README.md

metadata

configs:
  - config_name: javascript
    data_files:
      - split: train
        path: data/javascript-*.parquet

LangMap-TheStack-javascript-100M-olmo2

Code finetuning dataset for javascript streamed from bigcode/the-stack.

• Tokens collected: 100,000,000 (target: 100,000,000)
• Tokenizer: allenai/OLMo-2-1124-7B
• Schema: {"text": [...]} (sanitised source code)