Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Tensorflowjs converter not work using google collab #8356

Closed
nurulaqilah2207 opened this issue Aug 7, 2024 · 5 comments
Closed

Tensorflowjs converter not work using google collab #8356

nurulaqilah2207 opened this issue Aug 7, 2024 · 5 comments
Assignees

Comments

@nurulaqilah2207
Copy link

Hi. I tried to convert .h5 model to tensorflowjs using google collab. This is the error that I get:
2024-08-07 08:41:08.593697: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:479] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-08-07 08:41:08.618933: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:10575] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-08-07 08:41:08.618988: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1442] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2024-08-07 08:41:09.629240: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT

@nurulaqilah2207 nurulaqilah2207 added the type:bug Something isn't working label Aug 7, 2024
@gaikwadrahul8 gaikwadrahul8 self-assigned this Aug 7, 2024
@gaikwadrahul8
Copy link
Contributor

gaikwadrahul8 commented Aug 7, 2024

Hi, @nurulaqilah2207

Thank you for bringing this issue to our attention and to effectively investigate and resolve the issue, we kindly request that you provide your .h5 model file and the associated Google Colab notebook. These resources will enable us to conduct a thorough analysis and identify potential root causes

Your cooperation in sharing these materials will significantly expedite our troubleshooting process and help us provide resolution.

Thank you for your cooperation and patience.

@nurulaqilah2207
Copy link
Author

Hi @gaikwadrahul8. Thank you for responding. Here is the google Colab notebook: convert-to-tfjs, adn the .h5 model file: model.h5.

@gaikwadrahul8
Copy link
Contributor

gaikwadrahul8 commented Aug 8, 2024

Hi, @nurulaqilah2207

Thank you for providing the Google colab notebook along with your model, I have tried to convert your model into TensorFlow.js format and I'm able to convert your .h5 model to TensorFlow.js format, please refer this gist-file, it is generating the multiple .bin files and model.json file

cuFFT, cuDNN, cuBLAS factory registration errors indicate conflicts during registration of libraries used for hardware acceleration on NVIDIA GPU. This can happen due to multiple installations of CUDA, cuDNN, or cuBLAS, or version mismatches but I did not observe that behavior from my end in the colab notebook with GPU.

TF-TRT Warning: Could not find TensorRT this warning suggests TensorFlow is unable to locate TensorRT which is needed for optimizing models for inference on NVIDIA GPUs and it's an optional set up so you can ignore that warning message, However .h5 model is converting successfully to TensorFlow.js format so you may ignore those warning messages

Here is screenshot for reference :

image

If I have missed something here please let me know ?

Thank you for your cooperation and patience.

@gaikwadrahul8
Copy link
Contributor

Hi, @nurulaqilah2207

We are pleased to hear that your issue has been resolved as you confirmed in the other issue #8355. If you require no further assistance, kindly close this issue. For additional support related to TensorFlow.js please feel free to open a new issue.

Thank you for your cooperation and patience.

Copy link

Are you satisfied with the resolution of your issue?
Yes
No

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants