Tanim Jalal

Currently I have tried 4 models that fit in my gpu: 1. mistral nemo: works perfectly 2. gemma 2 27b: the code runs until inference using use_fast = True but...