DanielHesslow

Hi, I'm consistently seeing very high cpu overhead when launching (already jitted) kernels. Taking matrix multiply as an example, on an A6000 I'm getting similar performance for both in torch...

Access utf-8 byte sequence for each token

2

Hi, It would be great if it was possible to get the utf-8 byte sequence corresponding to each token id. Since tokenizers return strings, tokens which are not valid unicode...

🐛 hf-hub does not properly follow repo redirections with auth (after eg. renaming)

1

👋 Hey This Works: ``` rust let api = ApiBuilder::from_env() .build() .unwrap(); let repo = api.model("meta-llama/Llama-3.1-8B".to_string()); repo.download("config.json").unwrap(); ``` While this Fails: ``` rust let api = ApiBuilder::from_env() .build() .unwrap(); let...