mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-07-19 19:00:47 +00:00
561aca97cf
<del> **This PR is based on a few prerequisites PRs. They are listed as below:** - #17465 - #17469 - #17470 - #17472 - #17473 - #17484 Please review the current change by only looking at commit e2e6623e673ec6de55a5c1f8edcbd3a46b535a89 and later. </del> ### Description This PR introduces WebGPU IO binding. This new feature allows onnxruntime-web users to use tensors created from GPU as model input/output so that a model inferencing can be done without unnecessary data copy between CPU and GPU for model input/output. ### Examples An E2E demo/example is being worked on. Following is some simple demo with code snippet. Let's first check today how we do: ```js // STEP.1 - create an inference session: const mySession = await ort.InferenceSession.create('./my_model.onnx', { executionProviders: ['webgpu'] }); // STEP.2 - create model input: (supposing myImageCpuData is a Float32Array) const feeds = { 'input_image:0': new ort.Tensor('float32', myImageCpuData, [1, 224, 224, 3]) }; // STEP.3 - run model const myResults = await mySession.run(feeds); // STEP.4 - get output data const myData = myResults['output_image:0'].data; // Float32Array ``` #### for inputs (GPU tensor): Now, with IO binding, you can create a tensor from a GPU buffer, and feed it to the model: ```js // new STEP.2.A - create model input from a GPU buffer: (supposing myInputGpuBuffer is a `GPUBuffer` object with input data) const feeds = { 'input_image:0': ort.Tensor.fromGpuBuffer(myInputGpuBuffer, { dataType: 'float32', dims: [1, 224, 224, 3] }) }; ``` ### for outputs (pre-allocated GPU tensor) you can also do that for output, **if you know the output shape**: ```js // new STEP.2.B - create model output from a GPU buffer: (supposing myOutputGpuBuffer is a pre-allocated `GPUBuffer` object) const fetches = { 'output_image:0': ort.Tensor.fromGpuBuffer(myOutputGpuBuffer, { dataType: 'float32', dims: [1, 512, 512, 3] }) }; // new STEP.3 - run model with pre-allocated output (fetches) const myResults = await mySession.run(feeds, fetches); ``` ### for outputs (specify location) if you do not know the output shape, you can specify the output location when creating the session: ```js // new STEP.1 - create an inference session with an option "preferredOutputLocation": const mySession = await ort.InferenceSession.create('./my_model.onnx', { executionProviders: ['webgpu'], preferredOutputLocation: "gpu-buffer" }); ``` if the model has multiple outputs, you can specify them seperately: ```js // new STEP.1 - create an inference session with an option "preferredOutputLocation": const mySession = await ort.InferenceSession.create('./my_model.onnx', { executionProviders: ['webgpu'], preferredOutputLocation: { "output_image:0": "gpu-buffer" } }); ``` now you don't need to prepare the `fetches` object and onnxruntime-web will prepare output data on the location that specified. #### read data when you get the output tensor, you can: ```js // get the gpu buffer object: const gpuBuffer = myOutputTensor.gpuBuffer; // GPUBuffer // get the CPU data asynchronizely const cpuData = await myOutputTensor.getData(); // get the CPU data asynchronizely and release the underlying GPU resources const cpuData = await myOutputTensor.getData(true); // dispose the tensor (release the underlying GPU resources). This tensor object will be invalid after dispose() is called. myOutputTensor.dispose(); ``` #### resource management JavaScript has GC so you don't need to worry about managing JavaScript objects. But there are 2 types of resources that are not managed by GC: - GPU buffer that used in tensors - Underlying ORT native resources To simplify, most of the unmanaged resources and handled inside ORT web. But there are a few resources that need users to manage: - All external GPU resources, including GPU buffers inside all tensors created by `Tensor.fromGpuBuffer()`, will not be managed by ORT. User should manage those GPU buffers themselves. - When a session is created with `preferredOutputLocation` == "gpu-buffer" specified in session options, and the corresponding output is not pre-allocated, user need to call the output tensor's `dispose()` or `getData(true)` to manually release the underlying GPU buffers. - ORT internal errors (including providing a pre-allocated output tensor with wrong type/dims) will invalidate the whole wasm memory and is not recoverable. An exception is thrown in this situation. |
||
|---|---|---|
| .. | ||
| jobs | ||
| stages | ||
| android-binary-size-check-stage.yml | ||
| android-dump-logs-from-steps.yml | ||
| android-java-api-aar-test.yml | ||
| android-java-api-aar.yml | ||
| build-linux-wasm-step.yml | ||
| c-api-artifacts-package-and-publish-steps-posix.yml | ||
| c-api-artifacts-package-and-publish-steps-windows.yml | ||
| c-api-cpu.yml | ||
| c-api-linux-cpu.yml | ||
| check-cache-stats.yml | ||
| clean-agent-build-directory-step.yml | ||
| common-variables.yml | ||
| compliance.yml | ||
| component-governance-component-detection-steps.yml | ||
| download-deps.yml | ||
| esrp_nuget.yml | ||
| explicitly-defined-final-tasks.yml | ||
| flex-downloadPipelineArtifact.yml | ||
| get-docker-image-steps.yml | ||
| install-appcenter.yml | ||
| java-api-artifacts-package-and-publish-steps-posix.yml | ||
| linux-build-step-with-cache.yml | ||
| linux-ci.yml | ||
| linux-cpu-packaging-pipeline.yml | ||
| linux-gpu-tensorrt-packaging-pipeline.yml | ||
| linux-wasm-ci.yml | ||
| linux-web-init-and-check.yml | ||
| mac-build-step-with-cache.yml | ||
| mac-cpu-packaging-pipeline.yml | ||
| mac-cpu-packaging-steps.yml | ||
| mac-cpu-packing-jobs.yml | ||
| mac-esrp-dylib.yml | ||
| nodejs-artifacts-package-and-publish-steps-posix.yml | ||
| nodejs-artifacts-package-and-publish-steps-windows.yml | ||
| ondevice-training-cpu-packaging-pipeline.yml | ||
| orttraining-linux-gpu-test-ci-pipeline.yml | ||
| publish-nuget.yml | ||
| py-linux-gpu.yml | ||
| py-linux.yml | ||
| py-package-smoking-test.yml | ||
| py-packaging-linux-test-cpu.yml | ||
| py-packaging-linux-test-cuda.yml | ||
| py-packaging-selectable-stage.yml | ||
| py-packaging-stage.yml | ||
| py-packaging-training-cuda-stage.yml | ||
| py-win-gpu.yml | ||
| react-native-ci.yml | ||
| rocm.yml | ||
| run-docker-build-steps.yml | ||
| set-nightly-build-option-variable-step.yml | ||
| set-python-manylinux-variables-step.yml | ||
| set-version-number-variables-step.yml | ||
| telemetry-steps.yml | ||
| upload-code-coverage-data.yml | ||
| use-android-ndk.yml | ||
| use-xcode-version.yml | ||
| validate-package.yml | ||
| web-browserstack-ci.yml | ||
| web-ci.yml | ||
| win-ci.yml | ||
| win-esrp-dll.yml | ||
| win-wasm-ci.yml | ||
| win-web-ci.yml | ||
| win-web-multi-browsers.yml | ||
| with-container-registry-steps.yml | ||