Update mixed_precision.ipynb

Updated mixed precision guide with with correct links

Author: chunduriv

site/en/guide/mixed_precision.ipynb

@@ -819,7 +819,7 @@
         "* Double the training batch size if it does not reduce evaluation accuracy\n",
         "* On GPUs, ensure most tensor dimensions are a multiple of $8$ to maximize performance\n",
         "\n",
-        "For more examples of mixed precision using the `tf.keras.mixed_precision` API, check the [official models repository](https://github.com/tensorflow/models/tree/master/official). Most official models, such as [ResNet](https://github.com/tensorflow/models/tree/master/official/vision/image_classification) and [Transformer](https://github.com/tensorflow/models/blob/master/official/nlp/transformer), will run using mixed precision by passing `--dtype=fp16`.\n"
+        "For more examples of mixed precision using the `tf.keras.mixed_precision` API, check the [official models repository](https://github.com/tensorflow/models/tree/master/official). Most official models, such as [ResNet](https://github.com/tensorflow/models/tree/master/official/legacy/image_classification) and [Transformer](https://github.com/tensorflow/models/tree/master/official/legacy/transformer), will run using mixed precision by passing `--dtype=fp16`.\n"
       ]
     }
   ],