pythia-160m-deduped-gptq-4bit / 4bit-128g.safetensors

Commit History

Use g instead of b for groupsize
33ed16d

OccamRazor commited on