HK Relay Upload to HuggingFace

HK Relay Upload to HuggingFace | Generated by AI

Home 2026.06

Done! 🎉

model.safetensors uploaded to HuggingFace in 11 min 39 sec from HK.

Full pipeline completed:
  Mac → HK (SCP):       30 min   (475 MB, ~263 KB/s)
  HK → HuggingFace:     11 min   (498 MB, ~700 KB/s)
  Total:                ~42 min

Your model is live:
  https://huggingface.co/lzwjava/sec-edgar-gpt-124m
  Commit: d769c2a2887896bd98db2e8616734ffe31804fe9

Cleanup: You can now destroy the Aliyun HK ECS instance to stop paying:

# Via Aliyun console or CLI
aliyun ecs DeleteInstance --InstanceId <instance-id> --Force true

The HK relay trick worked well — China→HK tolerable (no stalls), HK→HuggingFace fast (~700 KB/s). Next time you need to upload to HF from China, this is the playbook.

Back Donate