Fix for aws#2949.

Hyperparameters are only json encoded at the end of setting up a Sagemaker job.

Author: Christian Osendorfer

src/sagemaker/estimator.py

@@ -2760,11 +2760,13 @@ def _prepare_init_params_from_job_description(cls, job_details, model_channel_na
         init_params = super(Framework, cls)._prepare_init_params_from_job_description(
             job_details, model_channel_name
         )
-
-        init_params["entry_point"] = init_params["hyperparameters"].get(SCRIPT_PARAM_NAME)
-        init_params["source_dir"] = init_params["hyperparameters"].get(DIR_PARAM_NAME)
-        init_params["container_log_level"] = init_params["hyperparameters"].get(
-            CONTAINER_LOG_LEVEL_PARAM_NAME)
+        init_params["entry_point"] = json.loads(
+            init_params["hyperparameters"].get(SCRIPT_PARAM_NAME)
+        )
+        init_params["source_dir"] = json.loads(init_params["hyperparameters"].get(DIR_PARAM_NAME))
+        init_params["container_log_level"] = json.loads(
+            init_params["hyperparameters"].get(CONTAINER_LOG_LEVEL_PARAM_NAME)
+        )
 
         hyperparameters = {}
         for k, v in init_params["hyperparameters"].items():

src/sagemaker/local/image.py

@@ -493,7 +493,7 @@ def _prepare_training_volumes(
         # If there is a training script directory and it is a local directory,
         # mount it to the container.
         if sagemaker.estimator.DIR_PARAM_NAME in hyperparameters:
-            training_dir = hyperparameters[sagemaker.estimator.DIR_PARAM_NAME]
+            training_dir = json.loads(hyperparameters[sagemaker.estimator.DIR_PARAM_NAME])
             parsed_uri = urlparse(training_dir)
             if parsed_uri.scheme == "file":
                 host_dir = os.path.abspath(parsed_uri.netloc + parsed_uri.path)
@@ -579,7 +579,7 @@ def _update_local_src_path(self, params, key):
             The updated parameters.
         """
         if key in params:
-            src_dir = params[key]
+            src_dir = json.loads(params[key])
             parsed_uri = urlparse(src_dir)
             if parsed_uri.scheme == "file":
                 new_params = params.copy()

tests/unit/test_estimator.py

@@ -1241,8 +1241,9 @@ def test_custom_code_bucket(time, sagemaker_session):
 
     expected_submit_dir = "s3://{}/{}".format(code_bucket, expected_key)
     _, _, train_kwargs = sagemaker_session.train.mock_calls[0]
-    assert train_kwargs["hyperparameters"]["sagemaker_submit_directory"] == expected_submit_dir
-
+    assert train_kwargs["hyperparameters"]["sagemaker_submit_directory"] == json.dumps(
+        expected_submit_dir
+    )
 
 @patch("time.strftime", return_value=TIMESTAMP)
 def test_custom_code_bucket_without_prefix(time, sagemaker_session):
@@ -1264,7 +1265,9 @@ def test_custom_code_bucket_without_prefix(time, sagemaker_session):
 
     expected_submit_dir = "s3://{}/{}".format(code_bucket, expected_key)
     _, _, train_kwargs = sagemaker_session.train.mock_calls[0]
-    assert train_kwargs["hyperparameters"]["sagemaker_submit_directory"] == expected_submit_dir
+    assert train_kwargs["hyperparameters"]["sagemaker_submit_directory"] == json.dumps(
+        expected_submit_dir
+    ) 
 
 
 def test_invalid_custom_code_bucket(sagemaker_session):
@@ -1336,9 +1339,11 @@ def test_shuffle_config(sagemaker_session):
 
 
 BASE_HP = {
-    "sagemaker_program": SCRIPT_NAME,
-    "sagemaker_submit_directory": "s3://mybucket/{}/source/sourcedir.tar.gz".format(JOB_NAME),
-    "sagemaker_job_name": JOB_NAME,
+    "sagemaker_program": json.dumps(SCRIPT_NAME),
+    "sagemaker_submit_directory": json.dumps(
+    "s3://mybucket/{}/source/sourcedir.tar.gz".format(JOB_NAME)
+    ),
+    "sagemaker_job_name": json.dumps(JOB_NAME),
 }
 
 
@@ -1383,8 +1388,8 @@ def test_start_new_convert_hyperparameters_to_str(strftime, sagemaker_session):
     t.fit("s3://{}".format(uri))
 
     expected_hyperparameters = BASE_HP.copy()
-    expected_hyperparameters["sagemaker_container_log_level"] = logging.INFO
-    expected_hyperparameters["learning_rate"] = 0.1
+    expected_hyperparameters["sagemaker_container_log_level"] = str(logging.INFO)
+    expected_hyperparameters["learning_rate"] = json.dumps(0.1)
     expected_hyperparameters["123"] = json.dumps([456])
     expected_hyperparameters["sagemaker_region"] = '"us-west-2"'
 
@@ -1407,7 +1412,7 @@ def test_start_new_wait_called(strftime, sagemaker_session):
     t.fit("s3://{}".format(uri))
 
     expected_hyperparameters = BASE_HP.copy()
-    expected_hyperparameters["sagemaker_container_log_level"] = logging.INFO
+    expected_hyperparameters["sagemaker_container_log_level"] = str(logging.INFO)
     expected_hyperparameters["sagemaker_region"] = '"us-west-2"'
 
     actual_hyperparameter = sagemaker_session.method_calls[1][2]["hyperparameters"]