model-caching
Shared Hugging Face cache for teams. Sync local HF cache to Backblaze B2 or S3, eliminate repeated downloads, and enable fast, consistent model access across machines and CI.