feature: add support for PyTorch 1.8.0 training, inference, SMDDP, SMDMP

Author: saimidu

src/sagemaker/fw_utils.py

@@ -60,7 +60,7 @@
 )
 SM_DATAPARALLEL_SUPPORTED_FRAMEWORK_VERSIONS = {
     "tensorflow": ["2.3.0", "2.3.1", "2.4.1"],
-    "pytorch": ["1.6.0", "1.7.1"],
+    "pytorch": ["1.6.0", "1.7.1", "1.8.0"],
 }
 SMDISTRIBUTED_SUPPORTED_STRATEGIES = ["dataparallel", "modelparallel"]
 

src/sagemaker/image_uri_config/pytorch.json

@@ -55,7 +55,8 @@
             "1.4": "1.4.0",
             "1.5": "1.5.0",
             "1.6": "1.6.0",
-            "1.7": "1.7.1"
+            "1.7": "1.7.1",
+            "1.8": "1.8.0"
         },
         "versions": {
             "0.4.0": {
@@ -352,6 +353,39 @@
                     "us-west-2": "763104351884"
                 },
                 "repository": "pytorch-inference"
+            },
+            "1.8.0": {
+                "py_versions": [
+                    "py3",
+                    "py36"
+                ],
+                "registries": {
+                    "af-south-1": "626614931356",
+                    "ap-east-1": "871362719292",
+                    "ap-northeast-1": "763104351884",
+                    "ap-northeast-2": "763104351884",
+                    "ap-south-1": "763104351884",
+                    "ap-southeast-1": "763104351884",
+                    "ap-southeast-2": "763104351884",
+                    "ca-central-1": "763104351884",
+                    "cn-north-1": "727897471807",
+                    "cn-northwest-1": "727897471807",
+                    "eu-central-1": "763104351884",
+                    "eu-north-1": "763104351884",
+                    "eu-west-1": "763104351884",
+                    "eu-west-2": "763104351884",
+                    "eu-west-3": "763104351884",
+                    "eu-south-1": "692866216735",
+                    "me-south-1": "217643126080",
+                    "sa-east-1": "763104351884",
+                    "us-east-1": "763104351884",
+                    "us-east-2": "763104351884",
+                    "us-gov-west-1": "442386744353",
+                    "us-iso-east-1": "886529160074",
+                    "us-west-1": "763104351884",
+                    "us-west-2": "763104351884"
+                },
+                "repository": "pytorch-inference"
             }
         }
     },
@@ -369,7 +403,8 @@
             "1.4": "1.4.0",
             "1.5": "1.5.0",
             "1.6": "1.6.0",
-            "1.7": "1.7.1"	
+            "1.7": "1.7.1",
+            "1.8": "1.8.0"
         },
         "versions": {
             "0.4.0": {
@@ -667,6 +702,39 @@
                     "us-west-2": "763104351884"
                 },
                 "repository": "pytorch-training"
+            },
+            "1.8.0": {
+                "py_versions": [
+                    "py3",
+                    "py36"
+                ],
+                "registries": {
+                    "af-south-1": "626614931356",
+                    "ap-east-1": "871362719292",
+                    "ap-northeast-1": "763104351884",
+                    "ap-northeast-2": "763104351884",
+                    "ap-south-1": "763104351884",
+                    "ap-southeast-1": "763104351884",
+                    "ap-southeast-2": "763104351884",
+                    "ca-central-1": "763104351884",
+                    "cn-north-1": "727897471807",
+                    "cn-northwest-1": "727897471807",
+                    "eu-central-1": "763104351884",
+                    "eu-north-1": "763104351884",
+                    "eu-west-1": "763104351884",
+                    "eu-west-2": "763104351884",
+                    "eu-west-3": "763104351884",
+                    "eu-south-1": "692866216735",
+                    "me-south-1": "217643126080",
+                    "sa-east-1": "763104351884",
+                    "us-east-1": "763104351884",
+                    "us-east-2": "763104351884",
+                    "us-gov-west-1": "442386744353",
+                    "us-iso-east-1": "886529160074",
+                    "us-west-1": "763104351884",
+                    "us-west-2": "763104351884"
+                },
+                "repository": "pytorch-training"
             }
         }
     }

tests/unit/test_fw_utils.py

@@ -634,6 +634,7 @@ def test_validate_smdataparallel_args_not_raises():
         ("ml.p3.16xlarge", "tensorflow", "2.4.1", "py3", smdataparallel_enabled),
         ("ml.p3.16xlarge", "pytorch", "1.6.0", "py3", smdataparallel_enabled),
         ("ml.p3.16xlarge", "pytorch", "1.7.1", "py3", smdataparallel_enabled),
+        ("ml.p3.16xlarge", "pytorch", "1.8.0", "py3", smdataparallel_enabled),
     ]
     for instance_type, framework_name, framework_version, py_version, distribution in good_args:
         fw_utils._validate_smdataparallel_args(