Request for awq of the gemma 4 26B A4B MoE
#1
by rks2302 - opened
When will the awq quant be released for the gemma4 moe !?
Thanks :)
Any update on this pls?!
i saw people have already flooded 4 bit 26B, is it needed still?
Yes particularly for interactive use case moe has been quite faster, previously had been using your qwen 3.5 35B A3B awq 4 bit quant, it was fast but sometimes it produced weird indic language (might be training bias), so in the hope of your gemma 4 moe quant being the perfect one :)
Hi, any plans for releasing the awq for gemma moe variants?!
Hi, any plans to do so? If not then what about gemopus as we had seen awq versions of qwopus models?! Awaiting reply...
Thanks :)