I have an auto generated ‘forward’ func:
def forward(a, b, c):
…tmp=torch.lstm(a, (torch.tensor(…), torch.tensor(…)), self.lstm.weight, self.lstm.bias)
…
return ret
After using ‘torch._dynamo.explain‘, i got one graph containing ops before torch.lstm and one graph break reason ‘step_unsupported‘ and no more. But there are still a lot of code left after torch.lstm. i thought torch.lstm may cause one graph break instead of ending torch.compile. Why is that and what should i do to make torch.compile work on the code after torch.lstm?