3 ms·
I've tried running 3B LLAMA on my old machine and it was sluggish and kinda bad, but Claude 3.5 Sonnet works great. I will try the 70B model, should be good eno
by ac130kz 2y ago
I've tried running 3B LLAMA on my old machine and it was sluggish and kinda bad, but Claude 3.5 Sonnet works great. I will try the 70B model, should be good enough.
- dartos 2y agoYeah, I’ve found local models are completely useless for coding tasks whereas openai and Claude can at least do some refactors with guidance
- replete 2y agoI didn't have much success with ollama either. You really need a great accelerator for it to be useful enough. Codestral however is working out great so thanks for the tip.