executorch icon indicating copy to clipboard operation
executorch copied to clipboard

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

Results 597 executorch issues
Sort by recently updated
recently updated
newest added

Summary: This diff is to rename the "sdk-etrecord" documentation page to "etrecord". Old URL: https://pytorch.org/executorch/main/sdk-bundled-io.html New URL ("sdk" is removed): https://pytorch.org/executorch/main/bundled-io.html Design doc: https://docs.google.com/document/d/1l6DYTq9Kq6VrPohruRFP-qScZDj01W_g4zlKyvqKGF4/edit?usp=sharing Differential Revision: D63651681

CLA Signed

This PR adds a job to upload Android benchmark results to the benchmark database. It transforms the `benchmark_results.json` file slightly to fit into the current schema. We are going to...

CLA Signed

### 🐛 Describe the bug Get https://github.com/pytorch/executorch/pull/5710 and run ``` python executorch.examples.apple.coreml.scripts.export -m resnet18 --quantize ``` The FP32 model runs fully resident on ANE at 0.9ms on average and 11.13ms...

triaged
module: coreml

Differential Revision: D63863399

CLA Signed
fb-exported

Summary: Add short term pybindings and training module api. The optimizer bindings are probably fairly stable in practice, but the training module is not. Training module is a wrapper around...

CLA Signed
fb-exported

### 🐛 Describe the bug We are running exercise on the [newly launched benchmarking infra](https://github.com/pytorch/executorch/tree/main/extension/benchmark) with the in-tree enabled models under `examples/models`. You can follow the instruction to extract and...

partner: apple
module: coreml

Integer placeholders, for example in this examples x = torch.Tensor([1.]) + 1 Are by default lowered as int64 tensors. As int64 is not a valid TOSA dtype, we cast these...

CLA Signed
partner: arm
ciflow/trunk

### 🐛 Describe the bug setup-with-qnn.sh ### Versions examples/demo-apps/android/LlamaDemo/setup-with-qnn.sh: 68: pushd: not found while running script getting specific error?

Summary: X-link: https://github.com/pytorch/pytorch/pull/137289 User can preserve custom ops as well. We do so by extending current logic to also query CIA ops from all possible op namespaces. Differential Revision: D63846832...

CLA Signed
fb-exported