serve
Serve, optimize and scale PyTorch models in production
파일 탐색기
최종 버전 다운로드 (.zip)- bug.yml
- documentation.yml
- feature-request.yml
- benchmark_nightly.yml
- benchmark_torch_compile_nightly.yml
- ci-cpu-cpp.yml
- ci_cpu.yml
- ci_gpu.yml
- ci_graviton_cpu.yml
- codeql.yml
- doc-automation.yml
- docker-ci.yaml
- docker-nightly-build.yml
- kserve_cpu_tests.yml
- kserve_gpu_tests.yml
- kubernetes_tests.yml
- lint.yml
- official_release.yml
- official_release_docker.yml
- regression_tests_cpu.yml
- regression_tests_cpu_binaries.yml
- regression_tests_docker.yml
- regression_tests_gpu.yml
- regression_tests_gpu_binaries.yml
- regression_tests_graviton_cpu.yml
- torchserve-nightly-build.yml
- dependabot.yml
- pull_request_template.md
- pytorch-probot.yml
- inf-tg.png
- inference.png
- register.png
- requestdefaults.png
- scaleup.png
- setup-tg.png
- teardown-tg.png
- testplan.png
- udv.png
- unregister.png
- batchImageInputModelPlan.jmx
- concurrentLoadPlan.jmx
- concurrentScaleCalls.jmx
- graphsGenerator.jmx
- imageInputModelPlan.jmx
- multipleModelsLoadPlan.jmx
- pingPlan.jmx
- bert_cpu.yaml
- bert_multi_gpu.yaml
- bert_multi_gpu_better_transformer.yaml
- bert_multi_gpu_no_better_transformer.yaml
- bert_neuron.yaml
- bert_neuronx.yaml
- bert_torch_compile_gpu.yaml
- fastrcnn.yaml
- mnist.yaml
- opt_6.7b_neuronx.yaml
- resnet50_torch_compile_gpu.yaml
- sam_fast_torch_compile_gpu_best_latency.yaml
- vgg16.yaml
- wf_dog_breed.yaml
- wf_nmt_dualtranslation.yaml
- wf_nmt_retranslation.yaml
- __init__.py
- benchmarks.py
- common.py
- gen_md_report.py
- gen_metrics_json.py
- gen_model_config_json.py
- locust_benchmark.py
- report.py
- reporting.py
- system_under_test.py
- testplans.py
- update_artifacts.py
- add_jmeter_test.md
- auto_benchmark.py
- benchmark-ab.py
- benchmark.py
- benchmark_config_cpu.yaml
- benchmark_config_gpu.yaml
- benchmark_config_neuron.yaml
- benchmark_config_neuronx.yaml
- benchmark_config_template.yaml
- benchmark_config_torch_compile_gpu.yaml
- benchmark_model_dependencies.sh
- config.properties
- config_kf.properties
- config_template.properties
- install_dependencies.sh
- jmeter.md
- mac_install_dependencies.sh
- predict_latency.png
- README.md
- requirements-ab.txt
- sample_report.md
- snake_viz.png
- upload_results_to_s3.sh
- validate_report.py
- windows_install_dependencies.py
- meta.yaml
- meta.yaml
- meta.yaml
- .gitignore
- build_packages.py
- README.md
- build.py
- install.py
- README.md
- s3_binary_upload.py
- upload.py
- __init__.py
- ec2.py
- launch_test.py
- requirements.txt
- buildspec.yml
- buildspec_cpu.yml
- buildspec_gpu.yml
- buildspec_windows.yml
- staging_packages_buildspec_cpu.yml
- staging_packages_buildspec_gpu.yml
- torchserve_nightly_build.yml
- backend.cc
- backend.hh
- model_instance.cc
- model_instance.hh
- base_handler.cc
- base_handler.hh
- handler_factory.hh
- torch_scripted_handler.cc
- torch_scripted_handler.hh
- model_worker.cc
- model_worker.hh
- model_worker_socket.cc
- isocket.hh
- otf_message.cc
- otf_message.hh
- socket.cc
- socket.hh
- CMakeLists.txt
- CMakeLists.txt
- logging.yaml
- cache.hh
- config.hh
- log_metric.cc
- log_metric.hh
- log_metrics_cache.cc
- log_metrics_cache.hh
- metric.hh
- registry.cc
- registry.hh
- units.cc
- units.hh
- yaml_config.cc
- yaml_config.hh
- CMakeLists.txt
- config.cc
- config.hh
- dl_loader.hh
- file_system.cc
- file_system.hh
- foo.cc
- foo.hh
- ifoo.hh
- json.cc
- json.hh
- logging.cc
- logging.hh
- message.hh
- model_archive.cc
- model_archive.hh
- mock_socket.hh
- otf_message_test.cc
- otf_protocol_and_handler_test.cc
- examples_test.cc
- MANIFEST.json
- index_to_name.json
- model-config.yaml
- sample_text.txt
- MANIFEST.json
- config.json
- MANIFEST.json
- 0_png.pt
- index_to_name.json
- model-config.yaml
- prompt.txt
- MANIFEST.json
- config.json
- config.properties
- prompt.txt
- MANIFEST.json
- config.json
- MANIFEST.json
- mnist_script.pt
- MANIFEST.json
- mnist_script.pt
- MANIFEST.json
- MANIFEST.json
- 0.png
- 0_png.pt
- invalid.yaml
- log_to_file.yaml
- default_config.yaml
- invalid_config_duplicate_dimension.yaml
- invalid_config_duplicate_metric_dimension.yaml
- invalid_config_duplicate_metric_name.yaml
- invalid_config_empty_dimension.yaml
- invalid_config_empty_metric_name.yaml
- invalid_config_missing_metric_name.yaml
- invalid_config_undefined_dimension.yaml
- log_to_file.yaml
- minimal_valid_config.yaml
- valid_config.yaml
- test.json
- torch_scripted_test.cc
- log_metric_test.cc
- log_metrics_cache_test.cc
- registry_test.cc
- units_test.cc
- yaml_config_test.cc
- common.hh
- dl_loader_test.cc
- json_test.cc
- logging_test.cc
- model_archiver_test.cc
- CMakeLists.txt
- .clang-tidy
- CMakeLists.txt
- Doxyfile
- install.cmake
- README.md
- build_image.sh
- build_upload_release.py
- config.properties
- docker_nightly.py
- dockerd-entrypoint.sh
- Dockerfile
- Dockerfile.cpp
- Dockerfile.dev
- Dockerfile.neuron.dev
- Dockerfile.rocm
- Dockerfile.vllm
- promote-docker.sh
- README.md
- start.sh
- test_build_image_tagging.sh
- test_container_health.sh
- test_container_model_prediction.sh
- test_container_python_version.sh
- dev_api.rst
- modules.rst
- run_circleci_tests.rst
- setup.rst
- test.rst
- torchserve_sanity.rst
- ts.metrics.rst
- ts.model_service.rst
- ts.protocol.rst
- ts.rst
- ts.torch_handler.request_envelope.rst
- ts.torch_handler.rst
- ts.torch_handler.unit_tests.models.rst
- ts.torch_handler.unit_tests.rst
- ts.torch_handler.unit_tests.test_utils.rst
- ts.utils.rst
- ts_scripts.rst
- amd_support.md
- apple_silicon_support.md
- hardware_support.rst
- linux_aarch64.md
- nvidia_mps.md
- plugins_sdk_class_uml_diagrams.puml
- 3dogs.jpg
- blank_image.jpg
- DogCatBreed.png
- dogs-after.jpg
- dogs-before.jpg
- kitten_small.jpg
- mps_g4_single.png
- mps_g4_two_worker.png
- mps_p3_single.png
- mps_p3_two_worker.png
- NMTBackTranslate.png
- NMTDualTranslate.png
- plugins_sdk_class_diagram.png
- project_structure.png
- SLA.jpg
- slack.png
- stateful_batch.jpg
- throughput.png
- tps_vs_delay.png
- ts-lmi-internal.png
- Makefile
- requirements.txt
- apis.rst
- batch_inference_with_ts.md
- code_coverage.md
- conf.py
- configuration.md
- contents.rst
- custom_directives.py
- custom_service.md
- default_handlers.md
- FAQs.md
- genai_use_cases.md
- getting_started.md
- github_actions.md
- grpc_api.md
- index.rst
- inference_api.md
- internals.md
- large_model_inference.md
- llm_deployment.md
- logging.md
- management_api.md
- metrics.md
- metrics_api.md
- model_api_control.md
- model_loading.md
- model_zoo.md
- performance_checklist.md
- performance_guide.md
- README.md
- request_envelopes.md
- rest_api.md
- server.md
- snapshot.md
- token_authorization_api.md
- torchserve_on_win_native.md
- torchserve_on_wsl.md
- Troubleshooting.md
- use_cases.md
- workflow_inference_api.md
- workflow_management_api.md
- workflows.md
- 00_save_jit_model.sh
- 01_create_model_archive.sh
- 02_configure_server.sh
- 03_predict.py
- 1089-134686-0000.wav
- 1089-134686.trans.txt
- config.properties
- handler.py
- README.md
- requirements.txt
- save_jit_model.py
- benchmark_profile.yaml
- model-config.yaml
- model.py
- README.md
- resnet50.yaml
- bert_captum_visualization.png
- bert_response.json
- Captum_visualization_for_bert.ipynb
- Captum_visualization_for_mnist.ipynb
- mnist_response.json
- config.properties
- index_to_name.json
- README.md
- setup_config.json
- stream_inference.py
- ec2-asg.yaml
- ec2.yaml
- README.md
- bert_handler.cc
- bert_handler.hh
- aot_compile_export.py
- CMakeLists.txt
- index_to_name.json
- model-config.yaml
- README.md
- LICENSE
- llama2.hh
- llama_handler.cc
- llama_handler.hh
- CMakeLists.txt
- compile.py
- config.json
- prompt1.txt
- prompt2.txt
- README.md
- resnet_handler.cc
- resnet_handler.hh
- CMakeLists.txt
- index_to_name.json
- model-config.yaml
- README.md
- resnet50_torch_export.py
- LICENSE
- llama2.h
- baby_llama_handler.cc
- baby_llama_handler.hh
- CMakeLists.txt
- config.json
- prompt1.txt
- prompt2.txt
- README.md
- llamacpp_handler.cc
- llamacpp_handler.hh
- CMakeLists.txt
- config.json
- prompt1.txt
- prompt2.txt
- README.md
- mnist_handler.cc
- mnist_handler.hh
- CMakeLists.txt
- ModelReady.java
- org.pytorch.serve.servingsdk.ModelServerEndpoint
- README.md
- config.properties
- metrics.yaml
- mnist_handler.py
- README.md
- img1.jpg
- img2.jpg
- img3.jpg
- create_mar.sh
- dcgan_fashiongen_handler.py
- Readme.md
- config.properties
- Download_model.py
- query.py
- Readme.md
- requirements.txt
- sample.txt
- sample_v1.json
- sample_v2.json
- stable_diffusion_handler.py
- Bert_FT_trace.py
- README.md
- sample_text.txt
- sample_text0.txt
- sample_text1.txt
- sample_text2.txt
- sample_text3.txt
- sample_text4.txt
- sample_text5.txt
- sample_text6.txt
- sample_text7.txt
- sample_text_captum_input.txt
- index_to_name.json
- sample_text.txt
- sample_text0.txt
- sample_text1.txt
- sample_text2.txt
- sample_text3.txt
- sample_text4.txt
- sample_text5.txt
- sample_text6.txt
- sample_text7.txt
- sample_text_captum_input.txt
- sample_text.txt
- sample_text_captum_input.txt
- index_to_name.json
- sample_text.txt
- sample_text_captum_input.txt
- bert_ts.json
- Download_Transformer_models.py
- index_to_name.json
- index_to_name_bert.json
- model-config.yaml
- README.md
- requirements.txt
- sample_text.txt
- torchscript.md
- Transformer_handler_generalized.py
- Transformer_handler_generalized_neuron.py
- model.py
- README.md
- model-config.yaml
- model.py
- README.md
- mnist_profiler_stats.png
- mnist_trace.png
- 0.png
- 1.png
- 2.png
- 3.png
- 4.png
- 5.png
- 6.png
- 7.png
- 8.png
- 9.png
- inference.py
- mnist_handler.py
- README.MD
- config.properties
- Docker.md
- mnist.py
- mnist_cnn.pt
- mnist_handler.py
- mnist_ts.json
- README.md
- arch.png
- cv_pipeline.png
- sample_video.mp4
- config.properties
- create_mar_file.py
- near_real_time_video_handler.py
- README.md
- request.py
- croco.jpg
- dog.jpg
- kitten.jpg
- index_to_name.json
- model.py
- README.md
- debugger_screenshot.png
- test_handler.py
- 2.c9f03951.chunk.css
- main.5e3647ae.chunk.css
- 2.792d00c1.chunk.js
- 2.792d00c1.chunk.js.LICENSE.txt
- 3.64cc616e.chunk.js
- main.2a22477d.chunk.js
- main.94f210aa.chunk.js
- runtime-main.8e9ee5a4.js
- asset-manifest.json
- favicon.ico
- index.html
- logo192.png
- logo512.png
- manifest.json
- robots.txt
- 1625836884119.png
- 1625898706091.png
- demo.gif
- README.md
- index_to_name.json
- model.py
- README.md
- model.py
- README.md
- model-config.yaml
- model.py
- README.md
- vgg_handler.py
- compile.json
- index_to_name.json
- kitten.jpg
- README.md
- deeplabv3.py
- fcn.py
- intermediate_layer_getter.py
- model.py
- README.md
- fcn.py
- intermediate_layer_getter.py
- model.py
- README.md
- persons.jpg
- README.md
- captum_logo.svg
- dlrm.png
- huggingface_logo-noborder.svg
- k8s.png
- kserve.png
- mii-white.svg
- mlflow.png
- mtail.png
- onnx.png
- prometheus-logo.svg
- seq2seq.png
- instructor_embedding_handler.py
- README.md
- requirements.txt
- intel_gpu.py
- intel_gpu_metric_collector.py
- README.md
- model-config.yaml
- custom_handler.py
- ds-config.json
- Readme.md
- requirements.txt
- sample_text.txt
- DeepSpeed_mii_handler.py
- mii-deepspeed-fastgen.ipynb
- model-config.yaml
- Readme.md
- requirements.txt
- sample.txt
- config.properties
- DeepSpeed_mii_handler.py
- deepspeed_mii_stable_diffusion.py
- Download_deepseed_mii_models.py
- query.py
- Readme.md
- requirements.txt
- setup_config.json
- config.properties
- diffusion_fast_handler.py
- Download_model.py
- model_config.yaml
- query.py
- README.md
- handler.py
- model_config.yaml
- model_config_speculative.yaml
- model_config_tp.yaml
- README.md
- request.json
- model_config.yaml
- model_config_tp.yaml
- README.md
- request.json
- config.properties
- custom_handler.py
- model-config.yaml
- Readme.md
- requirements.txt
- sample_text.txt
- config.properties
- custom_handler.py
- Download_model.py
- Readme.md
- requirements.txt
- sample_text.txt
- setup_config.json
- model-config.yaml
- pippy_handler.py
- Readme.md
- requirements.txt
- sample_text.txt
- base_neuronx_continuous_batching_handler.py
- inf2-llama-continuous-batching.ipynb
- llama2-model-config.yaml
- llama3-model-config.yaml
- Readme.md
- base_neuronx_microbatching_handler.py
- inf2-llama-micro-batching.ipynb
- llama2-model-config.yaml
- llama3-model-config.yaml
- Readme.md
- Readme.md
- requirements.txt
- inf2_handler.py
- model-config.yaml
- Readme.md
- requirements.txt
- sample_text.txt
- inf2_save_split_checkpoints.py
- config.properties
- llm_handler.py
- model-config-llama2-7b-bf16.yaml
- model-config-llama2-7b-int8-sq.yaml
- model-config-llama2-7b-int8-woq.yaml
- README.md
- requirements.txt
- sample_text_0.txt
- custom_handler.py
- inference.py
- install_segment_anything_fast.sh
- kitten.jpg
- kitten_mask_fast.png
- model-config.yaml
- README.md
- checkpoint_converter.py
- convert_checkpoints.py
- dialogs.txt
- generate.py
- llama2.py
- llama2_tokenizer.py
- llama_handler.py
- model-config.yaml
- README.md
- sample_text.txt
- model-config.yaml
- prompt.json
- README.md
- model-config.yaml
- prompt.json
- README.md
- Download_model.py
- test_llm_streaming_response.py
- chat.json
- model-config.yaml
- prompt.json
- Readme.md
- model-config.yaml
- prompt.json
- Readme.md
- model-config.yaml
- prompt.json
- Readme.md
- config.properties
- Readme.md
- requirements.txt
- build_image.sh
- client_app.py
- config.properties
- dockerd-entrypoint.sh
- Dockerfile
- Download_model.py
- llama_cpp_handler.py
- requirements.txt
- torchserve_server_app.py
- architecture.png
- batch_size.png
- Client.png
- Server.png
- Workers.png
- client_app.py
- llama_cpp_handler.py
- model-config.yaml
- package_llama.sh
- Readme.md
- requirements.txt
- torchserve_server_app.py
- llama.png
- software_stack_inf2.jpg
- README.md
- throughput_latency.png
- config.yaml
- micro_batching_handler.py
- README.md
- 372CC.info.json
- Generting_Charades_action_lables.py
- handler.py
- README.md
- request.py
- sample.txt
- sample1.txt
- sample2.txt
- sample3.txt
- sample4.txt
- config.properties
- create_mar.sh
- model_handler_generalized.py
- README.md
- requirements_script.py
- setup_config_script.py
- custom_handler.py
- model-config.yaml
- README.md
- requirements.txt
- serialize_dali_pipeline.py
- index_to_name.json
- model.py
- persons.jpg
- README.md
- model.py
- README.md
- bus.jpg
- custom_handler.py
- persons.jpg
- README.md
- requirements.txt
- index_to_name.json
- persons.jpg
- README.md
- model-config.yaml
- model.py
- README.md
- hpu_image_classifier.py
- model-config.yaml
- model.py
- README.md
- config.properties
- Download_model.py
- model-config.yaml
- pipeline_utils.py
- query.py
- README.md
- requirements.txt
- stable_diffusion_handler.py
- model-config.yaml
- model.py
- README.md
- requirements.txt
- model-config.yaml
- README.md
- resnet18_torch_export.py
- caching_handler.py
- model-config-cache-dir.yaml
- model-config-fx-cache.yaml
- README.md
- README.md
- archive_model.sh
- download_wav2vec2.py
- environment.yml
- handler.py
- README.md
- sample.wav
- sample1.txt
- sample2.txt
- sample3.txt
- model-config.yaml
- Readme.md
- stateful_handler.py
- model-config.yaml
- Readme.md
- stateful_handler.py
- config.properties
- model.py
- model_cnn.pt
- requirements.txt
- index_to_name.json
- model.py
- README.md
- run_script.py
- run_script.sh
- sample_text.txt
- spm_dataset.py
- text_classifier_ts.json
- train.py
- handler.py
- index_to_name.json
- README.md
- sample_text.txt
- script_tokenizer_and_model.py
- download_model.py
- model-config.yaml
- README.md
- sample_input.txt
- setup.sh
- text_to_speech_handler.py
- create_mar.sh
- README.md
- requirements.txt
- sample_text.txt
- waveglow_handler.py
- waveglow_model.py
- model-config.yaml
- model.py
- README.md
- requirements.txt
- config.properties
- model-config.yaml
- README.md
- requirements.txt
- sample_text.txt
- T5_handler.py
- README.md
- requirements.txt
- resnet_tensorrt.py
- create_dlrm_mar.py
- dlrm_factory.py
- dlrm_handler.py
- dlrm_model_config.py
- README.md
- sample_data.json
- benchmark_results.json
- image-eager-final.png
- image-tc_inductor-final.png
- image-tc_openvino-final.png
- profile-eager.txt
- profile-tc_inductor.txt
- profile-tc_openvino.txt
- pip_freeze_list.txt
- client-app-screen-1.png
- client-app-screen-2.png
- client-app-screen-3.png
- multi-image-gen-app.gif
- server-app-screen-1.png
- server-app-screen-2.png
- server-app-screen-3.png
- workflow-1.png
- workflow-2.png
- download_model_llm.py
- llm_handler.py
- model-config.yaml
- download_model_sd.py
- model-config.yaml
- requirements.txt
- stable_diffusion_handler.py
- build_image.sh
- client_app.py
- config.properties
- dockerd-entrypoint.sh
- Dockerfile
- requirements.txt
- sd-benchmark.py
- server_app.py
- README.md
- llm_ep.png
- rag_llm.png
- rag_perf.png
- benchmark_profile.yaml
- config.properties
- Deploy.md
- hf_custom_embeddings.py
- query_llama.py
- query_rag.py
- query_rag_llama.py
- rag-config.yaml
- rag.yaml
- rag_handler.py
- README.md
- requirements.txt
- sample_text.txt
- Cat.jpg
- Dog1.jpg
- Dog2.jpg
- cat_dog_classification.ipynb
- cat_dog_classification_arch.py
- cat_dog_classification_handler.py
- dog_breed_classification.ipynb
- dog_breed_classification_arch.py
- dog_breed_classification_handler.py
- index_to_name.json
- README.md
- workflow_dog_breed_classification.yaml
- workflow_dog_breed_classification_handler.py
- sample.txt
- sample1.txt
- sample2.txt
- sample3.txt
- sample4.txt
- config.properties
- nmt_workflow_dualtranslation.yaml
- nmt_workflow_handler_dualtranslation.py
- nmt_workflow_handler_retranslation.py
- nmt_workflow_retranslation.yaml
- README.md
- README.md
- index_to_name.json
- model-config.yaml
- README.md
- requirements.txt
- sample_input_1.txt
- sample_input_2.txt
- xgboost_iris_handler.py
- xgboost_train.py
- README.md
- AWS4SignerBase.java
- AWS4SignerForAuthorizationHeader.java
- BinaryUtils.java
- HttpUtils.java
- InvalidKeyException.java
- InvalidModelException.java
- Manifest.java
- ModelArchive.java
- ModelConfig.java
- ModelException.java
- ModelNotFoundException.java
- ModelVersionNotFoundException.java
- ArchiveUtils.java
- HexUtils.java
- InvalidArchiveURLException.java
- ZipUtils.java
- InvalidWorkflowException.java
- Manifest.java
- WorkflowArchive.java
- WorkflowException.java
- WorkflowNotFoundException.java
- DownloadArchiveException.java
- ModelArchiveTest.java
- ModelConfigTest.java
- ArchiveUtilsTest.java
- WorkFlowArchiveTest.java
- CoverageTest.java
- TestHelper.java
- invalid.yaml
- valid.yaml
- MANIFEST.json
- model.pt
- service.py
- MANIFEST.json
- model.pt
- service.py
- custom_invalid_python_dep.mar
- custom_python_dep.mar
- echo-client-timeout.mar
- echo.mar
- error_batch.mar
- init-error.mar
- loading-memory-error.mar
- mnist.mar
- mnist_scripted.mar
- mnist_scripted_cpp.mar
- mnist_traced.mar
- noop-customized.mar
- noop-no-manifest.mar
- noop-v1.0-config-tests.mar
- noop.mar
- noop.tar.gz
- noop_entrypntfunc.mar
- noop_handlername.mar
- noop_v2.mar
- pred-custom-return-code.mar
- prediction-memory-error.mar
- respheader-test.mar
- custom_invalid_python_dep.war
- custom_python_dep.war
- loading-memory-error.war
- prediction-memory-error.war
- retry_timeout_test.war
- smtest.war
- build.gradle
- gradlew
- testng.xml
- gradle-wrapper.jar
- gradle-wrapper.properties
- IAcceleratorUtility.java
- ICsvSmiParser.java
- IJsonSmiParser.java
- AppleUtil.java
- CudaUtil.java
- ROCmUtil.java
- XpuUtil.java
- Accelerator.java
- AcceleratorVendor.java
- SystemInfo.java
- Dag.java
- DagExecutor.java
- InvalidDAGException.java
- Node.java
- NodeOutput.java
- WorkFlow.java
- WorkflowInt.java
- WorkflowManifest.java
- WorkflowModel.java
- GRPCInterceptor.java
- GRPCServiceFactory.java
- InferenceImpl.java
- ManagementImpl.java
- OpenInferenceProtocolImpl.java
- ApiDescriptionRequestHandler.java
- InferenceRequestHandler.java
- ManagementRequestHandler.java
- OpenInferenceProtocolRequestHandler.java
- PrometheusMetricsRequestHandler.java
- DescribeModelResponse.java
- KFV1ModelReadyResponse.java
- ListModelsResponse.java
- RegisterModelRequest.java
- BadRequestException.java
- ConflictStatusException.java
- ErrorResponse.java
- ExtendedSSLHandler.java
- HttpRequestHandler.java
- HttpRequestHandlerChain.java
- InternalServerException.java
- InvalidModelVersionException.java
- InvalidPluginException.java
- InvalidRequestHandler.java
- MethodNotAllowedException.java
- RequestTimeoutException.java
- ResourceNotFoundException.java
- ServiceUnavailableException.java
- Session.java
- StatusResponse.java
- TokenAuthorizationHandler.java
- GRPCJob.java
- Job.java
- JobGroup.java
- RestJob.java
- MetricConfiguration.java
- MetricSpecification.java
- MetricTypes.java
- PrometheusCounter.java
- PrometheusGauge.java
- PrometheusHistogram.java
- Dimension.java
- IMetric.java
- LogMetric.java
- Metric.java
- MetricBuilder.java
- MetricCache.java
- MetricCollector.java
- MetricManager.java
- Encoding.java
- Info.java
- MediaType.java
- OpenApi.java
- OpenApiUtils.java
- Operation.java
- Parameter.java
- Path.java
- PathParameter.java
- QueryParameter.java
- RequestBody.java
- Response.java
- Schema.java
- ModelServerContext.java
- ModelServerModel.java
- ModelServerRequest.java
- ModelServerResponse.java
- ModelWorker.java
- PluginsManager.java
- FSSnapshotSerializer.java
- InvalidSnapshotException.java
- SnapshotManager.java
- SnapshotReadException.java
- SnapshotSerializerFactory.java
- CodecUtils.java
- ModelRequestEncoder.java
- ModelResponseDecoder.java
- JSONPatternLayout.java
- QLogLayout.java
- BaseModelRequest.java
- EnvironmentUtils.java
- InputParameter.java
- ModelInferenceRequest.java
- ModelLoadModelRequest.java
- ModelWorkerResponse.java
- Predictions.java
- RequestInput.java
- WorkerCommands.java
- ApiUtils.java
- ConfigManager.java
- Connector.java
- ConnectorType.java
- GRPCUtils.java
- JsonUtils.java
- NettyUtils.java
- OpenSslKey.java
- OSUtils.java
- ServerGroups.java
- TokenAuthorization.java
- AsyncBatchAggregator.java
- AsyncWorkerThread.java
- BatchAggregator.java
- ContinuousBatching.java
- Model.java
- ModelManager.java
- ModelVersionedRefs.java
- ModelVersionName.java
- SequenceBatching.java
- SequenceContinuousBatching.java
- WorkerInitializationException.java
- WorkerLifeCycle.java
- WorkerState.java
- WorkerStateListener.java
- WorkerThread.java
- WorkLoadManager.java
- WorkflowInferenceRequestHandler.java
- WorkflowMgmtRequestHandler.java
- DescribeWorkflowResponse.java
- ListWorkflowResponse.java
- ModelRegistrationResult.java
- RegisterWorkflowRequest.java
- WorkflowManager.java
- ModelServer.java
- ServerInitializer.java
- inference.proto
- management.proto
- open_inference_grpc.proto
- log4j.properties
- log4j2.xml
- AppleUtilTest.java
- CudaUtilTest.java
- ROCmUtilTest.java
- XpuUtilTest.java
- AcceleratorTest.java
- SystemInfoTest.java
- MetricCacheTest.java
- MetricConfigurationTest.java
- MetricTest.java
- TestHelper.java
- ModelRequestEncoderTest.java
- ConfigManagerTest.java
- ConnectorTest.java
- PluginSdkTest.java
- WorkerLifeCycleTest.java
- CoverageTest.java
- EnsembleTest.java
- ModelServerTest.java
- SnapshotTest.java
- TestUtils.java
- WorkflowTest.java
- invalid_configuration_missing_dimension.yaml
- invalid_configuration_missing_metric_name.yaml
- invalid_configuration_missing_metric_unit.yaml
- sample_amd_discovery.json
- sample_amd_metrics.json
- sample_amd_updated_metrics.json
- sample_apple_smi.json
- valid_configuration.yaml
- valid_configuration_empty_metric_dimensions.yaml
- snapshot1.cfg
- snapshot2.cfg
- snapshot3.cfg
- snapshot4.cfg
- snapshot5.cfg
- snapshot6.cfg
- snapshot7.cfg
- snapshot8.cfg
- snapshot9.cfg
- certs.pem
- classification_multiout_agg.py
- classification_multiout_agg.yaml
- classification_multiout_handler.py
- classification_multiout_workflow.yaml
- config.properties
- config_snapshot.properties
- config_test_env.properties
- config_test_workflow.properties
- densenet_workflow.yaml
- densenet_workflow_handler.py
- describe_api.json
- inference_open_api.json
- key.pem
- keystore.p12
- log4j2-json.xml
- log4j2-qlog.xml
- management_open_api.json
- metrics_default.yaml
- metrics_open_api.json
- model_management_api.json
- build.gradle
- gradlew
- testng.xml
- checkstyle.xml
- findbugs-exclude.xml
- pmd.xml
- suppressions.xml
- check.gradle
- formatter.gradle
- launcher.gradle
- proto.gradle
- spotbugs.gradle
- .gitignore
- build.gradle
- gradle.properties
- gradlew
- gradlew.bat
- README.md
- settings.gradle
- AKS_pv_claim.yaml
- Azure_file_sc.yaml
- model_store_pod.yaml
- nvidia-device-plugin-ds.yaml
- config.properties
- README.md
- efs_mount.png
- eks_cfn.png
- overview.png
- security_group.png
- efs_pv_claim.yaml
- eks_cluster.yaml
- config.properties
- README.md
- setup_efs.sh
- deployment.yaml
- MNIST.md
- service.yaml
- FasterTransformer_HuggingFace_Bert.md
- _helpers.tpl
- deployment.yaml
- NOTES.txt
- service.yaml
- .helmignore
- Chart.yaml
- values.yaml
- pod.yaml
- pv_pvc.yaml
- config.properties
- README.md
- torchserve.yaml
- .helmignore
- Chart.yaml
- values.yaml
- efs_mount.png
- eks_cfn.png
- grafana_datasource.png
- overview.png
- security_group.png
- Dockerfile
- llama2_v1_gpu.yaml
- README.md
- sample_text.json
- MNIST.md
- __init__.py
- __main__.py
- image_transformer.py
- transformer_model_repository.py
- input.json
- README.md
- setup.py
- transformer.Dockerfile
- bert.json
- img2bytearray.py
- mnist.json
- README.md
- text_classifier.json
- bert_bytes_v2.json
- bert_tensor_v2.json
- bert_tokenizer.py
- README.md
- tobytes.py
- Transformer_kserve_handler.py
- 0.png
- mnist_v2_bytes.json
- mnist_v2_bytes_grpc.json
- mnist_v2_tensor.json
- mnist_v2_tensor_grpc.json
- README.md
- tobytes.py
- totensor.py
- utils.py
- __main__.py
- gprc_utils.py
- README.md
- TorchserveModel.py
- TSModelRepository.py
- pv.yaml
- pvc.yaml
- pvc.yaml
- pv.yaml
- ts-sample-gpu.yaml
- ts_sample-gs.yaml
- ts_sample.yaml
- ts-sample-gpu.yaml
- ts_sample-gs.yaml
- ts_sample.yaml
- cluster.yaml
- pv_pod.yaml
- mnist_oip_grpc.yaml
- mnist_oip_http.yaml
- mnist_v1_cpu.yaml
- mnist_v1_gpu.yaml
- mnist_v2_cpu.yaml
- mnist_v2_gpu.yaml
- test_mnist.sh
- build_image.sh
- build_upload_release.py
- config.properties
- developer_guide.md
- docker_nightly.py
- dockerd-entrypoint.sh
- Dockerfile
- Dockerfile.dev
- README.md
- requirements.txt
- deployment.yaml
- config.properties
- Dockerfile
- test_mnist.sh
- adapter.yaml
- autoscale.md
- destination_rule.yaml
- gateway.yaml
- hpa.yaml
- README.md
- virtual_service.yaml
- __init__.py
- manifest.py
- model.py
- 1.py
- destination_vocab.pt
- dummy-artifacts.txt
- service.py
- source_vocab.pt
- test_handler.py
- test_index_to_name.json
- test_model.py
- test_serialized_file.pt
- configuration.json
- default_handler_configuration.json
- test_integration_model_archiver.py
- test_model_archiver.py
- test_model_packaging.py
- test_model_packaging_utils.py
- test_version.py
- conftest.py
- pylintrc
- __init__.py
- arg_parser.py
- model_archiver.py
- model_archiver_config.py
- model_archiver_error.py
- model_packaging.py
- model_packaging_utils.py
- version.py
- version.txt
- .coveragerc
- LICENSE
- MANIFEST.in
- PyPiDescription.rst
- README.md
- setup.py
- DDBClient.java
- DDBSnapshot.java
- DDBSnapshotSerializer.java
- org.pytorch.serve.servingsdk.snapshot.SnapshotSerializer
- DDBSnapshotSerializerTest.java
- LocalDynamoDB.java
- build.gradle
- testng.xml
- ddb_endpoint.md
- README.md
- ExecutionParameters.java
- Ping.java
- org.pytorch.serve.servingsdk.ModelServerEndpoint
- build.gradle
- gradle-wrapper.jar
- gradle-wrapper.properties
- checkstyle.xml
- findbugs-exclude.xml
- pmd.xml
- suppressions.xml
- check.gradle
- formatter.gradle
- spotbugs.gradle
- build.gradle
- gradle.properties
- gradlew
- gradlew.bat
- settings.gradle
- common.txt
- developer.txt
- neuronx.txt
- torch_common.txt
- torch_cu101_linux.txt
- torch_cu101_windows.txt
- torch_cu102_linux.txt
- torch_cu102_windows.txt
- torch_cu111_linux.txt
- torch_cu111_windows.txt
- torch_cu113_linux.txt
- torch_cu113_windows.txt
- torch_cu116_linux.txt
- torch_cu116_windows.txt
- torch_cu117_linux.txt
- torch_cu117_windows.txt
- torch_cu118_linux.txt
- torch_cu118_windows.txt
- torch_cu121_linux.txt
- torch_cu121_windows.txt
- torch_cu92.txt
- torch_cu92_linux.txt
- torch_darwin.txt
- torch_linux.txt
- torch_linux_aarch64.txt
- torch_neuronx_linux.txt
- torch_rocm60.txt
- torch_rocm61.txt
- torch_rocm62.txt
- torch_windows.txt
- trt_llm.txt
- maven-wrapper.jar
- maven-wrapper.properties
- EndpointTypes.java
- Endpoint.java
- Request.java
- Response.java
- BaseDimension.java
- BaseMetric.java
- DimensionRegistry.java
- InbuiltMetricsRegistry.java
- MetricEventListener.java
- MetricEventListenerRegistry.java
- MetricEventPublisher.java
- MetricLogEvent.java
- MetricPluginException.java
- Snapshot.java
- SnapshotSerializer.java
- Context.java
- Model.java
- ModelServerEndpoint.java
- ModelServerEndpointException.java
- Worker.java
- ModelServerEndpointTest.java
- checkstyle.xml
- mvnw
- mvnw.cmd
- pom.xml
- environment.json
- explanation_api_test_collection.json
- explanation_data.json
- https_test_collection.json
- increased_timeout_inference.json
- inference_api_test_collection.json
- inference_data.json
- inference_stream2_data.json
- inference_stream_data.json
- kf_api_test_collection.json
- kf_https_test_collection.json
- kf_inference_api_test_collection.json
- kf_inference_data.json
- kfv2_api_test_collection.json
- kfv2_https_test_collection.json
- kfv2_inference_data.json
- large_model_inference_data.json
- management_api_test_collection.json
- management_data.json
- workflow_api_test_collection.json
- workflow_data.json
- workflow_inference_collection.json
- workflow_inference_data.json
- resnet_custom.py
- resnet_profiler_override.py
- conftest.py
- test_config_snapshot.py
- test_model_registering.py
- config.properties
- mnist_custom_dependencies_handler.py
- model_config.yaml
- requirements.txt
- metrics_auto_detect.yaml
- __init__.py
- fake_streaming_model.py
- model_config.yaml
- stream.ipynb
- stream_handler.py
- compile_handler.py
- config.properties
- model.py
- pt2.yaml
- pt2_dict.yaml
- pt2_enable_default.yaml
- pt2_enable_false.yaml
- pt2_enable_true.yaml
- xla.yaml
- 0.png
- __init__.py
- dlrm_bs_1.json
- dlrm_bs_2.json
- dlrm_model.py
- kitten.jpg
- conftest.py
- onnx_handler.py
- test_async_worker_comm.py
- test_auto_recover.py
- test_benchmark.py
- test_continuous_batching.py
- test_cpp_backend.py
- test_dali_preprocess.py
- test_device_config.py
- test_distributed_inference_handler.py
- test_example_dcgan.py
- test_example_gpt_fast.py
- test_example_intel_extension_for_pytorch.py
- test_example_micro_batching.py
- test_example_near_real_time_video.py
- test_example_scriptable_tokenzier.py
- test_example_stateful_sequence_batching_http.py
- test_example_stateful_sequence_continuous_batching_http.py
- test_example_torch_compile.py
- test_example_torchrec_dlrm.py
- test_example_vllm.py
- test_gRPC_allowed_urls.py
- test_gRPC_inference_api.py
- test_gRPC_management_apis.py
- test_gRPC_model_control.py
- test_gRPC_token_authorization.py
- test_gRPC_utils.py
- test_handler.py
- test_handler_traceback_logging.py
- test_ipex_serving.py
- test_metrics.py
- test_metrics_kf.py
- test_mnist_template.py
- test_model_archiver.py
- test_model_config.py
- test_model_control_mode.py
- test_model_custom_dependencies.py
- test_onnx.py
- test_parallelism.py
- test_pytorch_profiler.py
- test_send_intermediate_prediction_response.py
- test_sm_mme_requirements.py
- test_snapshot.py
- test_startup_timeout.py
- test_token_authorization.py
- test_torch_compile.py
- test_torch_export.py
- test_torch_xla.py
- test_tp_llama.py
- test_url_path.py
- test_utils.py
- mnist.mar
- smtest.war
- certs.pem
- config.properties
- config_https.properties
- config_kf.properties
- config_kfv2.properties
- config_model_mode.properties
- config_token.properties
- config_token_priority.properties
- config_token_priority2.properties
- config_token_workflow.properties
- key.pem
- postman.png
- __init__.py
- buildspec.yml
- buildspec_gpu.yml
- buildspec_windows.yml
- config_ipex.properties
- config_kf.properties
- config_ts.properties
- data_file_config.md
- README.md
- regression_tests.py
- rpc
- metrics.yaml
- sagemaker_config.properties
- __init__.py
- deepspeed.py
- pt_pippy.py
- __init__.py
- dali_pipeline_generation.py
- default.dali
- Readme.md
- requirements.txt
- __init__.py
- dali.py
- __init__.py
- hf_batch_streamer.py
- micro_batching.py
- text_utils.py
- timer.py
- utils.py
- __init__.py
- caching_metric.py
- dimension.py
- metric.py
- metric_abstract.py
- metric_cache_abstract.py
- metric_cache_errors.py
- metric_cache_yaml_impl.py
- metric_collector.py
- metric_encoder.py
- metric_type_enum.py
- metrics_store.py
- process_memory_metric.py
- system_metrics.py
- unit.py
- __init__.py
- model_service.py
- __init__.py
- otf_message_handler.py
- metric_cache_unit_test.py
- metrics.yaml
- metrics_api.yaml
- metrics_empty_dims.yml
- metrics_empty_fields.yml
- metrics_mismatching_dims.yaml
- metrics_missing_types.yaml
- metrics_model_empty.yaml
- metrics_wo_model_metrics.yaml
- dummy_model_service.py
- MANIFEST.json
- MANIFEST.json
- dummy_class_model_service.py
- dummy_func_model_service.py
- test_backend_metric.py
- test_decorator.py
- test_handler_utils.py
- test_hf_batch_streamer.py
- test_micro_batching.py
- test_model_loader.py
- test_model_service_worker.py
- test_otf_codec_protocol.py
- test_version.py
- test_worker_service.py
- pylintrc
- README.md
- __init__.py
- base_deepspeed_handler.py
- base_pippy_handler.py
- __init__.py
- base.py
- body.py
- json.py
- kserve.py
- kservev2.py
- .gitignore
- __init__.py
- base_model.py
- __init__.py
- mock_context.py
- model_dir.py
- __init__.py
- conftest.py
- test_base_handler.py
- test_envelopes.py
- test_image_classifier.py
- test_image_segmenter.py
- test_mnist_kf.py
- test_object_detector.py
- __init__.py
- base_handler.py
- contractions.py
- dali_handler.py
- dali_image_classifier.py
- densenet_handler.py
- image_classifier.py
- image_segmenter.py
- object_detector.py
- text_classifier.py
- text_handler.py
- trt_llm_handler.py
- vision_handler.py
- vllm_handler.py
- __init__.py
- hf_utils.py
- setup_model_venv.py
- timeit_decorator.py
- util.py
- .gitignore
- __init__.py
- arg_parser.py
- async_service.py
- context.py
- launcher.py
- llm_launcher.py
- model_loader.py
- model_server.py
- model_service_worker.py
- service.py
- version.py
- version.txt
- sanity_models.json
- spellcheck.yaml
- wordlist.txt
- __init__.py
- api_utils.py
- backend_utils.py
- frontend_utils.py
- get_aws_credential.sh
- install_dependencies.py
- install_from_src.py
- install_utils
- mac_npm_deps
- mar_config.json
- markdown_link_check_config.json
- marsgen.py
- modelarchiver_utils.py
- print_env_info.py
- push_nightly.sh
- regression_utils.py
- sanity_utils.py
- setup_wsl_ubuntu
- shell_utils.py
- spellcheck.sh
- torchserve_grpc_client.py
- tsutils.py
- utils.py
- validate_model_on_gpu.py
- workflow_archiver_utils.py
- __init__.py
- manifest.py
- workflow.py
- test_handler.py
- test_spec.yaml
- __init__.py
- configuration.json
- test_integration_workflow_archiver.py
- test_version.py
- test_workflow_packaging.py
- test_workflow_packaging_utils.py
- pylintrc
- __init__.py
- arg_parser.py
- version.py
- version.txt
- workflow_archiver_error.py
- workflow_packaging.py
- workflow_packaging_utils.py
- .coveragerc
- LICENSE
- MANIFEST.in
- PyPiDescription.rst
- README.md
- setup.py
- .gitignore
- .gitmodules
- .pre-commit-config.yaml
- _config.yml
- CODE_OF_CONDUCT.md
- codecov.yml
- CONTRIBUTING.md
- LICENSE
- link_check_config.json
- MANIFEST.in
- mypy.ini
- PyPiDescription.rst
- README.md
- SECURITY.md
- setup.py
- torchserve_sanity.py
# 설치 가이드
git clone https://github.com/pytorch/serve
깃허브에서 프로젝트 코드 전체를 내 컴퓨터로 내려받습니다.
cd serve
방금 내려받은 프로젝트 폴더 안으로 이동합니다.
2. 공식 설치 스크립트
쉬움 추천- Python 3 pip 명령어를 쓰려면 Python이 필요합니다.
pip install torchserve torch-model-archiver torch-workflow-archiver
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
pip install torchserve-nightly torch-model-archiver-nightly torch-workflow-archiver-nightly
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.
3. Docker
쉬움- Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
- Docker Desktop 컨테이너를 빌드하고 실행하려면 필요합니다. 설치 후 실행해서 백그라운드에 켜두세요.
docker pull pytorch/torchserve
이 명령어를 터미널에 그대로 입력해 실행하세요.
docker pull pytorch/torchserve-nightly
이 명령어를 터미널에 그대로 입력해 실행하세요.
docker build --pull . -f docker/Dockerfile.vllm -t ts/vllm
Dockerfile을 기반으로 실행 가능한 이미지를 빌드합니다.
docker run --rm -ti --shm-size 10g --gpus all -e HUGGING_FACE_HUB_TOKEN=$token -p 8080:8080 -v data:/data ts/vllm --model_id meta-llama/Meta-Llama-3-8B-Instruct --disable_token_auth
빌드된 이미지를 실제 컨테이너로 실행합니다.
이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.
4. CMake
보통cd cpp
이 프로젝트의 관련 파일이 하위 폴더 안에 있어서, 먼저 그 폴더로 이동합니다.
mkdir build && cd build
빌드 결과물을 담을 폴더를 만들고 그 안으로 이동합니다.
cmake ..
소스코드를 분석해 빌드 설정 파일을 생성합니다 (build 폴더 안에서 실행해야 함).
make
생성된 빌드 설정을 바탕으로 실제 컴파일을 진행해 실행 파일을 만듭니다.
5. Maven (Java)
보통- Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
- JDK (Java) Java 프로젝트를 빌드/실행하려면 필요합니다.
- Maven mvn 명령어를 쓰기 위한 빌드 도구입니다.
cd serving-sdk
이 프로젝트의 관련 파일이 하위 폴더 안에 있어서, 먼저 그 폴더로 이동합니다.
mvn clean install
Maven으로 의존성 설치와 빌드를 진행합니다.
6. Gradle (Java/Kotlin)
보통- Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
- JDK (Java) Java/Kotlin 프로젝트를 빌드/실행하려면 필요합니다.
- Gradle 레포에 포함된 gradlew(Gradle Wrapper)를 쓰면 Gradle을 따로 설치할 필요가 없습니다.
cd frontend/archive
이 프로젝트의 관련 파일이 하위 폴더 안에 있어서, 먼저 그 폴더로 이동합니다.
./gradlew build
Gradle로 빌드를 진행합니다.
7. Python
쉬움pip install torchserve torch-model-archiver torch-workflow-archiver
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
pip install torchserve-nightly torch-model-archiver-nightly torch-workflow-archiver-nightly
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
python -m ts.llm_launcher --model_id meta-llama/Llama-3.2-3B-Instruct --disable_token_auth
파이썬 스크립트(또는 모듈)를 실행합니다.
python -m ts.llm_launcher --model_id meta-llama/Meta-Llama-3.1-8B-Instruct --engine trt_llm --disable_token_auth
파이썬 스크립트(또는 모듈)를 실행합니다.
이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.
