This is a test run. I trained it just to see whether it was possible to train 34B 200K ctx Yi on 24GB card and for it to work fine after exl2 quant.
I actually really like the vocabulary that this lora has. It spits out a bit of chinese due to light English training, but for the most part it gives pretty natural outputs.
- Downloads last month
- 12
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support