maria-mh07

Hi! I'm on windows, using master 5a5f8b1 - I downloaded 13b and 30b alpaca models as mentioned by @madmads11 and @Puncia - Ran `python convert-unversioned-ggml-to-ggml.py models\Alpaca\13B models/LLaMA/tokenizer.model` and `python convert-unversioned-ggml-to-ggml.py...

I get the same error after cloning the repository and running ``` python generate.py \ --load_8bit \ --base_model 'decapoda-research/llama-7b-hf' \ --lora_weights 'tloen/alpaca-lora-7b' ``` Bitsandbytes installed from source. How much gpu...