My own (ZeroWw) quantizations. output and embed tensors quantized to f16. all other tensors quantized to q5_k or q6_k.

Result: both f16.q6 and f16.q5 are smaller than q8_0 standard quantization and they perform as well as the pure f16.

GGUF

Model size

9.24B params

Architecture

gemma2

5-bit

6-bit

8-bit

16-bit

Inference API

Unable to determine this model's library. Check the docs .

Collection including ZeroWw/gemma-2-9b-it-GGUF