Bang DaeYeong

@byshiue Is there a part where you set the seed... Could you tell me which c++ script it is in? I'm trying to modify

I have a question and I am writing. quantized model(hivemind/gpt-j-6B-8bit) and of the original model(EleutherAI/gpt-j-6B) The generate inference speed is almost doubled(quantized model is much slower) I wonder if it...