Enable PT 2.4 Training Autopatch (#4462)

* Enable PT 2.4 Training Autopatch

* revert toml

Author: sirutBuasai

pytorch/training/buildspec-2-4-sm.yml

@@ -5,7 +5,7 @@ framework: &FRAMEWORK pytorch
 version: &VERSION 2.4.0
 short_version: &SHORT_VERSION "2.4"
 arch_type: x86
-autopatch_build: "False"
+autopatch_build: "True"
 
 repository_info:
   training_repository: &TRAINING_REPOSITORY
@@ -41,7 +41,7 @@ images:
     tag_python_version: &TAG_PYTHON_VERSION py311
     os_version: &OS_VERSION ubuntu22.04
     tag: !join [ *VERSION, "-", *DEVICE_TYPE, "-", *TAG_PYTHON_VERSION, "-", *OS_VERSION, "-sagemaker" ]
-    # latest_release_tag: !join [ *VERSION, "-", *DEVICE_TYPE, "-", *TAG_PYTHON_VERSION, "-", *OS_VERSION, "-sagemaker" ]
+    latest_release_tag: !join [ *VERSION, "-", *DEVICE_TYPE, "-", *TAG_PYTHON_VERSION, "-", *OS_VERSION, "-sagemaker" ]
     # build_tag_override: "beta:2.4.0-cpu-py311-ubuntu22.04-sagemaker"
     docker_file: !join [ docker/, *SHORT_VERSION, /, *DOCKER_PYTHON_VERSION, /Dockerfile., *DEVICE_TYPE ]
     target: sagemaker
@@ -57,7 +57,7 @@ images:
     cuda_version: &CUDA_VERSION cu124
     os_version: &OS_VERSION ubuntu22.04
     tag: !join [ *VERSION, "-", *DEVICE_TYPE, "-", *TAG_PYTHON_VERSION, "-", *CUDA_VERSION, "-", *OS_VERSION, "-sagemaker" ]
-    # latest_release_tag: !join [ *VERSION, "-", *DEVICE_TYPE, "-", *TAG_PYTHON_VERSION, "-", *CUDA_VERSION, "-", *OS_VERSION, "-sagemaker" ]
+    latest_release_tag: !join [ *VERSION, "-", *DEVICE_TYPE, "-", *TAG_PYTHON_VERSION, "-", *CUDA_VERSION, "-", *OS_VERSION, "-sagemaker" ]
     # build_tag_override: "beta:2.4.0-gpu-py311-cu124-ubuntu22.04-sagemaker"
     docker_file: !join [ docker/, *SHORT_VERSION, /, *DOCKER_PYTHON_VERSION, /, *CUDA_VERSION, /Dockerfile.,
                          *DEVICE_TYPE ]