ZyangLee comments

Results 4 comments of


                                            ZyangLee

AttributeError: module 'torch.onnx' has no attribute 'set_training'

Same problem

RuntimeError: The size of tensor a (32000) must match the size of tensor b (32001) at non-singleton dimension 0

> My setting is same as u, u can just low your fastchat version with the command ' pip install fschat==0.1.10', which can fix this issue. You can see 'FastChat...

RuntimeError: The size of tensor a (32000) must match the size of tensor b (32001) at non-singleton dimension 0

> > > > My setting is same as u, u can just low your fastchat version with the command ' pip install fschat==0.1.10', which can fix this issue. You...

出现如下warning: tried to get lr value before scheduler/optimizer started stepping, returning lr=0

我是在使用deepspeed微调flant5系列模型时遇到的该问题，lr一直为0，上述方法只有对Transformers版本降级有效，且deepspeed不需要降级；transformers==4.40 --> 4.28.1, deepspeed=0.9.3