Add model card and fix base_vlm to public model ID
#2
by wcy1122 - opened
- Add README.md with checkpoint details, benchmark results, and usage
- config.yaml: base_vlm -> StarVLA/Qwen3-VL-4B-Instruct-Action (the previous
ID is not publicly accessible; the action-extended vocab of 153984 matches
this checkpoint) - Add training_config.original.yaml preserving the original run config
wcy1122 changed pull request status to merged