Bump version and link ASR notebook and tutorial.

PiperOrigin-RevId: 195755605

Author: Lukasz Kaiser

docs/cloud_tpu.md

@@ -18,6 +18,11 @@ See the official tutorial for [running Transformer
 on Cloud TPUs](https://cloud.google.com/tpu/docs/tutorials/transformer)
 for some examples and try out your own problems.
 
+You can train an Automatic Speech Recognition (ASR) model with Transformer
+on TPU by using `transformer` as `model` with `transformer_librispeech_tpu` as
+`hparams_set` and `librispeech` as `problem`. See this [tutorial](tutorials/ast_with_transformer.md) for more details on training it and this
+[notebook](https://colab.research.google.com/github/tensorflow/tensor2tensor/blob/master/tensor2tensor/notebooks/asr_transformer.ipynb) to see how the resulting model transcribes your speech to text.
+
 Image Transformer:
 * `imagetransformer` with `imagetransformer_base_tpu` (or
     `imagetransformer_tiny_tpu`)

docs/index.md

@@ -16,10 +16,14 @@ accessible and [accelerate ML
 research](https://research.googleblog.com/2017/06/accelerating-deep-learning-research.html).
 
 
-## Basics
+## Introduction
 
 * [Walkthrough](walkthrough.md): Install and run.
 * [IPython notebook](https://colab.research.google.com/github/tensorflow/tensor2tensor/blob/master/tensor2tensor/notebooks/hello_t2t.ipynb): Get a hands-on experience.
+* [Automatic Speech Recognition notebook](https://colab.research.google.com/github/tensorflow/tensor2tensor/blob/master/tensor2tensor/notebooks/asr_transformer.ipynb): Transcribe speech to text with a T2T model.
+
+## Basics
+
 * [Overview](overview.md): How all parts of T2T code are connected.
 * [New Problem](new_problem.md): Train T2T models on your data.
 * [New Model](new_model.md): Create your own T2T model.
@@ -29,6 +33,7 @@ research](https://research.googleblog.com/2017/06/accelerating-deep-learning-res
 * [Training on Google Cloud ML](cloud_mlengine.md)
 * [Training on Google Cloud TPUs](cloud_tpu.md)
 * [Distributed Training](distributed_training.md)
+# [Automatic Speech Recognition (ASR) with Transformer](tutorials/asr_with_transformer.md)
 
 ## Solving your task
 

docs/tutorials/asr_with_transformer.md

@@ -1,10 +1,13 @@
 # Automatic Speech Recognition (ASR) with Transformer
 
+Check out the [Automatic Speech Recognition notebook](https://colab.research.google.com/github/tensorflow/tensor2tensor/blob/master/tensor2tensor/notebooks/asr_transformer.ipynb) to see how the resulting model transcribes your speech to text.
+
 ## Data set
 
 This tutorial uses the publicly available
 [Librispeech](http://www.openslr.org/12/) ASR corpus.
 
+
 ## Generate the dataset
 
 To generate the dataset use `t2t-datagen`. You need to create environment

setup.py

@@ -5,7 +5,7 @@
 
 setup(
     name='tensor2tensor',
-    version='1.6.1',
+    version='1.6.2',
     description='Tensor2Tensor',
     author='Google Inc.',
     author_email='no-reply@google.com',