Difference between revisions of "TensorFlow"

Revision as of 15:36, 7 November 2019

Description

TensorFlow is an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) that flow between them. This flexible architecture lets you deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device without rewriting code. TensorFlow also includes TensorBoard, a data visualization toolkit, and Keras, a high-level (easier to use) neural networks API, written in Python and capable of running on top of TensorFlow.

Required Modules

Run module spider tensorflow to find out what environment modules are available for this application.

System Variables

HPC_{{#uppercase:tensorflow}}_DIR - installation directory

Additional Information

As of version 2.0 tensorflow includes keras as tf.keras module. This is the module that you should use. See keras vs. tf.keras article for details.

To start a tensorflow session on a GPU or GPUs on HiPerGator you must request the '--gpus' resource and specify the 'gpu' partition in your job script or on the command line as described in GPU Access help page. E.g.

srun --partition=gpu --gpus=1 --ntasks=1 --mem=4gb --time=08:00:00 --pty -u bash -i

Citation

If you publish research that uses tensorflow you have to cite it as follows:

https://www.tensorflow.org/versions/r0.11/resources/bib.html

@@ Line 18: / Line 18: @@
 {{App_Description|app={{#var:app}}|url={{#var:url}}|name={{#var:app}}}}|}}
-TensorFlow is an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) that flow between them. This flexible architecture lets you deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device without rewriting code. TensorFlow also includes TensorBoard, a data visualization toolkit.
+TensorFlow is an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) that flow between them. This flexible architecture lets you deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device without rewriting code. TensorFlow also includes TensorBoard, a data visualization toolkit, and Keras, a high-level (easier to use) neural networks API, written in Python and capable of running on top of TensorFlow.
 <!--Modules-->
@@ Line 31: / Line 31: @@
 <!--Run-->
 {{#if: {{#var: exe}}|==Additional Information==
+As of version 2.0 tensorflow includes keras as tf.keras module. This is the module that you should use. See [https://www.pyimagesearch.com/2019/10/21/keras-vs-tf-keras-whats-the-difference-in-tensorflow-2-0/ keras vs. tf.keras] article for details.
-TensorFlow is built from source and optimized for GPGPUs we have. There are two modules. The default module e.g. 'tensorflow/1.7.0' is python3 based. There is a separate py2 moduel e.g. 'tensorflow/1.7.0py2' for legacy scripts.
+To start a tensorflow session on a GPU or GPUs on HiPerGator you must request the '--gpus' resource and specify the 'gpu' partition in your job script or on the command line as described in [[GPU_Access|GPU Access]] help page. E.g.
-  $ python3
+  srun --partition=gpu --gpus=1 --ntasks=1 --mem=4gb --time=08:00:00 --pty -u bash -i
- >>> import tensorflow as tf
-A number of useful python modules have been installed within the tensorflow python virtual environment on request e.g. keras.
- $ python3 -c 'import keras'
 |}}
 <!--Job Scripts-->