why the size more than double than original gemma 4b
#2
by
ryg81
- opened
Hey I just downloaded 4B model and found that on ollama website size of it is around 8.6 GB
while original model is just 3.3gb
gemma-3-abliterated defaults to f16, not Q4_K_M, and the effect of Q4_K_M is not very good, nor is it provided.
OK so your provided model is F16 and that is the reason. I apologize for not seeing those small details down there.
All versions of gemma-3-abliterated have been released for testing.
So in future we can hope to see quantization?