pvop

Results 4 comments of pvop

[BUG]: fp32 param and grad have different shape torch.Size([5064704]) vs torch.Size([128000]) when use lora_rank=4 at stage 1

the same issue too

[BUG]: fp32 param and grad have different shape torch.Size([5064704]) vs torch.Size([128000]) when use lora_rank=4 at stage 1

the same issue too

[BUG]: fp32 param and grad have different shape torch.Size([5064704]) vs torch.Size([128000]) when use lora_rank=4 at stage 1

AssertionError: fp32 param and grad have different shape I have solved this error. I use GLM-10B to train reward model. The outputs of 'mems' is used as last_hidden_states. But the...

When I ran the 'activations.py', I met an error

I get the same error. Because the tensorflow.keras version is different, this code use keras.layers.merge, but there is not merge in new keras. So you should use keras.layers.Concatenate(), so the...