Comment on DeepSeek Coder: Let the Code Write ItselfparentComments−findjashua2ydid a comparison on LM Studio - the answers are eerily similar, but Mixtral is way way faster. Codellama-70B is slow to the point of being unusable.(M1 Max, 64 GB RAM)
Comments
did a comparison on LM Studio - the answers are eerily similar, but Mixtral is way way faster. Codellama-70B is slow to the point of being unusable.
(M1 Max, 64 GB RAM)