... Zoom:
- Install https://lmstudio.ai or similar app
- Select a model and download
The real bottleneck is hardware. I could only use upto 13B models. Results were worse than GPT 3.5 and slow.
Posted via D.Buzz
... Zoom:
The real bottleneck is hardware. I could only use upto 13B models. Results were worse than GPT 3.5 and slow.
Posted via D.Buzz