Who contributed to PyTorch on July 13, 2026?

6 developers shipped this update, including qflen, Jason Ansel, PyTorch UpdateBot, Aaron Orenstein, Yang, Shuxin, and jewelkm89.

What were the notable PyTorch updates?

Add an accurate Metal erfc for MPS (eager + inductor) (#189291), Remove stale Dynamo skip for max_pool1d pooling test (#186531), and Handle deleted push dispatches (ciflow tag cleanup).

@pytorch

PyTorch and the broader machine-learning ecosystem

github ↗

Pick a date

Topics: Python AI / ML Full archive →

The Wire · Showcase

PYTORCH FIXES CRITICAL MPS MATH BUG WHILE TAMING CI GHOSTS

By RepoJournal · Filed 06:58 UTC on July 13, 2026 · About PyTorch

6 people shipped this

qflen @qflen 1 cited

Jason Ansel @jansel 1 cited

Aaron Orenstein @aorenste 1 cited

Yang, Shuxin @orrangetabby17 1 cited

jewelkm89 @jewelkm89 1 cited

PyTorch UpdateBot @pytorchupdatebot

PyTorch's Metal backend shipped an accurate erfc function that eliminates 100% relative error on upper tail computations, fixing a cascade of broken numerics in GELU and special functions.

The Metal Performance Shaders backend had a silent catastrophe: `c10::metal::erfc` was computing `1.0 - erf(x)`, which saturates to zero once erf hits fp32 limits around x=3.9, poisoning everything downstream [1]. That means GELU activations, gamma function completements, and `torch.special.erfc` all inherited the flush. This fix splits out a dedicated Metal implementation that holds accuracy across the entire range. In parallel, PyTorch axed a stale Dynamo skip decorator on max_pool1d tests [2] now that the CPU pooling suite no longer triggers LLVM OOM failures, restoring regression coverage that had been dormant. On the CI side, the team hardened crcr-test to handle deleted push dispatches without false L1 probe failures [3]. GitHub sends a null SHA when branches or tags get deleted during PR cleanup; the old code tried checking out ref `0000...`, which crashed. The fix adds a dedicated `push-deleted` job that validates deletion semantics and skips checkout entirely. Routine inductor work continues: fbcode xfails for numerical property mismatches [4] and XPU maxpool backward fallback tuning [5].

Action items

→ Test MPS GELU and special function numerics on your M-series hardware after this lands pytorch/pytorch [plan]
→ Monitor L1 CI for clean push-deleted handling across next 5 tag cleanup cycles pytorch/crcr-test [monitor]

References

[1] Add an accurate Metal erfc for MPS (eager + inductor) (#189291) pytorch/pytorch
[2] Remove stale Dynamo skip for max_pool1d pooling test (#186531) pytorch/pytorch
[3] Handle deleted push dispatches (ciflow tag cleanup) ↗ pytorch/crcr-test
[4] [inductor] IS_FBCODE xfails for opinfo numeric-property tests (#189640) pytorch/pytorch
[5] [xpu][inductor] fix the fallback behavior of maxpool backward for performance (#187940) pytorch/pytorch

Quick answers

What shipped in PyTorch on July 13, 2026?: PyTorch's Metal backend shipped an accurate erfc function that eliminates 100% relative error on upper tail computations, fixing a cascade of broken numerics in GELU and special functions. In total, 13 commits and 1 pull requests landed.
Who contributed to PyTorch on July 13, 2026?: 6 developers shipped this update, including qflen, Jason Ansel, PyTorch UpdateBot, Aaron Orenstein, Yang, Shuxin, and jewelkm89.
What were the notable PyTorch updates?: Add an accurate Metal erfc for MPS (eager + inductor) (#189291), Remove stale Dynamo skip for max_pool1d pooling test (#186531), and Handle deleted push dispatches (ciflow tag cleanup).

PYTORCH FIXES H100 FLASH ATTENTION CI BREAKAGE, HELION CONSOLIDATES PALLAS LAUNCHERS

PyTorch patched a critical H100 test failure caused by upstream flash-attention directory churn, while Helion merged a major Pallas launcher unification that simplifies the runtime API.

python 20 shipped 1-min read

@pytorch 2 days ago

DYNAMO FIXES ITERATOR CORRUPTION AND FAKE INT ARITHMETIC

Dynamo's deque iteration was silently failing to detect mutations and reversing into the wrong iterator type, while custom hash functions using id() arithmetic crashed the compiler.

python 56 shipped 1-min read

@pytorch 3 days ago

TORCHRL CHECKPOINTS NOW MEMORY-MAPPED BY DEFAULT, FIXING WINDOWS FILE-LOCK CORRUPTION

TorchRL's trainer checkpoint loader was materializing entire files into non-evictable RAM and aliasing external state into live components, causing permission errors on Windows and potential corruption if checkpoints were rewritten externally.

python 104 shipped 2-min read

@pytorch 4 days ago

PYTORCH 2.13 SHIPS WITH FLEXATTENTION ON APPLE SILICON, NCCL2 BACKEND LANDS IN-TREE

PyTorch's biggest release in months brings native Apple Silicon acceleration, refactors distributed training infrastructure, and ships quantized embedding support across CPU and accelerators.

python 112 shipped 1-min read

— or follow the whole beat:

Elsewhere on the wire

Elixir & Phoenix about 6 hours ago

ELIXIR FIXES SPEC AND ARITY BUGS ACROSS CORE MODULES

Elixir core shipped three precision fixes overnight that tighten specs on IEx and EEx utilities, catching the kind of type mismatches that slip past dialyzer.

elixir 7 shipped 1-min read

Mobile platforms about 6 hours ago

FLUTTER CLEANS HOUSE ON ANALYSIS RULES WHILE SWIFT FIXES C-INTEROP CRASHES

Flutter is migrating away from deprecated strict analysis modes in favor of lint rules, while Swift patches critical crashes in bit-field accessors and C interoperability.

mobile 21 shipped 1-min read

@dotnet about 6 hours ago

ASPNETCORE VALIDATION GENERATOR FIXED, MSBUILD SYNCS ROSLYN AND NUGET

ASP.NET Core's validation source generator now correctly unwraps array types, closing a gap that could slip validation past your models.

dotnet 6 shipped 1-min read

@spring-projects about 6 hours ago

SPRING-AI FIXES CRITICAL TOOL CALLING BUG THAT BROKE OPENAI STRICT MODE

The `strict` parameter in OpenAI tool calling was silently failing across all Spring AI implementations because it was serialized in the wrong place in the payload.

java 8 shipped 1-min read