pytorch/executorch

View on GitHub

On-device AI across mobile, embedded and edge for PyTorch

Stars: 5k
Forks: 1.2k
Open beginner issues: 0
Indexed issues: 851
Avg merge: 2d 10h
Merged PRs (30d): 581

Dominant language: Python
License: No license data
Last GitHub push: Sep 19, 2026
Latest indexed: Sep 19, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: good first issue help wanted

Beginner friendly Everything

Open All

851 open issues indexed

[ET-VK] Segmentation Faults when running Vulkan Backend on certain Qualcomm chipsets Open

module: vulkan

pytorch/executorch#17557 · 1 assignee · Opened Feb 19, 2026
ExecuTorch CoreML Delegate Segfault for `aten::where` (1-input) and `aten::nonzero_numpy` Open

module: coreml

pytorch/executorch#17537 · 3 comments · 1 assignee · Opened Feb 18, 2026
Support quantize_ for convolution on XNNPACK Open

module: xnnpack

pytorch/executorch#17432 · 1 reaction · 1 assignee · Opened Feb 13, 2026
Update demo video on adding packages for Using ExecuTorch on iOS Open

advanced docathon-2026 module: doc module: ios

pytorch/executorch#17425 · 11 comments · 1 assignee · Opened Feb 12, 2026
Fix CPU performance on Whisper (index_put) Open

pytorch/executorch#17420 · 1 assignee · Opened Feb 12, 2026
Analyze XNNPACK linear + conv perf across thread count + hw Open

pytorch/executorch#17345 · 1 assignee · Opened Feb 10, 2026
DirectML support on Windows Open

module: windows

pytorch/executorch#17298 · 1 comment · 1 assignee · Opened Feb 8, 2026
LoweringException: NoValidChoicesError: No choices to select. Open

module: cuda

pytorch/executorch#17197 · 10 comments · 1 reaction · 1 assignee · Opened Feb 4, 2026
gemma3 int4 tile packed delay after slimtner Open

pytorch/executorch#17162 · 1 comment · 1 assignee · Opened Feb 3, 2026
add device attribute in ETensor Open

pytorch/executorch#17161 · 7 comments · 1 assignee · Opened Feb 3, 2026
MCU hardware benchmark compared to TFLite Open

module: microcontrollers partner: arm

pytorch/executorch#17157 · 32 comments · 2 reactions · 2 assignees · Opened Feb 3, 2026
Point OpenVINO Backend Requierments to a Stable PyPi Version Open

partner: intel

pytorch/executorch#17148 · 1 assignee · Opened Feb 3, 2026
ddr memory requrement in long-context (8k or more) Open

module: qnn partner: qualcomm

pytorch/executorch#17146 · 1 comment · 1 assignee · Opened Feb 3, 2026
Long-term fix is to use a tiny LLM-like model designed for CI that fits on Ethos-U with realistic weights. Open

Difficulty 5/5 Over a week Newbie friendliness 15/100

pytorch/executorch#17133 · Opened Feb 3, 2026
Gemma: LLM - Spanish response to the English prompt. Open

module: llm

pytorch/executorch#17120 · 1 assignee · Opened Feb 2, 2026
Add native support for Vision-Language-Action (VLA) models on edge devices Open

module: examples

pytorch/executorch#17079 · 8 comments · 1 assignee · Opened Jan 31, 2026
autotune custom kernels in cuda backend Open

pytorch/executorch#17008 · 2 comments · 1 assignee · Opened Jan 29, 2026
flatc is not found but is present Open

module: exir module: xnnpack triaged

Difficulty 3/5 1-2 days Newbie friendliness 65/100

pytorch/executorch#17000 · 8 comments · Opened Jan 29, 2026
Are there plans to add multi-LoRA support to the Qualcomm backend Open

module: qnn partner: qualcomm

pytorch/executorch#16999 · 5 comments · 3 reactions · 1 assignee · Opened Jan 29, 2026
Investigate failure when running metal delegated whisper sampler (argmax) Open

module: metal triaged

pytorch/executorch#16984 · 1 reaction · 2 assignees · Opened Jan 28, 2026
Bug Report: LLaMA Model Generating Gibberish Output (tokenXXXX) on Android with QNN Backend Open

module: qnn partner: qualcomm triaged

Difficulty 5/5 Over a week Newbie friendliness 25/100

pytorch/executorch#16972 · 8 comments · Opened Jan 28, 2026
The Cadence runtime example needs update Open

module: cadence triaged

pytorch/executorch#16898 · 6 comments · 1 assignee · Opened Jan 27, 2026
Add Test cases & Delete old weights here if they have no users left? Open

pytorch/executorch#16866 · 1 assignee · Opened Jan 26, 2026
Safely deserialize inputs/buffers/tensors from the PTE file. Open

pytorch/executorch#16810 · 1 assignee · Opened Jan 23, 2026
Explore Nvidia Nano Jetson Open

pytorch/executorch#16786 · 1 comment · 1 assignee · Opened Jan 22, 2026
want to set npu cores and select npu cores used Open

module: qnn partner: qualcomm triaged

Difficulty 5/5 Over a week Newbie friendliness 28/100

pytorch/executorch#16762 · 14 comments · Opened Jan 22, 2026
Have example of GatedDeltaNet exporting Open

pytorch/executorch#16708 · 1 assignee · Opened Jan 20, 2026
Create VCPKG port Open

triaged

pytorch/executorch#16685 · 3 comments · 1 reaction · 2 assignees · Opened Jan 17, 2026
Enable fp16 GeLU on XNNPACK post-update Open

pytorch/executorch#16679 · 1 assignee · Opened Jan 16, 2026
Backend init configs Open

pytorch/executorch#16569 · 1 assignee · Opened Jan 13, 2026
CoreML sometimes produces garbage output on cached models Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

pytorch/executorch#16492 · 1 comment · Opened Jan 7, 2026
Analyze CPU operator performance Open

pytorch/executorch#16471 · 1 assignee · Opened Jan 6, 2026
Implement torchvision NMS + BoxWithNMSLimit custom op Open

pytorch/executorch#16470 · 2 comments · 1 assignee · Opened Jan 6, 2026
Accuracy drop : DeiT-tiny & LLaMA behavior differences between FP / Decomposition / FQ / TOSA Open

module: arm partner: arm

Difficulty 4/5 3-5 days Newbie friendliness 45/100

pytorch/executorch#16426 · 6 comments · Opened Jan 2, 2026
java linux cannot work , we need executorch java jar format package ，please support Open

module: android

Difficulty 5/5 Over a week Newbie friendliness 20/100

pytorch/executorch#16422 · 4 comments · Opened Dec 31, 2025
Batch Inference On 8255 device Open

module: qnn partner: qualcomm

Difficulty 5/5 Over a week Newbie friendliness 25/100

pytorch/executorch#16413 · 12 comments · Opened Dec 30, 2025
Errors while trying to create a .pte file for gemma 3 Open

module: cuda module: quantization triaged

Difficulty 4/5 3-5 days Newbie friendliness 42/100

pytorch/executorch#16411 · 19 comments · Opened Dec 30, 2025
examples/qualcomm/qaihub_scripts/stable_diffusion runtime error Open

module: qnn partner: qualcomm

Difficulty 4/5 3-5 days Newbie friendliness 30/100

pytorch/executorch#16407 · 11 comments · Opened Dec 29, 2025
WearOS executorch `dlopen failed: library "libexecutorch.so" not found` Open

module: android

pytorch/executorch#16405 · 5 comments · 1 assignee · Opened Dec 27, 2025
XNNPACK is not taking effect on Android and x86 platforms. Open

module: xnnpack

Difficulty 4/5 3-5 days Newbie friendliness 35/100

pytorch/executorch#16397 · 4 comments · Opened Dec 26, 2025
Numeric Discrepancy detector for ExecuTorch Qualcomm backend. Open

pytorch/executorch#16381 · 3 comments · 1 reaction · 2 assignees · Opened Dec 24, 2025
Android ExecuTorch :Improve ET Dump (Prod / Release ready) with Runtime control Open

pytorch/executorch#16360 · 1 comment · 1 assignee · Opened Dec 22, 2025
Benchmark CMSIS Conv & DW Conv and improve operator selection Open

Difficulty 4/5 3-5 days Newbie friendliness 42/100

pytorch/executorch#16347 · Opened Dec 19, 2025
Support message matching in death tests Open

pytorch/executorch#16281 · 1 comment · 1 assignee · Opened Dec 16, 2025
custom op auto tuning on triton kernel Open

pytorch/executorch#16279 · 1 assignee · Opened Dec 16, 2025
Efficient SAM3 demo Open

module: examples

pytorch/executorch#16269 · 13 comments · 1 assignee · Opened Dec 15, 2025
Android Documentation - Improve Llama example Open

android_ux

pytorch/executorch#16263 · 1 assignee · Opened Dec 15, 2025
Android Demo: Enable QWEN for QNN Open

android_ux

pytorch/executorch#16259 · 1 assignee · Opened Dec 15, 2025
Android CI : Demo App, Incrementally test the app with CI Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

pytorch/executorch#16258 · Opened Dec 15, 2025
Device mismatch error when evaluating converted PT2E quantized ViT model on CUDA/MPS Open

module: mps partner: arm

pytorch/executorch#16250 · 8 comments · 2 assignees · Opened Dec 15, 2025

pytorch/executorch

Get new issues in your inbox