[Question] Step 3 loss curve #89
Description
Question
I performed the step 3 moe-finetunning on phi-2 model, the loss doesn't seemed to drop that much. I wonder if that's normal. Thanks!
{{ message }}
I performed the step 3 moe-finetunning on phi-2 model, the loss doesn't seemed to drop that much. I wonder if that's normal. Thanks!