GPT2 model used in LLama example?
madisonPickering opened this issue · comments
Hello,
It looks like line 39 of run.py in modelzoo/modelzoo/transformers/pytorch/llama/run.py is causing a GPT2 model to be trained, instead of a Llama model. I suspect this is not the intended behavior? It is also possible I am reading the file incorrectly, which if this is the case I apologize for the inconvenience