Stack Overflow
Stack Overflow

Questions tagged [tpu]

Use this tag for Tensor Processing Unit (TPU). TPU is an application-specific integrated circuit developed by Google.

421 questions
2
votes
1 answer

Segmentation Fault Tensorflow Lite

Hi I got the following error: "Segmentation fault" on my Raspberry Pi 4 Model B. I am following David Tian's Guide to build a Smart Robot Car which uses Deep Learning. Some important informations: Raspberry Pi 4 Model B Raspbian Buster Python3…
Micha93
  • 628
  • 1
  • 9
  • 22
2
votes
0 answers

Using TensorFlow 2.0 keras: NotFoundError FetchOutputs node : not found [Op:AutoShardDataset]

I am trying to make use of TPU on Google Colab. I have followed the guide official tensorflow guide to use tpu as follows: resolver = tf.distribute.cluster_resolver.TPUClusterResolver(tpu='grpc://' +…
Shashikant Kadam
  • 71
  • 1
  • 5
2
votes
1 answer

Permission denied on resource project None on 1 thread only in tpu_cluster_resolver

I'm running an BERT pretraining code on cloud TPUs from the Compute Engine. Each time i run it, i get this error on 1 thread but the training continues normally. I ran the same code on google Colab TPUs and it worked fine. for the…
wissam antoun
  • 65
  • 5
2
votes
0 answers

TPU in colab fails while GPU works

I am working on a modified version of the TPU fmnist example. While the original works fine, in my version I reduce the frequency of some classes according to a percentile (also omitting some classes, but that does not seem to cause the…
Dinari
  • 2,487
  • 13
  • 28
2
votes
1 answer

TPU core error on Google Cloud Platform (Cannot find any TPU cores in the system. Please double check Tensorflow master address and TPU worker)

The error message is this Traceback (most recent call last): File "./run_classifier.py", line 914, in tf.app.run() File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/platform/app.py", line 40, in run _run(main=main,…
crwarwa
  • 121
  • 1
  • 2
  • 10
2
votes
1 answer

Should I use a 2D or 3D convolution for a 3D grayscale image?

I'm working with a TFRecord dataset consisting of multiple grayscale images of cross-sections of a 3D object, ending up with the shape [32, 256, 256]. The dimension of 32 represents the number of cross-sections, and it is significantly less than the…
Justin Zhang
  • 45
  • 1
  • 7
2
votes
1 answer

No OpKernel was registered to support Op 'TPUReplicateMetadata' used by node TPUReplicateMetadata

When I run the following .ipynb: https://colab.research.google.com/drive/1DpUCBm58fruGNRtQL_DiSVbT90spdZgm I got: No OpKernel was registered to support Op 'TPUReplicateMetadata' used by node TPUReplicateMetadata (defined at…
user7862197
2
votes
1 answer

How to use local Coral USB TPU with Google Colab (instead of Cloud TPU)

I have a USB TPU and would like to use it as LOCAL RUNTIME in Google Colab. I was not able to find any resources on this topic. You can use a local Runtime (local Jupyter) and it is explained here :…
RodR
  • 21
  • 1
2
votes
1 answer

Is it possible to print value of TPU cloud variable in tensorflow?

Having disabled eager execution, I am able to connect to my cloud TPU and run my custom training loop. After calculating the loss, I would like to print that variable. Given that loss is a tensor of a Cloud TPU, I haven't found any way to print it…
gdaras
  • 9,401
  • 2
  • 23
  • 39
2
votes
1 answer

Colab TPU: TensorFlow '2.0.0-beta0' LinearClassifier .train Bug

Attempting to get LinearClassifier running with Colab TPU. https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/estimator/LinearClassifier TPUStrategy is supported in TensorFlow 2.0…
Machine Learning
  • 485
  • 6
  • 15
2
votes
0 answers

'strided_slice_19' (op: 'StridedSlice') with input shapes

I tried to use keras.preprocessing.image.ImageDataGenerator on TPU,but i get this error from the first epoch.the same code work with jupyter notebook but take hours for trainig. My MODEL: model = Sequential() model.add(Conv2D(32, (3, 3),…
Youssef Mhamdi
  • 51
  • 7
2
votes
0 answers

Using TPU on GKE: Error recorded from infeed: Socket closed

Once in a while our GKE TPUEstimator based training job using TPUs fail with: Error recorded from infeed: Socket closed An error was raised. This may be due to a preemption in a connected worker or parameter server. The current session will be…
Prashast
  • 5,645
  • 3
  • 30
  • 31
2
votes
0 answers

Is it possible to run python tensorflow code on TPU without using the Estimator API?

I have spent weeks now trying to write a Python level Tensorflow code that could communicate with TPUs directly. How would it be possible to implement the system that could run on a TPU without the Estimator API? Resources I tried: All the…
user3815403
  • 57
  • 5
2
votes
1 answer

How does BERT utilize TPU memories?

README in the Google's BERT repo says, even a single sentence of length 512 can not sit in a 12 GB Titan X for the BERT-Large model. But in the BERT paper, it says 64 TPU chips are used to train BERT-Large with a maximum length 512 and batch size…
soloice
  • 980
  • 8
  • 17
2
votes
2 answers

ValueError: slice index 0 of dimension 0 out of bounds. while using google colab TPU

I tried to run simple autoencoder with DENSE layer on TPU. But this error happened. It was OK when I run it on GPU and CPU. I'm using Google Colab I tried to change the input_dims to input_shape in the first layer as recommended by the Tensorflow.…
Muhammad Faiz Misman
  • 21
  • 1
  • 2
1 2 3
…
28 29