No difference in size ?
#2
by
Pumba2
- opened
Hi
Isnt this model suppose to be smaller in size and should consume less ram ?
Coz from what I see this one is 7.8gb (lm studio app shows 7.8gb when choosing llm model in search tab) and the regular gguf gemma 3 12b is 8.1gb...
I asked this question in https://www.reddit.com/r/LocalLLaMA/... Is this quant the one mentioned in today's Google blogpost as is differs from the original QAT quant published in the Google's HF repository!
Yeah to be completely honest I'm not 100% sure what their own uploaded one is, it's massive for some reason :o