Hi! I'm Niels, part of the community science team at Hugging Face.
This PR improves the model card and metadata for the IAPO checkpoints:
- Changed the `pipeline_tag` from `reinforcement-learning` to `text-generation` for better discoverability and to enable the inference widget/standard LLM evaluation.
- Added `library_name: transformers` to the metadata to enable the "Use in Transformers" code snippet button.
- Cleaned up the Markdown layout and ensured clear links to the paper and code repository.
- Provided a clear usage example for loading the different subfolders available in this repository.
Upload Qwen2.5-7B-Instruct_GSM8K/model-00001-of-00004.safetensors with huggingface_hub
