Safe to Delete

Hugging Face Tokenizers Cache (~/.cache/huggingface/tokenizers)

Space: 500 MB – 5 GB developer Windows, macOS, Linux
Quick Answer: The tokenizers cache stores BPE vocabulary files for LLM models. Deleting tokenizer cache files is safe.

Bottom Line: It is 100% safe to delete Hugging Face tokenizers cache. Deleting it clears cached vocabulary JSON files.

What is the Tokenizers Cache?

Located at ~/.cache/huggingface/tokenizers, Hugging Face transformers caches Byte-Pair Encoding (BPE) vocabulary and tokenizer config files.

Can You Delete Tokenizer Files?

✅ Yes, safe to delete.

rm -rf ~/.cache/huggingface/tokenizers/*

Discussion

Loading authentication...

Related in developer

The AI Engineer's Model Disk Cleanup Guide: Hugging Face, Ollama & PyTorch

The ultimate 1,200-word field guide for AI developers and local LLM hobbyists to reclaim 200GB+ from Ollama, Hugging Face, PyTorch, and vLLM model weights.

Safe to Delete

Android Studio System Index Cache (AndroidStudio Cache)

Android Studio caches Gradle project indexes, VFS caches, and compiler artifacts. Invalidate and clear is safe.

Safe to Delete

Ansible Automation Temp & Fact Cache (~/.ansible/tmp)

Ansible stores temporary playbook execution modules and target host fact JSON caches in ~/.ansible/tmp. Clearing Ansible tmp is safe.

Safe to Delete
Back to all files