178 billion tokens across 179 million documents (in the minhash subset). Rather than scraping the web again, AraMix combines seven publicly available Arabic datasets, applies Arabic-specific quality filtering, and performs cross-dataset deduplication.
|
| 23 |
|
|
|
|
|
|
|
| 24 |
We train a 1.4B parameter language model through nanotron on 30 billion tokens to show that the `consensus` subset of AraMix outperforms the previous state-of-the-art, [arabicweb24](https://huggingface.co/datasets/lightonai/ArabicWeb24) (see [Appendix A9 in the Fineweb-2 paper](https://arxiv.org/pdf/2506.20920)) while having more total tokens. Furthermore, the `minhash_deduped` subset performs on-par with nearly 5 times the total number of tokens.
|
| 25 |
|
| 26 |
In this setup, we remove all samples in `consensus` with more than 5 duplicates.
|