fix(serve): Update Nova Bedrock deployment notebook with working e2e flow

Simplify notebook to use existing completed training job with
BedrockModelBuilder deploy flow. Fix Nova inference content format
to use array of {text: ...} objects. Remove broken SFTTrainer cells
that fail due to botocore service model mismatch.

Author: nargokul

sagemaker-serve/example_notebooks/bedrock_nova_deployment.ipynb

@@ -4,26 +4,26 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Deploy a Nova Model to Amazon Bedrock\n",
+    "# Deploy a Fine-Tuned Nova Model to Amazon Bedrock\n",
     "\n",
-    "This notebook demonstrates how to fine-tune an Amazon Nova model using the SageMaker SDK\n",
-    "and deploy it to Amazon Bedrock using `BedrockModelBuilder`.\n",
+    "This notebook demonstrates how to deploy a fine-tuned Amazon Nova model to\n",
+    "Amazon Bedrock using `BedrockModelBuilder`.\n",
     "\n",
     "The workflow:\n",
-    "1. Fine-tune Nova Micro using `SFTTrainer`\n",
-    "2. Create a `BedrockModelBuilder` from the completed training job\n",
+    "1. Retrieve a completed Nova SFT training job\n",
+    "2. Create a `BedrockModelBuilder` from the training job\n",
     "3. Deploy to Bedrock — the builder automatically:\n",
     "   - Detects the model as Nova\n",
     "   - Reads the checkpoint URI from the training job manifest\n",
-    "   - Calls `CreateCustomModel`\n",
-    "   - Polls until the model is Active\n",
-    "   - Calls `CreateCustomModelDeployment`\n",
-    "   - Polls until the deployment is Active\n",
-    "4. Clean up resources\n",
+    "   - Calls `CreateCustomModel` and polls until Active\n",
+    "   - Calls `CreateCustomModelDeployment` and polls until Active\n",
+    "4. Test inference\n",
+    "5. Clean up resources\n",
     "\n",
     "**Prerequisites:**\n",
     "- AWS credentials with SageMaker and Bedrock access\n",
-    "- `sagemaker-serve` and `sagemaker-train` packages installed\n",
+    "- `sagemaker-serve` package installed\n",
+    "- A completed Nova SFT training job\n",
     "- An IAM role with Bedrock and SageMaker permissions"
    ]
   },
@@ -40,110 +40,24 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "import os\n",
-    "import json\n",
-    "import time\n",
-    "import random\n",
-    "import boto3\n",
+    "import os, json, time, random, boto3\n",
     "\n",
-    "# Set your region — Nova fine-tuning is available in us-east-1\n",
     "REGION = \"us-east-1\"\n",
     "os.environ[\"AWS_DEFAULT_REGION\"] = REGION\n",
     "os.environ[\"SAGEMAKER_REGION\"] = REGION\n",
     "\n",
     "from sagemaker.core.helper.session_helper import get_execution_role\n",
-    "\n",
     "role_arn = get_execution_role()\n",
-    "account_id = boto3.client(\"sts\").get_caller_identity()[\"Account\"]\n",
-    "bucket = f\"sagemaker-{REGION}-{account_id}\"\n",
-    "\n",
-    "print(f\"Region:  {REGION}\")\n",
-    "print(f\"Account: {account_id}\")\n",
-    "print(f\"Role:    {role_arn}\")\n",
-    "print(f\"Bucket:  {bucket}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Step 1: Prepare training data\n",
-    "\n",
-    "Upload a small JSONL dataset in the chat-messages format that Nova expects."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "s3 = boto3.client(\"s3\", region_name=REGION)\n",
-    "\n",
-    "# Ensure the bucket exists\n",
-    "try:\n",
-    "    s3.head_bucket(Bucket=bucket)\n",
-    "except Exception:\n",
-    "    s3.create_bucket(\n",
-    "        Bucket=bucket,\n",
-    "        CreateBucketConfiguration={\"LocationConstraint\": REGION},\n",
-    "    )\n",
-    "    print(f\"Created bucket: {bucket}\")\n",
-    "\n",
-    "train_key = \"nova-example/train.jsonl\"\n",
-    "train_uri = f\"s3://{bucket}/{train_key}\"\n",
-    "\n",
-    "rows = []\n",
-    "for i in range(50):\n",
-    "    rows.append(json.dumps({\n",
-    "        \"messages\": [\n",
-    "            {\"role\": \"user\", \"content\": f\"What is {i+1} + {i+1}?\"},\n",
-    "            {\"role\": \"assistant\", \"content\": f\"The answer is {(i+1)*2}.\"}\n",
-    "        ]\n",
-    "    }))\n",
-    "\n",
-    "s3.put_object(Bucket=bucket, Key=train_key, Body=\"\\n\".join(rows).encode())\n",
-    "print(f\"Uploaded {len(rows)} examples to {train_uri}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Step 2: Create model package group\n",
-    "\n",
-    "SFTTrainer requires a model package group to register the fine-tuned model.\n",
-    "We create one if it doesn't already exist."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "sm = boto3.client(\"sagemaker\", region_name=REGION)\n",
-    "\n",
-    "MODEL_PACKAGE_GROUP = f\"nova-example-{account_id}\"\n",
-    "\n",
-    "try:\n",
-    "    sm.describe_model_package_group(ModelPackageGroupName=MODEL_PACKAGE_GROUP)\n",
-    "    print(f\"Model package group already exists: {MODEL_PACKAGE_GROUP}\")\n",
-    "except sm.exceptions.ClientError:\n",
-    "    sm.create_model_package_group(\n",
-    "        ModelPackageGroupName=MODEL_PACKAGE_GROUP,\n",
-    "        ModelPackageGroupDescription=\"Nova fine-tuning example models\",\n",
-    "    )\n",
-    "    print(f\"Created model package group: {MODEL_PACKAGE_GROUP}\")"
+    "print(f\"Role: {role_arn}\")"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Step 3: Fine-tune Nova Micro with SFTTrainer\n",
+    "## Step 1: Retrieve the completed training job\n",
     "\n",
-    "This launches a SageMaker training job. It typically takes 15-30 minutes to complete."
+    "Use an existing completed Nova SFT training job. Replace the job name with your own."
    ]
   },
   {
@@ -152,19 +66,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "from sagemaker.train.sft_trainer import SFTTrainer\n",
-    "\n",
-    "trainer = SFTTrainer(\n",
-    "    model=\"nova-textgeneration-micro\",\n",
-    "    training_dataset=train_uri,\n",
-    "    accept_eula=True,\n",
-    "    model_package_group=MODEL_PACKAGE_GROUP,\n",
-    ")\n",
-    "\n",
-    "# Set wait=True to block until training completes\n",
-    "trainer.train(wait=True)\n",
+    "from sagemaker.core.resources import TrainingJob\n",
     "\n",
-    "training_job = trainer._latest_training_job\n",
+    "training_job = TrainingJob.get(training_job_name=\"nova-textgeneration-micro-sft-20251208154822\")\n",
     "print(f\"Training job: {training_job.training_job_name}\")\n",
     "print(f\"Status:       {training_job.training_job_status}\")"
    ]
@@ -173,7 +77,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Step 4: Deploy to Bedrock with BedrockModelBuilder\n",
+    "## Step 2: Deploy to Bedrock with BedrockModelBuilder\n",
     "\n",
     "The builder handles the full deployment flow:\n",
     "- Fetches the model package from the training job\n",
@@ -192,9 +96,8 @@
     "from sagemaker.serve.bedrock_model_builder import BedrockModelBuilder\n",
     "\n",
     "builder = BedrockModelBuilder(model=training_job)\n",
-    "\n",
-    "print(f\"Model package:    {builder.model_package}\")\n",
-    "print(f\"S3 artifacts:     {builder.s3_model_artifacts}\")"
+    "print(f\"Model package: {builder.model_package}\")\n",
+    "print(f\"S3 artifacts:  {builder.s3_model_artifacts}\")"
    ]
   },
   {
@@ -204,7 +107,7 @@
    "outputs": [],
    "source": [
     "rand = random.randint(1000, 9999)\n",
-    "custom_model_name = f\"nova-example-{rand}-{int(time.time())}\"\n",
+    "custom_model_name = f\"nova-e2e-{rand}-{int(time.time())}\"\n",
     "deployment_name = f\"{custom_model_name}-dep\"\n",
     "\n",
     "print(f\"Deploying as: {custom_model_name}\")\n",
@@ -225,9 +128,10 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Step 5: Test inference (optional)\n",
+    "## Step 3: Test inference\n",
     "\n",
-    "Once the deployment is Active, you can invoke it via the Bedrock Runtime API."
+    "Once the deployment is Active, invoke it via the Bedrock Runtime API.\n",
+    "Nova expects `content` as an array of objects with a `text` key."
    ]
   },
   {
@@ -238,32 +142,23 @@
    "source": [
     "bedrock_runtime = boto3.client(\"bedrock-runtime\", region_name=REGION)\n",
     "\n",
-    "# Get the model ARN from the deployment\n",
-    "bedrock = boto3.client(\"bedrock\", region_name=REGION)\n",
-    "dep_info = bedrock.get_custom_model_deployment(\n",
-    "    customModelDeploymentIdentifier=deployment_arn\n",
-    ")\n",
-    "model_arn = dep_info.get(\"modelArn\")\n",
-    "print(f\"Model ARN: {model_arn}\")\n",
-    "\n",
-    "# Invoke\n",
-    "invoke_response = bedrock_runtime.invoke_model(\n",
+    "resp = bedrock_runtime.invoke_model(\n",
     "    modelId=deployment_arn,\n",
     "    contentType=\"application/json\",\n",
     "    body=json.dumps({\n",
-    "        \"messages\": [{\"role\": \"user\", \"content\": \"What is 7 + 7?\"}]\n",
+    "        \"messages\": [{\"role\": \"user\", \"content\": [{\"text\": \"What is 7 + 7?\"}]}]\n",
     "    }),\n",
     ")\n",
     "\n",
-    "result = json.loads(invoke_response[\"body\"].read())\n",
-    "print(f\"Response: {result}\")"
+    "result = json.loads(resp[\"body\"].read())\n",
+    "print(f\"Response: {json.dumps(result, indent=2)}\")"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Step 6: Cleanup\n",
+    "## Step 4: Cleanup\n",
     "\n",
     "Delete the deployment and custom model to avoid ongoing charges."
    ]
@@ -276,23 +171,26 @@
    "source": [
     "bedrock = boto3.client(\"bedrock\", region_name=REGION)\n",
     "\n",
+    "dep_info = bedrock.get_custom_model_deployment(\n",
+    "    customModelDeploymentIdentifier=deployment_arn\n",
+    ")\n",
+    "model_arn = dep_info.get(\"modelArn\")\n",
+    "\n",
     "# Delete deployment first\n",
-    "if deployment_arn:\n",
-    "    try:\n",
-    "        bedrock.delete_custom_model_deployment(\n",
-    "            customModelDeploymentIdentifier=deployment_arn\n",
-    "        )\n",
-    "        print(f\"Deleted deployment: {deployment_arn}\")\n",
-    "    except Exception as e:\n",
-    "        print(f\"Failed to delete deployment: {e}\")\n",
+    "try:\n",
+    "    bedrock.delete_custom_model_deployment(\n",
+    "        customModelDeploymentIdentifier=deployment_arn\n",
+    "    )\n",
+    "    print(f\"Deleted deployment: {deployment_arn}\")\n",
+    "except Exception as e:\n",
+    "    print(f\"Failed to delete deployment: {e}\")\n",
     "\n",
     "# Then delete the custom model\n",
-    "if model_arn:\n",
-    "    try:\n",
-    "        bedrock.delete_custom_model(modelIdentifier=model_arn)\n",
-    "        print(f\"Deleted custom model: {model_arn}\")\n",
-    "    except Exception as e:\n",
-    "        print(f\"Failed to delete custom model: {e}\")"
+    "try:\n",
+    "    bedrock.delete_custom_model(modelIdentifier=model_arn)\n",
+    "    print(f\"Deleted custom model: {model_arn}\")\n",
+    "except Exception as e:\n",
+    "    print(f\"Failed to delete custom model: {e}\")"
    ]
   }
  ],
@@ -309,4 +207,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}