File size: 329 Bytes
03ce227 |
1 2 3 4 5 6 |
Quantized from https://huggingface.co/ausboss/llama-30b-supercot
Quantized with https://github.com/0cc4m/GPTQ-for-LLaMa for compatibility with 0cc4m's fork of KoboldAI
Command used to quantize:
```python llama.py c:\llama-30b-supercot c4 --wbits 4 --true-sequential --groupsize 128 --save_safetensors 4bit-128g.safetensors``` |