the same fine tune code, flan t5 work, and this does not.
#3
by
CUIGuy
- opened
ValueError: You are trying to save a non contiguous tensor: encoder.block.0.layer.0.SelfAttention.q.weight which is not allowed. It either means you are trying to save tensors which are reference of each other in which case it's recommended to save only the full tensors, and reslice at load time, or simply call .contiguous() on your tensor to pack it before saving.
I have some code that fine-tune flan t5 on some nlp tasks, and when I try to switch to this version instead, the same code give me this, what I am missing? I thought we should use this instead flan T5 for fine tuning, the transformer code is not working on this one?