Skip to content

Pull requests: pytorch/executorch

New pull request New
436 Open 18,171 Closed
436 Open 18,171 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

portable: add fast path for contiguous innermost-dim reduction in amax.out and amin.out CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21142 opened Jul 22, 2026 by vacu9708 Contributor Loading…
2
Arm backend: Remove statically proven TOSA no-ops ciflow/trunk CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: arm Issues related to arm backend partner: arm For backend delegation, kernels, demo, etc. from the 3rd-party partner, Arm release notes: none Do not include this in the release notes
#21141 opened Jul 22, 2026 by YufengShi-dudu Collaborator Loading…
1
NXP backend: Building MCUXpresso example CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: nxp Issues related to NXP Neutron NPU delegation and code under backends/nxp/ release notes: nxp Changes to the NXP Neutron backend delegate
#21140 opened Jul 22, 2026 by jirioc Collaborator • Draft
2
Make agent instruction for attribution broader. CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21139 opened Jul 22, 2026 by robell Collaborator Loading…
3
Arm backend: Avoid unnecessary lowering pass retraces ciflow/trunk CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: arm Issues related to arm backend partner: arm For backend delegation, kernels, demo, etc. from the 3rd-party partner, Arm release notes: none Do not include this in the release notes
#21138 opened Jul 22, 2026 by YufengShi-dudu Collaborator Loading…
1
fbcode/executorch/backends/cadence/aot/compiler_utils.py CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21137 opened Jul 22, 2026 by meta-codesync Bot Loading…
3
[ExecuTorch][WebGPU] Add Qwen3 K16 attention and fp16 KV cache CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21136 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Add HuggingFace rotate-half RoPE CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21135 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Repair dynamic SDPA routing, add attestation CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21134 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Select emdawnwebgpu for Emscripten builds CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21133 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Add Llama K16 online causal attention CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21132 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Replace broad QKV fusion with BK64 kernel CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21131 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Route exact prefill shapes to the BK64 GEMM CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21130 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Fuse the SwiGLU elementwise chain CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21129 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Coordinate dynamic dispatch routes CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21128 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][WebGPU] Add certified scoped output suppression CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21127 opened Jul 22, 2026 by JCNTH Contributor Loading…
2
[ExecuTorch][llm] Wire --use_moe_quantized_op through export pipeline and LLM config CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21126 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
[ExecuTorch][llm] Add MoE quantized custom op documentation CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21125 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
[ExecuTorch][llm] Fuse w1+w3 into single GEMM in quantized_moe_ffn CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21124 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
[ExecuTorch][llm] Add numerical kernel tests for quantized_moe_ffn CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21123 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
[ExecuTorch][llm] Add source-transform and Meta-kernel tests for MoE op CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21122 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
[ExecuTorch][llm] Add MoE source transformation with INT4 quantized packing CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21121 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
[ExecuTorch][llm] Add Meta kernel and registration probe for quantized_moe_ffn CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21120 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
[ExecuTorch][llm] Add quantized_moe_ffn custom op with C++ kernel CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#21119 opened Jul 22, 2026 by digantdesai Contributor Loading…
2
Apply NoPE on global attention layers (fix ANE/static RoPE mismatch) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. meta-exported
#21115 opened Jul 22, 2026 by YIWENX14 Contributor Loading…
3
ProTip! Mix and match filters to narrow down what you’re looking for.