3 ms·
80GB is enough, yeah. I'm not sure what exact LORA/quantization settings would be ideal, but check out https://github.com/OpenAccess-AI-Collective/axolotl#conf
by brucethemoose2 3y ago
80GB is enough, yeah.
I'm not sure what exact LORA/quantization settings would be ideal, but check out https://github.com/OpenAccess-AI-Collective/axolotl#config https://github.com/OpenAccess-AI-Collective/axolotl#config