Who contributed to PyTorch on July 26, 2026?

5 developers shipped this update, including Guan-Ming (Wesley) Chiu, Denis Boyda, Jim Meyering, drisspg, and Lisa Jin.

What were the notable PyTorch updates?

[Dynamo] Support SDPAParams construction in fullgraph (#190839), [c10d][torchcomms] Add MCCL backend wrapper example (#190292) (#191034), and [Dynamo] Reland: Use getattr_static when probing internal attributes (#190970).

@pytorch

PyTorch and the broader machine-learning ecosystem

github ↗

Pick a date

Topics: Python AI / ML Full archive →

The Wire · Showcase

DYNAMO FIXES GRAPH-BREAKING SDPA PARAMS, PAT SVD LEAK BREAKS DOWNSTREAM TESTS

By RepoJournal · Filed 06:58 UTC on July 26, 2026 · About PyTorch

5 people shipped this

Guan-Ming (Wesley) Chiu @guan404ming 2 cited

Denis Boyda @dboyda 1 cited

Jim Meyering @meyering 1 cited

drisspg @drisspg 1 cited

Lisa Jin @lisjin 1 cited

PyTorch shipped fixes for two critical compilation and module patching bugs overnight that were silently breaking downstream workflows.

Dynamo can now construct SDPAParams inside compiled regions without graph-breaking under fullgraph=True [1], routing the pybind11 class through TorchInGraphFunctionVariable following the torch.nn.Parameter pattern. The same Dynamo reland [3] restores getattr_static for internal attribute probing after it was reverted for inflating frame counts, but the fix is back in and the frame-count assertions pass. On the CUDA front, _scaled_mm with float8 output was silently dropping the scale_result parameter [5], causing identical bitwise outputs for different scales. Over in pytorch/ao, insert_svd_modules_ was mutating the shared nn.Linear class directly, leaving AttributeError traps in weight access across unrelated modules [6], breaking torch.compile tests in int4/int8/float8 workflows after any pat SVD test ran. The fix patches a private per-instance subclass instead. Meanwhile, c10d's TorchComms integration no longer requires users to mutate c10d internals before calling init_process_group [2], and out-of-range float conversions in tensor creation tests are now restricted to valid ranges [4].

Action items

→ If running torch.compile with int4/int8/float8 quantization, upgrade pytorch/ao immediately pytorch/ao [immediate]
→ Rebuild with latest Dynamo if you rely on SDPAParams in compiled graphs pytorch/pytorch [plan]
→ Verify float8 _scaled_mm scaling behavior in your production models pytorch/pytorch [monitor]

References

[1] [Dynamo] Support SDPAParams construction in fullgraph (#190839) pytorch/pytorch
[2] [c10d][torchcomms] Add MCCL backend wrapper example (#190292) (#191034) pytorch/pytorch
[3] [Dynamo] Reland: Use getattr_static when probing internal attributes (#190970) pytorch/pytorch
[4] [caffe2] Avoid out-of-range float->integral conversions (UB) in test_tensor_creation_ops (#191025) (#191025) pytorch/pytorch
[5] [CUDA] Restore scale_result for float8 _scaled_mm output (#190979) pytorch/pytorch
[6] [pat] Fix SVD class leakage into `nn.Linear` from #4587 (#4613) pytorch/ao

Quick answers

What shipped in PyTorch on July 26, 2026?: PyTorch shipped fixes for two critical compilation and module patching bugs overnight that were silently breaking downstream workflows. In total, 17 commits and 2 pull requests landed.
Who contributed to PyTorch on July 26, 2026?: 5 developers shipped this update, including Guan-Ming (Wesley) Chiu, Denis Boyda, Jim Meyering, drisspg, and Lisa Jin.
What were the notable PyTorch updates?: [Dynamo] Support SDPAParams construction in fullgraph (#190839), [c10d][torchcomms] Add MCCL backend wrapper example (#190292) (#191034), and [Dynamo] Reland: Use getattr_static when probing internal attributes (#190970).

CRITICAL OIDC INJECTION IN DOCS PREVIEW WORKFLOW PATCHED

PyTorch's docs-preview CI trusted fork-controlled artifacts in a context with token-write permissions, exposing the entire build pipeline to code injection.

python 66 shipped 1-min read

@pytorch 2 days ago

PYTORCH AUTOGRAD GETS 7% FASTER, AOTI FIXES SILENT FAILURES

Interned attribute names in autograd.Function shaved microseconds off the hot path while AOTI's scatter operations now properly report errors instead of silently corrupting results.

python 64 shipped 1-min read

@pytorch 5 days ago

DYNAMO REVERTS BREAKING CHANGE, EXECUTORCH CLEANS UP DEPRECATED TYPES

PyTorch reverted a Dynamo optimization that broke internal tests, while ExecutorTorch is aggressively deprecating c10 shims in favor of standard library types.

python 91 shipped 1-min read

@pytorch 6 days ago

PYTORCH SHIPS BUILD FIX WHILE HELION TUNES H100 KERNELS TO DEFAULT

A critical build regression in cusparselt.cpp is now patched, while the kernel autotuner promotes its pointwise seed heuristic to production defaults on H100 and B200.

python 36 shipped 1-min read

Elsewhere on the wire

AI Agents about 5 hours ago

OPENHANDS CUTS EVENT NOISE, CLINE SCRUBS DEAD CODE ACROSS TWO DESKS

OpenHands shipped a filter to stop forwarding unrequested automation events, while Cline burned down five legacy UI bugs in one sprint.

ai-agents 14 shipped 1-min read

Local LLMs about 5 hours ago

LLAMA.CPP LOCKS DOWN WASM, ADDS MCP STDIO, HEXAGON GETS IM2COL ACCELERATION

llama.cpp disabled OpenMP for Emscripten builds to fix WASM compilation, while shipping full MCP stdio support and unlocking 1.5x to 3.5x latency gains on Hexagon vision models.

llms 139 shipped 2-min read

@CachyOS about 5 hours ago

LINUX 7.1.5 AND 6.18.40 SHIP WITH UPSTREAM REBASES AND DRIVER FIXES

CachyOS dropped two kernel releases overnight, pulling in upstream stability patches and refining the apple-bce driver stack for T2 Macs.

infra 61 shipped 1-min read

Elixir & Phoenix about 5 hours ago

ELIXIR STRIPS REDUNDANT GUARDS, FIXES TUPLE PROJECTION BUG

Barna Kovacs landed three optimization PRs removing dead code from core Enum and Calendar modules, while a separate fix addresses a flawed tuple projection algorithm that was producing inconsistent type representations.