TensorFlow 2 User Guide: updates to links to TF1 examples and adding note about TF1 examples

Summary:
From SDK 3.1, TensorFlow 1 is only supported in CentOS 7 and the Examples and Tutorials repos for TensorFlow 1 are fixed at version 3.0. This diff adds a note and updates the
links to the examples and tutorials repos.

Reviewers: #tensorflow, #framework_ip_review_-_any_oss_or_third-party_code_use_has_been_approved, #tech_docs, gauthamg

Reviewed By: #tensorflow, #framework_ip_review_-_any_oss_or_third-party_code_use_has_been_approved, gauthamg

Subscribers: timl

Maniphest Tasks: T71447

Differential Revision: https://phabricator.sourcevertex.net/D78013

Author: jayniep-gc

tensorflow/compiler/plugin/poplar/docs/custom_codelet.rst

@@ -391,7 +391,7 @@ files. Only the Poplar and PopLibs headers, and the corresponding libraries are
 required.
 
 You can add ``-g`` to the above command to compile the custom operation with
-debugging symbols. This allows you to debug the C++ code with ``gdb``. 
+debugging symbols. This allows you to debug the C++ code with ``gdb``.
 
 
 Using the custom op in TensorFlow
@@ -461,9 +461,14 @@ Some examples of using a custom op in TensorFlow are shown in the following
 sections. There are further examples in the Graphcore GitHub tutorials
 repository:
 
-* :tutorials-repo:`Custom op with gradient <feature_examples/tensorflow/custom_gradient>`
+* :tutorials-repo-tf1:`Custom op with gradient <feature_examples/tensorflow/custom_gradient>`
+
+* :tutorials-repo-tf1:`Custom op with codelet for custom vertex code <feature_examples/tensorflow/custom_op>`
+
+.. note:: These examples are for TensorFlow 1.
+
+  From Poplar SDK 3.1, TensorFlow 1 will only be supported in CentOS 7. In addition, `Examples <https://github.com/graphcore/examples/tree/v3.0.0>`__ and `Tutorials <https://github.com/graphcore/tutorials/tree/sdk-release-3.0>`__ for TensorFlow 1 are only available up to version 3.0 of the SDK. There has been limited testing of the 3.0 versions of the TensorFlow 1 tutorials and examples with Poplar SDK 3.1.
 
-* :tutorials-repo:`Custom op with codelet for custom vertex code <feature_examples/tensorflow/custom_op>`
 
 
 .. _inplace_operations:

tensorflow/compiler/plugin/poplar/docs/intro.rst

@@ -5,13 +5,16 @@ The purpose of this document is to introduce the TensorFlow framework from the
 perspective of developing and training models for the IPU. It assumes you have
 some knowledge of machine learning and TensorFlow.
 
+For more information about the IPU architecture, abstract programming model and tools, as well as algorithmic techniques, refer to the :external+ipu-programmers-guide:doc:`index`. The :external+memory-performan-optimisation:doc:`index` contains guidelines for optimising performance in machine learning models running on the IPU.
+
+
 .. note:: This document is for TensorFlow 2. For information on TensorFlow 1
           please refer to `Targeting the IPU from TensorFlow 1
           <https://docs.graphcore.ai/projects/tensorflow1-user-guide/>`_.
 
-See the "Getting Started" guide for your IPU system in the
-`Graphcore documentation portal <https://docs.graphcore.ai>`_
-for information on installing the Poplar SDK and the Graphcore implementation of TensorFlow.
+See the `Getting Started" guide for your IPU system <https://docs.graphcore.ai/en/latest/getting-started.html>`__ in the
+`Graphcore documentation <https://docs.graphcore.ai>`_
+for information on installing the Poplar SDK and refer to the :external+tensorflow1-quick-start:doc:`index` for installing TensorFlow 1 and running a simple application.
 
 TensorFlow is a powerful graph-modelling framework that can be used for the
 development, training and deployment of deep learning models. In the Graphcore

tensorflow/compiler/plugin/poplar/docs/poplar_ops.rst

@@ -17,9 +17,13 @@ for the op that can perform this padding, in addition to normalising and casting
 performed on-device, after the data has been transferred to the IPU.
 
 An example of its use can be found in the ``fused_normalise_image()`` function in the `CNN training application
-example <https://github.com/graphcore/examples/blob/master/vision/cnns/tensorflow1/training/Datasets/imagenet_preprocessing.py>`_
+example <https://github.com/graphcore/examples/blob/v3.0.0/vision/cnns/tensorflow1/training/Datasets/imagenet_preprocessing.py>`_
 in Graphcore's examples repository on GitHub.
 
+.. note:: This example is for TensorFlow 1.
+    From Poplar SDK 3.1, TensorFlow 1 will only be supported in CentOS 7. In addition, `Examples <https://github.com/graphcore/examples/tree/v3.0.0>`__ and `Tutorials <https://github.com/graphcore/tutorials/tree/sdk-release-3.0>`__ for TensorFlow 1 are only available up to version 3.0 of the SDK. There has been limited testing of the 3.0 versions of the TensorFlow 1 tutorials and examples with Poplar SDK 3.1.
+
+
 Matmul serialisation
 ~~~~~~~~~~~~~~~~~~~~
 
@@ -28,9 +32,13 @@ the code size of the multiplication and the temporary memory requirements of the
 
 See :py:func:`tensorflow.python.ipu.math_ops.serialized_matmul` for details of the op.
 
-An example of its use can be found in the ``mlm_head()`` function in the `BERT application example <https://github.com/graphcore/examples/blob/master/nlp/bert/tensorflow1/modeling.py>`_
+An example of its use can be found in the ``mlm_head()`` function in the `BERT application example <https://github.com/graphcore/examples/blob/v3.0.0/nlp/bert/tensorflow1/modeling.py>`_
 in Graphcore's examples repository on GitHub.
 
+.. note:: This example is for TensorFlow 1.
+    From Poplar SDK 3.1, TensorFlow 1 will only be supported in CentOS 7. In addition, `Examples <https://github.com/graphcore/examples/tree/v3.0.0>`__ and `Tutorials <https://github.com/graphcore/tutorials/tree/sdk-release-3.0>`__ for TensorFlow 1 are only available up to version 3.0 of the SDK. There has been limited testing of the 3.0 versions of the TensorFlow 1 tutorials and examples with Poplar SDK 3.1.
+
+
 Dropout
 ~~~~~~~