GPT4 x Vicuna 13B 4-bit

From NousResearch: https://huggingface.co/NousResearch/gpt4-x-vicuna-13b

Folders

ggml: q4_0 and q4_1

gptq: works with Triton and CUDA branches