| --- | |
| A mixture of datasets focused on Source/Intermediate Representation training for multilingual code generation. | |
| For SCU CSEN346. | |
| **Usage**: | |
| Pass to continued_pretrain.py with | |
| ``` | |
| --dataset_name './tokenized_dataset' | |
| ``` | |
| Sources: | |
| UKPLab/SLTrans | |
| bigcode/the-stack | |
| allenai/peS2o | |
| open-web-math/open-web-math |