Armin W. Thomas
Armin W. Thomas
Thanks for bringing this to my attention! I've added this to my todo list and will try to get to it as soon as possible (but likely won't be able...
Thank you for your response @pbontrager. See below for the differences resulting from the script above: ```bash Comparing gemma 2b Loading checkpoint shards: 100%|██████████████████████████████████████████████████████████████████████████| 2/2 [00:00
thank you @pbontrager for tracking this down! if helpful, I am happy to help with this?
> Note that this script compares the models in float32 precision on CPU. When I modify the script to compare the models on GPU in bf16 precision, I suddenly see...