Search Results
6/27/2025, 6:49:27 PM
>>105723940
>I hate chinese for conning the world with their deepseek garbage that according to the square root moe law is an equivalent of a 26B dense model.
...but it has 37b active params.
>I hate chinese for conning the world with their deepseek garbage that according to the square root moe law is an equivalent of a 26B dense model.
...but it has 37b active params.
6/27/2025, 6:27:45 PM
>>105723602
>I hate chinese for conning the world with their deepseek garbage that according to the square root moe law is an equivalent of a 26B dense model.
...but it has 37b active params.
>I hate chinese for conning the world with their deepseek garbage that according to the square root moe law is an equivalent of a 26B dense model.
...but it has 37b active params.
Page 1