Fix new MXNet parameter for distributed training (aws#453)

laurenyu · web-flow · commit b6b9feea3ce1 · 2018-11-05T17:44:52.000-08:00
diff --git a/sagemaker-python-sdk/mxnet_gluon_sentiment/mxnet_sentiment_analysis_with_gluon.ipynb b/sagemaker-python-sdk/mxnet_gluon_sentiment/mxnet_sentiment_analysis_with_gluon.ipynb
@@ -113,7 +113,7 @@
     "          train_instance_type='ml.c4.2xlarge',\n",
     "          framework_version='1.3.0',\n",
     "          py_version='py2',\n",
-    "          launch_parameter_server=True,\n",
+    "          distributions={'parameter_server': {'enabled': True}},\n",
     "          hyperparameters={'batch-size': 8,\n",
     "                           'epochs': 2,\n",
     "                           'learning-rate': 0.01,\n",
diff --git a/sagemaker-python-sdk/mxnet_mnist/mxnet_mnist.ipynb b/sagemaker-python-sdk/mxnet_mnist/mxnet_mnist.ipynb
@@ -90,7 +90,7 @@
     "                        train_instance_count=1,\n",
     "                        train_instance_type='ml.m4.xlarge',\n",
     "                        framework_version='1.3.0',\n",
-    "                        launch_parameter_server=True,\n",
+    "                        distributions={'parameter_server': {'enabled': True}},\n",
     "                        hyperparameters={'learning-rate': 0.1})"
    ]
   },

Original file line number	Diff line number	Diff line change
`@@ -90,7 +90,7 @@`
`90`	`90`	`" train_instance_count=1,\n",`
`91`	`91`	`" train_instance_type='ml.m4.xlarge',\n",`
`92`	`92`	`" framework_version='1.3.0',\n",`
`93`		`- " launch_parameter_server=True,\n",`
	`93`	`+ " distributions={'parameter_server': {'enabled': True}},\n",`
`94`	`94`	`" hyperparameters={'learning-rate': 0.1})"`
`95`	`95`	`]`
`96`	`96`	`},`