Davide Manco

icon company Fourthline icon location Amsterdam

Results 26 comments of Davide Manco

uvloop is grouping responses to multiple requests/messages

Can someone review the PR from #619? This issue has been here for a while and it's quite an annoying one.

'jit_compile=True' raises InternalTorchDynamoError with EfficientNetV2 models on torch backend.

Thank you @dhantule 🙏

'jit_compile=True' raises InternalTorchDynamoError with EfficientNetV2 models on torch backend.

Hi @SamanehSaadat, no unfortunately that's not a possibility for us 🙁

'jit_compile=True' raises InternalTorchDynamoError with EfficientNetV2 models on torch backend.

Hi @james77777778, I see you made a PR related to this issue, did you fix the problem?

'jit_compile=True' raises InternalTorchDynamoError with EfficientNetV2 models on torch backend.

@james77777778 I see, thanks for the answer! Do you know which latest version of pytorch is still fully compatible with Keras?

Use Multi GPU(Mirrored Strategy) training with XLA and AMP(Mixed Precision)

Hi! I'm still experiencing this error, is there any update?