![](https://cdn-avatars.huggingface.co/v1/production/uploads/654b3a5c8f78bd30275cc6a7/Y5B3cEIArRptUH3fMF7M3.png)
minpeter/QLoRA-Llama-3.2-1B-alpaca
Updated
•
3
•
1
Note QLoRA-Llama-3.2-1B-chatml is fundamentally the same, but trained on a much larger dataset (21k steps).
Note 52 steps, 2 minutes to train