Merge branch 'master' into lambda-model

Author: ahsan-z-khan

src/sagemaker/debugger/__init__.py

@@ -15,6 +15,7 @@
 
 from sagemaker.debugger.debugger import (  # noqa: F401
     CollectionConfig,
+    DEBUGGER_FLAG,
     DebuggerHookConfig,
     framework_name,
     get_default_profiler_rule,

src/sagemaker/debugger/debugger.py

@@ -32,6 +32,7 @@
 from sagemaker.utils import build_dict
 
 framework_name = "debugger"
+DEBUGGER_FLAG = "USE_SMDEBUG"
 
 
 def get_rule_container_image_uri(region):

src/sagemaker/estimator.py

@@ -29,6 +29,7 @@
 from sagemaker.analytics import TrainingJobAnalytics
 from sagemaker.debugger import TensorBoardOutputConfig  # noqa: F401 # pylint: disable=unused-import
 from sagemaker.debugger import (
+    DEBUGGER_FLAG,
     DebuggerHookConfig,
     FrameworkProfile,
     get_default_profiler_rule,
@@ -2269,6 +2270,11 @@ def _validate_and_set_debugger_configs(self):
                     )
                     self.debugger_hook_config = False
 
+        if self.debugger_hook_config is False:
+            if self.environment is None:
+                self.environment = {}
+            self.environment[DEBUGGER_FLAG] = "0"
+
     def _stage_user_code_in_s3(self):
         """Upload the user training script to s3 and return the location.
 

src/sagemaker/huggingface/estimator.py

@@ -70,10 +70,10 @@ def __init__(
                 ``image_uri`` is provided. The current supported version is ``4.6.1``.
             tensorflow_version (str): TensorFlow version you want to use for
                 executing your model training code. Defaults to ``None``. Required unless
-                ``pytorch_version`` is provided. The current supported version is ``1.6.0``.
+                ``pytorch_version`` is provided. The current supported version is ``2.4.1``.
             pytorch_version (str): PyTorch version you want to use for
                 executing your model training code. Defaults to ``None``. Required unless
-                ``tensorflow_version`` is provided. The current supported version is ``2.4.1``.
+                ``tensorflow_version`` is provided. The current supported versions are ``1.7.1`` and ``1.6.0``.
             source_dir (str): Path (absolute, relative or an S3 URI) to a directory
                 with any other training source code dependencies aside from the entry
                 point file (default: None). If ``source_dir`` is an S3 URI, it must

tests/integ/test_debugger.py

@@ -18,6 +18,7 @@
 import pytest
 
 from sagemaker.debugger.debugger import (
+    DEBUGGER_FLAG,
     DebuggerHookConfig,
     Rule,
     rule_configs,
@@ -748,6 +749,7 @@ def test_mxnet_with_debugger_hook_config_disabled(
         job_description = mx.latest_training_job.describe()
 
         assert job_description.get("DebugHookConfig") is None
+        assert job_description.get("Environment", {}).get(DEBUGGER_FLAG) == "0"
 
 
 def _get_rule_evaluation_statuses(job_description):

tests/unit/sagemaker/workflow/test_steps.py

@@ -23,7 +23,7 @@
     patch,
 )
 
-from sagemaker.debugger import ProfilerConfig
+from sagemaker.debugger import DEBUGGER_FLAG, ProfilerConfig
 from sagemaker.estimator import Estimator
 from sagemaker.tensorflow import TensorFlow
 from sagemaker.inputs import TrainingInput, TransformInput, CreateModelInput
@@ -275,6 +275,7 @@ def test_training_step_tensorflow(sagemaker_session):
                 "sagemaker_distributed_dataparallel_custom_mpi_options": '""',
             },
             "ProfilerConfig": {"S3OutputPath": "s3://my-bucket/"},
+            "Environment": {DEBUGGER_FLAG: "0"},
         },
         "CacheConfig": {"Enabled": True, "ExpireAfter": "PT1H"},
     }