Anonymous
8/5/2025, 4:20:35 PM
No.106149955
>>106149915
It does happen for deepseek but deepseek at Q1 is still the best model you can run in that amount of memory. A model with fewer parameters at a larger quant will be worse.
It does happen for deepseek but deepseek at Q1 is still the best model you can run in that amount of memory. A model with fewer parameters at a larger quant will be worse.