Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
[Fix] Fetching the Git-LFS tokenizer files (#1954)
Prior to this PR, when running commands like ```shell python3 -m mlc_chat chat HF://mlc-ai/gemma-7b-it-q4f16_2-MLC ``` only the binary weight files are downloaded, among all the Git LFS files. For models like Gemma whose tokenizer is large and also in Git LFS file, the tokenizer files are not effectively downloaded automatically. For example, the cloned Gemma `tokenizer.json` file has content ``` version https://git-lfs.github.com/spec/v1 oid sha256:05e97791a5e007260de1db7e1692e53150e08cea481e2bf25435553380c147ee size 17477929 ``` and this content is never realized to the actual tokenizer. This will lead to the issue of #1913. This PR fixes the issue by pulling all the Git LFS files that are not binary files.
- Loading branch information