Remove Granite 2B

#1
by adamelliotfields - opened

Should only use models that fit in a single safetensors file; must be under 2B.

As of right now, the only alternative non-thinking model is Gemma 2 2B Gemma 3 1B. Thinking models eat up ZeroGPU time, so best saved for API.

Edit: forgot that we're not using gated models; will just delete Granite since there's no viable alternatives.

adamelliotfields changed discussion title from Replace Granite 2B with Qwen 1.5B R1 Distilled to Remove Granite 2B
Your need to confirm your account before you can post a new comment.

Sign up or log in to comment