Hfrxdjt

Have I saved too much for retirement so far?

Two-sided logarithm inequality

What linear sensor for a keyboard?

Can somebody explain Brexit in a few child-proof sentences?

How to align and center standalone amsmath equations?

Reply 'no position' while the job posting is still there

If a character with the Alert feat rolls a crit fail on their Perception check, are they surprised?

Open a doc from terminal, but not by its name

Folder comparison

How do ground effect vehicles perform turns?

What is the gram­mat­i­cal term for “‑ed” words like these?

Greco-Roman egalitarianism

How do I implement a file system driver driver in Linux?

Translation of Scottish 16th century church stained glass

Is a model fitted to data or is data fitted to a model?

Can I sign legal documents with a smiley face?

Did arcade monitors have same pixel aspect ratio as TV sets?

On a tidally locked planet, would time be quantized?

Some numbers are more equivalent than others

Global amount of publications over time

What (else) happened July 1st 1858 in London?

Journal losing indexing services

My friend sent me a screenshot of a transaction hash, but when I search for it I find divergent data. What happened?

Why is Arduino resetting while driving motors?

How can I increase CUDA load on a Tensorflow deep learning task after reducing batch size to fit the GPU?

How to calculate the mini-batch memory impact when training deep learning models?Public cloud GPU support for TensorFlowOnline vs minibatch training for speedWhy Tensorflow does NOT quit when CUDA_ERROR_OUT_OF_MEMORYTraining Inception V3 based model using Keras with Tensorflow BackendTensorflow Deep learning network not utilizing GPU?Execution time of the same algorithm for different runsFully convolutional networks with partially segmented dataWhy i get OOM error although my model is not that large?Using CPU after training in GPU

0

$begingroup$

I am running this TensorFlow task for Swahili-to-English on an NVidia GeForce 1060 GPU with 6GB of VRAM:
https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/contrib/eager/python/examples/nmt_with_attention/nmt_with_attention.ipynb

I have to reduce the batch size to get the example code not to run out of memory on the GPU. With trial and error, I find a batch size (minimum is 1) where the code runs and uses the most memory. I notice that, as I reduce the batch size, the CUDA core load as reported by Windows Task Manager GPU view goes down.

The application described in the link above creates a complex TensorFlow network. I don't know whether Tensorflow creates one copy of the network or multiple copies to load the GPU.

If it can create multiple copies, is there a TensorFlow switch for that? I don't think memory speed should be a bottleneck in feeding the GPU. That is, I should be able to optimize between batch size or number of jobs resident in the GPU, and number of compute networks in the GPU.

Is there an easy way in TensorFlow to assess the size in CUDA cores of a compute network?

What are the factors I can use to optimize CUDA load in a small GPU with a large deep learning task?

machine-learning tensorflow machine-translation

share|improve this question

asked 22 mins ago

Lars EricsonLars Ericson

1334

$endgroup$

add a comment | 

0

$begingroup$

I am running this TensorFlow task for Swahili-to-English on an NVidia GeForce 1060 GPU with 6GB of VRAM:
https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/contrib/eager/python/examples/nmt_with_attention/nmt_with_attention.ipynb

I have to reduce the batch size to get the example code not to run out of memory on the GPU. With trial and error, I find a batch size (minimum is 1) where the code runs and uses the most memory. I notice that, as I reduce the batch size, the CUDA core load as reported by Windows Task Manager GPU view goes down.

The application described in the link above creates a complex TensorFlow network. I don't know whether Tensorflow creates one copy of the network or multiple copies to load the GPU.

If it can create multiple copies, is there a TensorFlow switch for that? I don't think memory speed should be a bottleneck in feeding the GPU. That is, I should be able to optimize between batch size or number of jobs resident in the GPU, and number of compute networks in the GPU.

Is there an easy way in TensorFlow to assess the size in CUDA cores of a compute network?

What are the factors I can use to optimize CUDA load in a small GPU with a large deep learning task?

machine-learning tensorflow machine-translation

share|improve this question

asked 22 mins ago

Lars EricsonLars Ericson

1334

$endgroup$

add a comment | 

$begingroup$

I am running this TensorFlow task for Swahili-to-English on an NVidia GeForce 1060 GPU with 6GB of VRAM:
https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/contrib/eager/python/examples/nmt_with_attention/nmt_with_attention.ipynb

I have to reduce the batch size to get the example code not to run out of memory on the GPU. With trial and error, I find a batch size (minimum is 1) where the code runs and uses the most memory. I notice that, as I reduce the batch size, the CUDA core load as reported by Windows Task Manager GPU view goes down.

The application described in the link above creates a complex TensorFlow network. I don't know whether Tensorflow creates one copy of the network or multiple copies to load the GPU.

If it can create multiple copies, is there a TensorFlow switch for that? I don't think memory speed should be a bottleneck in feeding the GPU. That is, I should be able to optimize between batch size or number of jobs resident in the GPU, and number of compute networks in the GPU.

Is there an easy way in TensorFlow to assess the size in CUDA cores of a compute network?

What are the factors I can use to optimize CUDA load in a small GPU with a large deep learning task?

machine-learning tensorflow machine-translation

share|improve this question

asked 22 mins ago

Lars EricsonLars Ericson

1334

$endgroup$

share|improve this question

asked 22 mins ago

Lars EricsonLars Ericson

1334

share|improve this question

asked 22 mins ago

Lars EricsonLars Ericson

1334

asked 22 mins ago

Lars EricsonLars Ericson

1334

asked 22 mins ago

Lars EricsonLars Ericson

1334