search

qwopqwop200 / GPTQ-for-LLaMa

4 bits quantization of LLaMA using GPTQ
Apache License 2.0
2.98k stars 457 forks source link

compare with llama.cpp int4 quantize? #257

Open luohao123 opened 1 year ago

luohao123 commented 1 year ago

compare with llama.cpp int4 quantize?