Env migration from acpt-pytorch-2.2-cuda12.1 to acpt-pytorch-2.8-cuda12.6

cli/jobs/nebulaml/PyTorch_CNN_MNIST/job.yml

@@ -22,15 +22,15 @@ resources:
 #       https://learn.microsoft.com/en-us/azure/machine-learning/concept-environments
 #
 # for the latest ACPT image using PyTorch 2.2, CUDA 12.1 on Python 3.10
-# environment: azureml:AzureML-acpt-pytorch-2.2-cuda12.1@latest
-environment: azureml:AzureML-acpt-pytorch-2.2-cuda12.1@latest
+# environment: azureml:AzureML-acpt-pytorch-2.8-cuda12.6@latest
+environment: azureml:AzureML-acpt-pytorch-2.8-cuda12.6@latest
 # for a specific version
-#environment: azureml:AzureML-acpt-pytorch-2.2-cuda12.1:3
+#environment: azureml:AzureML-acpt-pytorch-2.8-cuda12.6:1
 # for the latest version of a custom environment:  (====install Nebula in dockerfile environment=====)
 #environment: azureml:ACPT-Extended@latest
 # for a specific Docker image (has to be compatible with Azure ML):
 #environment:
-#    image: mcr.microsoft.com/azureml/curated/acpt-pytorch-2.2-cuda12.1:2
+#    image: mcr.microsoft.com/azureml/curated/acpt-pytorch-2.8-cuda12.6:1
 
 # command that should be run by the job (====install Nebula with comand OR use dockerfile environment=====)
 command: >-

cli/jobs/nebulaml/PyTorch_CNN_MNIST/train.py

@@ -7,8 +7,6 @@
 from torchvision import datasets, transforms
 from torch.optim.lr_scheduler import StepLR
 
-import nebulaml as nm
-
 
 class Net(nn.Module):
     def __init__(self):
@@ -169,9 +167,6 @@ def main():
     else:
         device = torch.device("cpu")
 
-    # Nebula Initialization
-    nm.init(persistent_storage_path="/tmp/test", persistent_time_interval=2)
-
     train_kwargs = {"batch_size": args.batch_size}
     test_kwargs = {"batch_size": args.test_batch_size}
     if use_cuda:
@@ -197,10 +192,7 @@ def main():
         scheduler.step()
 
         if args.save_model:
-            # torch.save(model.state_dict(), "mnist_cnn.pt")
-            # Nebula Save
-            ckpt = nm.Checkpoint(f"global_step{epoch}", 1)
-            ckpt.save("mnist_cnn.pt", model.state_dict())
+            torch.save(model.state_dict(), "mnist_cnn.pt")
 
 
 if __name__ == "__main__":

cli/jobs/pipelines-with-components/basics/6b_pytorch_hello_world/component.yml

@@ -7,7 +7,7 @@ name: py_torch_hello_world
 display_name: PyTorch-hello-world
 version: 1
 
-environment: azureml://registries/azureml/environments/acpt-pytorch-2.2-cuda12.1/labels/latest
+environment: azureml://registries/azureml/environments/acpt-pytorch-2.8-cuda12.6/labels/latest
 command: >-
   echo "
   RANK: $RANK \n

cli/jobs/pipelines/cifar-10/pipeline.yml

@@ -33,7 +33,7 @@ jobs:
       model_dir:
         type: uri_folder
         mode: upload
-    environment: azureml://registries/azureml/environments/acpt-pytorch-2.2-cuda12.1/labels/latest
+    environment: azureml://registries/azureml/environments/acpt-pytorch-2.8-cuda12.6/labels/latest
     compute: azureml:gpu-cluster
     distribution:
       type: pytorch
@@ -47,7 +47,7 @@ jobs:
       --data-dir ${{inputs.cifar}}
       --model-dir ${{inputs.model_dir}}/model
     code: src/eval-model
-    environment: azureml://registries/azureml/environments/acpt-pytorch-2.2-cuda12.1/labels/latest
+    environment: azureml://registries/azureml/environments/acpt-pytorch-2.8-cuda12.6/labels/latest
     compute: azureml:gpu-cluster
     distribution:
       type: pytorch

cli/jobs/single-step/pytorch/cifar-distributed/job.yml

@@ -11,7 +11,7 @@ inputs:
   cifar:
      type: uri_folder
      path: azureml:cifar-10-example@latest
-environment: azureml:AzureML-acpt-pytorch-2.2-cuda12.1@latest
+environment: azureml:AzureML-acpt-pytorch-2.8-cuda12.6@latest
 compute: azureml:gpu-cluster
 distribution:
   type: pytorch

cli/jobs/single-step/pytorch/word-language-model/job.yml

@@ -16,7 +16,7 @@ inputs:
   corpus:
     path: data
     mode: download
-environment: azureml:AzureML-acpt-pytorch-2.2-cuda12.1@latest
+environment: azureml:AzureML-acpt-pytorch-2.8-cuda12.6@latest
 compute: azureml:gpu-cluster
 display_name: pytorch-word-language-model-example
 experiment_name: pytorch-word-language-model-example

sdk/python/jobs/pipelines/2b_train_cifar_10_with_pytorch/eval-model.yml

@@ -17,4 +17,4 @@ distribution:
   process_count_per_instance: 1
 resources:
   instance_count: 2
-environment: azureml://registries/azureml/environments/acpt-pytorch-2.2-cuda12.1/labels/latest
+environment: azureml://registries/azureml/environments/acpt-pytorch-2.8-cuda12.6/labels/latest

sdk/python/jobs/pipelines/2b_train_cifar_10_with_pytorch/train-model.yml

@@ -20,4 +20,4 @@ distribution:
   process_count_per_instance: 1
 resources:
   instance_count: 2
-environment: azureml://registries/azureml/environments/acpt-pytorch-2.2-cuda12.1/labels/latest
+environment: azureml://registries/azureml/environments/acpt-pytorch-2.8-cuda12.6/labels/latest

sdk/python/jobs/single-step/pytorch/distributed-training/distributed-cifar10.ipynb

@@ -354,7 +354,7 @@
     "    code=\"./src\",  # local path where the code is stored\n",
     "    command=\"python train.py --data-dir ${{inputs.cifar}} --epochs ${{inputs.epoch}} --batch-size ${{inputs.batchsize}} --workers ${{inputs.workers}} --learning-rate ${{inputs.lr}} --momentum ${{inputs.momen}} --print-freq ${{inputs.prtfreq}} --model-dir ${{inputs.output}}\",\n",
     "    inputs=inputs,\n",
-    "    environment=\"azureml:AzureML-acpt-pytorch-2.2-cuda12.1@latest\",\n",
+    "    environment=\"azureml:AzureML-acpt-pytorch-2.8-cuda12.6@latest\",\n",
     "    instance_count=2,  # In this, only 2 node cluster was created.\n",
     "    distribution={\n",
     "        \"type\": \"PyTorch\",\n",

sdk/python/jobs/single-step/pytorch/train-hyperparameter-tune-deploy-with-pytorch/train-hyperparameter-tune-deploy-with-pytorch.ipynb

@@ -199,7 +199,7 @@
    },
    "outputs": [],
    "source": [
-    "curated_env_name = \"AzureML-acpt-pytorch-2.2-cuda12.1@latest\""
+    "curated_env_name = \"AzureML-acpt-pytorch-2.8-cuda12.6@latest\""
    ]
   },
   {

tutorials/e2e-distributed-pytorch-image/e2e-object-classification-distributed-pytorch.ipynb

@@ -412,7 +412,7 @@
     "        # \"register_model_as\": \"places_dev\",\n",
     "        \"enable_profiling\": False,\n",
     "    },\n",
-    "    environment=\"AzureML-acpt-pytorch-2.2-cuda12.1@latest\",\n",
+    "    environment=\"AzureML-acpt-pytorch-2.8-cuda12.6@latest\",\n",
     "    compute=\"gpu-cluster\"\n",
     "    if (gpu_cluster)\n",
     "    else None,  # No compute needs to be passed to use serverless\n",