You can deploy this model and start earning money today!

owenqwenllmwine/
code-v1

continued-pretraining

Views

No views yet

	Dataset Card for Code Corpus code-v1




	


	Dataset Description




	


	Dataset Summary

Code Corpus code-v1 is a token-balanced collection of source code, commit messages and diffs, programming Q&A, and technical text. It is intended for language-model pretraining and continued pretraining. All examples use a common JSONL schema with per-record provenance, license metadata, exact token counts, and content hashes. The released dataset contains 2 billion tokens, 2… See the full description on the dataset page: https://huggingface.co/datasets/owenqwenllmwine/code-v1.