Add `library_name: transformers` to metadata
This PR adds the library_name: transformers metadata tag to the model card.
Evidence from the Quickstart section clearly shows the model's compatibility with the transformers library, as it imports AutoTokenizer and AutoModelForCausalLM from transformers. The config.json also defines auto_map entries for transformers-compatible classes like Qwen3ForCausalLM.
Adding this tag enables the automated "how to use" widget on the Hugging Face Hub, providing users with a quick and convenient way to interact with the model.
Hi @nielsr , thanks for the suggestion regarding the library_name tag!
I have manually updated the README.md to include the library_name: transformers metadata.
However, I will close this PR without merging the code changes for two reasons:
- The change to the variable name (
cur_gen_pos_ids->cur_gen_blocks_pos_ids) is incorrect and would break the logic (or cause an undefined variable error). - The prompt modification introduces a Python syntax error (multi-line strings require triple quotes).
I appreciate the help with the metadata integration! Closing this now as the tag is live.