Welcome to Amazon SageMaker. This site highlights example Jupyter notebooks for a variety of machine learning use cases that you can run in SageMaker.
This site is based on the SageMaker Examples repository on GitHub. To run these notebooks, you will need a SageMaker Notebook Instance or SageMaker Studio. Refer to the SageMaker developer guide's Get Started page to get one of these set up.
On a Notebook Instance, the examples are pre-installed and available from the examples menu item in JupyterLab. On SageMaker Studio, you will need to open a terminal, go to your home folder, then clone the repo with the following:
git clone https://github.com/aws/amazon-sagemaker-examples.git
.. toctree:: :maxdepth: 1 :caption: Introduction intro.rst
We recommend the following notebooks as a broad introduction to the capabilities that SageMaker offers. To explore in even more depth, we provide additional notebooks covering even more use cases and frameworks.
.. toctree:: :maxdepth: 1 :caption: End to End ML Lifecycle end_to_end_ml_lifecycle/sm-autopilot_customer_churn end_to_end_ml_lifecycle/sm-autopilot_linear_regression_california_housing end_to_end_ml_lifecycle/sm-autopilot_time_series_forecasting
.. toctree:: :maxdepth: 1 :caption: Prepare data prepare_data/sm-data_wrangler_data_prep_widget/sm-data_wrangler_data_prep_widget prepare_data/sm-feature_store_feature_processor/sm-feature_store_feature_processor prepare_data/sm-feature_store_ground_truth_classification_output_to_store/sm-feature_store_ground_truth_classification_output_to_store prepare_data/sm-feature_store_introduction/sm-feature_store_introduction prepare_data/sm-ground_truth_active_learning_workflow_bring_your_own_model/sm-ground_truth_active_learning_workflow_bring_your_own_model prepare_data/sm-ground_truth_annotation_consolidation_image_classification/sm-ground_truth_annotation_consolidation_image_classification prepare_data/sm-ground_truth_object_detection_example/sm-ground_truth_object_detection_example prepare_data/sm-ground_truth_video_quality_metrics/sm-ground_truth_video_quality_metrics prepare_data/sm-marketplace_augmented_ai_with_marketplace_ml_models/sm-marketplace_augmented_ai_with_marketplace_ml_models prepare_data/sm-processing_feature_transformation_with_dask/sm-processing_feature_transformation_with_dask prepare_data/sm-processing_spark_processing/sm-processing_spark_processing prepare_data/sm-spark_pca_kmeans/sm-spark_pca_kmeans prepare_data/sm-ground_truth_3d_pointcloud_labeling prepare_data/sm-ground_truth_chained_streaming_labeling_job prepare_data/sm-ground_truth_create_streaming_labeling_job prepare_data/sm-ground_truth_labeling_adjustment_job_adaptation prepare_data/sm-ground_truth_object_detection_augmented_manifest_training prepare_data/sm-ground_truth_pretrained_model_labeling prepare_data/sm-ground_truth_rlhf_llm_finetuning prepare_data/sm-ground_truth_text_classification_labeling_accuracy_analysis prepare_data/sm-processing_introduction prepare_data/sm-processing_r prepare_data/sm-spark_kmeans
.. toctree:: :maxdepth: 1 :caption: Build and Train Models build_and_train_models/sm-distributed_data_parallelism_pytorch/sm-distributed_data_parallelism_pytorch build_and_train_models/sm-distributed_model_parallel/sm-distributed_model_parallel build_and_train_models/sm-forecast_deepar_time_series_modeling/sm-forecast_deepar_time_series_modeling build_and_train_models/sm-fsdp_finetuning_of_llama_v2/sm-fsdp_finetuning_of_llama_v2 build_and_train_models/sm-heterogeneous_clusters_for_model_training/sm-heterogeneous_clusters_for_model_training build_and_train_models/sm-heterogeneous_clusters_training/sm-heterogeneous_clusters_training build_and_train_models/sm-hyperparameter_tuning_pytorch/sm-hyperparameter_tuning_pytorch build_and_train_models/sm-introduction_to_blazingtext_word2vec_text8/sm-introduction_to_blazingtext_word2vec_text8 build_and_train_models/sm-introduction_to_ip_insights/sm-introduction_to_ip_insights build_and_train_models/sm-introduction_to_lda/sm-introduction_to_lda build_and_train_models/sm-introduction_to_ntm/sm-introduction_to_ntm build_and_train_models/sm-introduction_to_object2vec_sentence_similarity/sm-introduction_to_object2vec_sentence_similarity build_and_train_models/sm-jax_bring_your_own/sm-jax_bring_your_own build_and_train_models/sm-managed_spot_training_xgboost/sm-managed_spot_training_xgboost build_and_train_models/sm-marketplace_build_model_package_for_listing/sm-marketplace_build_model_package_for_listing build_and_train_models/sm-marketplace_building_your_own_container_as_package/sm-marketplace_building_your_own_container_as_package build_and_train_models/sm-object_detection_birds/sm-object_detection_birds build_and_train_models/sm-random_cut_forest_example/sm-random_cut_forest_example build_and_train_models/sm-regression_xgboost/sm-regression_xgboost build_and_train_models/sm-remote_function_pytorch_mnist/sm-remote_function_pytorch_mnist build_and_train_models/sm-remote_function_quick_start/sm-remote_function_quick_start build_and_train_models/sm-scikit_build_your_own_container/sm-scikit_build_your_own_container build_and_train_models/sm-script_mode_distributed_training_horovod_tensorflow/sm-script_mode_distributed_training_horovod_tensorflow build_and_train_models/sm-semantic_segmentation/sm-semantic_segmentation build_and_train_models/sm-smddp_bert/sm-smddp_bert build_and_train_models/sm-training_compiler_language_modeling_multi_gpu_multi_node/sm-training_compiler_language_modeling_multi_gpu_multi_node build_and_train_models/sm-training_compliler_single-node-single-gpu_bert/sm-training_compliler_single-node-single-gpu_bert build_and_train_models/sm-automatic_model_tune_hyperparameter_tuning_early_stopping build_and_train_models/sm-deepar_example build_and_train_models/sm-distributed_training_model_parallel_v2_mixtral_on_p4 build_and_train_models/sm-hpo_warmstart_image_classification build_and_train_models/sm-hyperparameter_tuning_hyperband_automatic_model_tuning build_and_train_models/sm-introduction_to_auogluon_tabular_regression build_and_train_models/sm-introduction_to_factorization_machines build_and_train_models/sm-introduction_to_pca build_and_train_models/sm-k_nearest_neighbors_multi_class_classification build_and_train_models/sm-lightgbm_catboost_tabular_classification build_and_train_models/sm-linear_learner_mnist build_and_train_models/sm-tabtransformer_tabular_classification
.. toctree:: :maxdepth: 1 :caption: Deploy and Monitor deploy_and_monitor/sm-a_b_testing/sm-a_b_testing deploy_and_monitor/sm-async_inference_walkthrough/sm-async_inference_walkthrough deploy_and_monitor/sm-async_inference_with_python_sdk/sm-async_inference_with_python_sdk deploy_and_monitor/sm-batch_inference_with_torchserve/sm-batch_inference_with_torchserve deploy_and_monitor/sm-batch_transform_pca_dbscan_movie_clusters/sm-batch_transform_pca_dbscan_movie_clusters deploy_and_monitor/sm-batch_transform_pytorch/sm-batch_transform_pytorch deploy_and_monitor/sm-batch_transform_with_torchserve/sm-batch_transform_with_torchserve deploy_and_monitor/sm-clarify_model_bias_monitor_batch_transform/sm-clarify_model_bias_monitor_batch_transform deploy_and_monitor/sm-clarify_model_bias_monitor_batch_transform_json/sm-clarify_model_bias_monitor_batch_transform_json deploy_and_monitor/sm-clarify_model_bias_monitor_for_endpoint/sm-clarify_model_bias_monitor_for_endpoint deploy_and_monitor/sm-clarify_model_bias_monitor_for_endpoint_json/sm-clarify_model_bias_monitor_for_endpoint_json deploy_and_monitor/sm-deployment_guardrails_update_inference_endpoint_with_linear_traffic_shifting/sm-deployment_guardrails_update_inference_endpoint_with_linear_traffic_shifting deploy_and_monitor/sm-deployment_guardrails_update_inference_endpoint_with_rolling_deployment/sm-deployment_guardrails_update_inference_endpoint_with_rolling_deployment deploy_and_monitor/sm-deployment_guardrails_update_inference_endpoint_with_with_canary_traffic_shifting/sm-deployment_guardrails_update_inference_endpoint_with_with_canary_traffic_shifting deploy_and_monitor/sm-host_pretrained_model_bert/sm-host_pretrained_model_bert deploy_and_monitor/sm-inference_pipeline_with_scikit_linear_learner/sm-inference_pipeline_with_scikit_linear_learner deploy_and_monitor/sm-lineage_cross_account_queries_with_ram/sm-lineage_cross_account_queries_with_ram deploy_and_monitor/sm-marketplace_using_model_package_arn/sm-marketplace_using_model_package_arn deploy_and_monitor/sm-mme_with_torchserve/sm-mme_with_torchserve deploy_and_monitor/sm-model_monitor_batch_transform_data_quality_on_schedule/sm-model_monitor_batch_transform_data_quality_on_schedule deploy_and_monitor/sm-model_monitor_batch_transform_model_quality_on_schedule/sm-model_monitor_batch_transform_model_quality_on_schedule deploy_and_monitor/sm-model_monitor_bias_and_explainability_monitoring/sm-model_monitor_bias_and_explainability_monitoring deploy_and_monitor/sm-model_monitor_introduction/sm-model_monitor_introduction deploy_and_monitor/sm-model_monitor_model_quality_monitoring/sm-model_monitor_model_quality_monitoring deploy_and_monitor/sm-multi_container_endpoint_direct_invocation/sm-multi_container_endpoint_direct_invocation deploy_and_monitor/sm-multi_model_endpoint_bring_your_own_container/sm-multi_model_endpoint_bring_your_own_container deploy_and_monitor/sm-serverless_inference_huggingface_text_classification/sm-serverless_inference_huggingface_text_classification deploy_and_monitor/sm-shadow_variant_shadow_api/sm-shadow_variant_shadow_api deploy_and_monitor/sm-triton_inferentia2/sm-triton_inferentia2 deploy_and_monitor/sm-triton_mme_bert_trt/sm-triton_mme_bert_trt deploy_and_monitor/sm-triton_mme_gpu_ensemble_dali/sm-triton_mme_gpu_ensemble_dali deploy_and_monitor/sm-triton_nlp_bert/sm-triton_nlp_bert deploy_and_monitor/sm-triton_realtime_sme_flan_t5/sm-triton_realtime_sme_flan_t5 deploy_and_monitor/sm-triton_tensorrt-sentence_transformer/sm-triton_tensorrt-sentence_transformer deploy_and_monitor/sm-xgboost_bring_your_own_model/sm-xgboost_bring_your_own_model deploy_and_monitor/sm-inference_recommender_introduction deploy_and_monitor/sm-serverless_inference deploy_and_monitor/sm-triton_realtime_sme deploy_and_monitor/sm-triton_tensorflow_model_deploy deploy_and_monitor/sm-app_autoscaling_realtime_endpoints/sm-app_autoscaling_realtime_endpoints deploy_and_monitor/sm-app_autoscaling_realtime_endpoints_inference_components/sm-app_autoscaling_realtime_endpoints_inference_components deploy_and_monitor/sm-app_autoscaling_realtime_endpoints_step_scaling/sm-app_autoscaling_realtime_endpoints_step_scaling deploy_and_monitor/sm-model_monitor_byoc_llm_monitor/sm-model_monitor_byoc_llm_monitor
.. toctree:: :maxdepth: 1 :caption: Generatative AI generative_ai/sm-finetuning_huggingface_with_your_own_scripts_and_data/sm-finetuning_huggingface_with_your_own_scripts_and_data generative_ai/sm-mixtral_8x7b_fine_tune_and_deploy/sm-mixtral_8x7b_fine_tune_and_deploy generative_ai/sm-djl_deepspeed_bloom_176b_deploy generative_ai/sm-fsdp_training_of_llama_v2_with_fp8_on_p5 generative_ai/sm-jumpstart_foundation_code_llama_fine_tuning_human_eval generative_ai/sm-jumpstart_foundation_finetuning_gpt_j_6b_domain_adaptation generative_ai/sm-jumpstart_foundation_gemma_fine_tuning generative_ai/sm-jumpstart_foundation_llama_2_finetuning generative_ai/sm-jumpstart_foundation_llama_3_finetuning generative_ai/sm-jumpstart_foundation_llama_3_text_completion generative_ai/sm-jumpstart_foundation_llama_3_8b_domain_adaption_finetuning.ipynb generative_ai/sm-jumpstart_foundation_llama_3_2_3b_finetuning.ipynb generative_ai/sm-jumpstart_foundation_llama_3_2_vision_language_finetuning.ipynb generative_ai/sm-jumpstart_foundation_llama_guard_text_moderation generative_ai/sm-jumpstart_foundation_mistral_7b_domain_adaption_finetuning generative_ai/sm-jumpstart_foundation_rag_langchain_question_answering generative_ai/sm-jumpstart_foundation_text_generation_inference generative_ai/sm-jumpstart_foundation_trainium_inferentia_finetuning_deployment generative_ai/sm-jumpstart_huggingface_text_classification generative_ai/sm-jumpstart_llama_2_chat_completion generative_ai/sm-jumpstart_llama_2_text_completion generative_ai/sm-jumpstart_rag_question_answering_with_cohere_and_langchain generative_ai/sm-jumpstart_stable_diffusion_text_to_image generative_ai/sm-jumpstart_text_embedding generative_ai/sm-jumpstart_time_series_forecasting
.. toctree:: :maxdepth: 1 :caption: ML Ops ml_ops/sm-ml_lineage_tracking_model_governance_graph/sm-ml_lineage_tracking_model_governance_graph ml_ops/sm-mlflow_deployment/sm-mlflow_deployment ml_ops/sm-mlflow_hpo/sm-mlflow_hpo ml_ops/sm-mlflow_pipelines/sm-mlflow_pipelines ml_ops/sm-mlflow_setup/sm-mlflow_setup ml_ops/sm-mlflow_training/sm-mlflow_training ml_ops/sm-model_monitor_batch_transform_data_quality_with_pipelines_on_demand/sm-model_monitor_batch_transform_data_quality_with_pipelines_on_demand ml_ops/sm-model_monitor_batch_transform_model_quality_with_pipelines_on_demand/sm-model_monitor_batch_transform_model_quality_with_pipelines_on_demand ml_ops/sm-pipelines_batch_inference_step_decorator/sm-pipelines_batch_inference_step_decorator ml_ops/sm-pipelines_callback_step/sm-pipelines_callback_step ml_ops/sm-pipelines_clarify_model_monitor_integration/sm-pipelines_clarify_model_monitor_integration ml_ops/sm-pipelines_emr_step_using_step_decorator/sm-pipelines_emr_step_using_step_decorator ml_ops/sm-pipelines_lambda_step/sm-pipelines_lambda_step ml_ops/sm-pipelines_launching_autopilot_with_auto_ml_step/sm-pipelines_launching_autopilot_with_auto_ml_step ml_ops/sm-pipelines_local_mode/sm-pipelines_local_mode ml_ops/sm-pipelines_preprocess_train_evaluate_batch_transform/sm-pipelines_preprocess_train_evaluate_batch_transform ml_ops/sm-pipelines_selective_execution/sm-pipelines_selective_execution ml_ops/sm-pipelines_step_caching/sm-pipelines_step_caching ml_ops/sm-pipelines_step_decorator/sm-pipelines_step_decorator ml_ops/sm-pipelines_step_decorator_with_classic_training_step/sm-pipelines_step_decorator_with_classic_training_step ml_ops/sm-pipelines_step_decorator_with_condition_step/sm-pipelines_step_decorator_with_condition_step ml_ops/sm-pipelines_emr_step_with_running_emr_cluster ml_ops/sm-pipelines_emr-step-with_cluster_lifecycle_management ml_ops/sm-pipelines_hyperparameter_tuning ml_ops/sm-pipelines_local_mode ml_ops/sm-pipelines_train_model_registry_deploy
.. toctree:: :maxdepth: 1 :caption: Responsible AI responsible_ai/sm-autopilot_model_explanation_with_shap/sm-autopilot_model_explanation_with_shap responsible_ai/sm-clarify_explainability_image_classification/sm-clarify_explainability_image_classification responsible_ai/sm-clarify_natural_language_processing_online_explainability/sm-clarify_natural_language_processing_online_explainability responsible_ai/sm-clarify_object_detection/sm-clarify_object_detection responsible_ai/sm-clarify_online_explainability_mme_xgboost/sm-clarify_online_explainability_mme_xgboost responsible_ai/sm-clarify_online_explainability_tabular/sm-clarify_online_explainability_tabular responsible_ai/sm-clarify_text_explainability_text_sentiment_analysis/sm-clarify_text_explainability_text_sentiment_analysis responsible_ai/sm-clarify_time_series_bring_your_own_model/sm-clarify_time_series_bring_your_own_model responsible_ai/sm-model_governance_model_card/sm-model_governance_model_card responsible_ai/sm-model_governance_model_card_with_model_package/sm-model_governance_model_card_with_model_package responsible_ai/sm-clarify_fairness_and_explainability_bring_your_own_container responsible_ai/sm-clarify_fairness_and_explainability_with_boto3