Text Generation
Transformers
Safetensors
minimax_m2
conversational
custom_code
fp8

Question about XET storage: Why are some safetensors files 40 Bytes?

#35
by yingchaox - opened

I noticed an interesting detail about the model files and I'm curious about the technical implementation.

Observation:
When I download the model, some .safetensors files (like model-00129-of-00130.safetensors onwards) show as 40 Bytes on disk, but the model loads and runs perfectly fine.

Sign up or log in to comment