Disk Cleanup Guide and Commands | Generated by AI
OK here’s the full picture. Your disk is at 94% (only 900MB truly free, not 12GB — that was including purgeable space).
TOP SPACE HOGS (sorted by size):
CATEGORY SIZE WHAT TO DO
─────────────────────────────────────────────────────────────
~/.cache/huggingface/ 21 GB Models you downloaded — safe to prune
~/projects/lzwjava.github.io 58 GB Huge! Check for large media/old posts
~/Library/Containers/ 37 GB WeChat 19GB, Docker 8GB, QQ 4.5GB
~/.cache/ (total) 31 GB HF 21GB + torch 4.6GB + pre-commit 3GB + uv 1.3GB
~/Library/Caches/ 22 GB JetBrains 7.1GB, Homebrew 3.7GB, pip 2.6GB
~/Library/Application Sup. 24 GB JetBrains 5.4GB, nomic.ai 3.5GB, Lark 2GB
~/.npm/ 7.2 GB Old npm cache
~/.cache/torch/ 4.6 GB PyTorch hub models
~/Library/Developer/ 3.8 GB CoreSimulator 3.3GB
QUICK WINS (safe to delete now, ~40+ GB):
# 1. HuggingFace cache (21 GB) — re-download on demand
huggingface-cli scan-cache
huggingface-cli delete-cache
# 2. npm cache (7.2 GB)
npm cache clean --force
# 3. Homebrew cache (3.7 GB)
brew cleanup --prune=all
# 4. pip cache (2.6 GB)
pip3 cache purge
# 5. pre-commit cache (3 GB)
pre-commit clean
# 6. PyTorch hub cache (4.6 GB) — check first
ls ~/.cache/torch/hub/
rm -rf ~/.cache/torch/hub/*
# 7. JetBrains caches (7.1 GB)
rm -rf ~/Library/Caches/JetBrains/*
# 8. uv cache (1.3 GB)
uv cache prune
BIGGER CLEANUP (need your call):
# WeChat (19 GB!) — open WeChat settings → Storage → Clear
# This is the single biggest easy win after HF cache
# Docker (8 GB) — if not using Docker on Air:
docker system prune -a --volumes
# lzwjava.github.io (58 GB) — what's in there?
du -sh ~/projects/lzwjava.github.io/*/ | sort -rh | head
# Likely large images, videos, or build artifacts
Want me to run the safe cleanup commands (items 1-8)? That should free ~40+ GB. For WeChat and the blog repo, I need your decision.