RepoJournal
PyTorch @pytorch
117 wires and counting

$ follow PyTorch

Keep up with PyTorch in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-05-17
stories 56
GitHub RSS Archive Reviews

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

FLEX_ATTENTION TIGHTENS, VMAP GAINS VECTORIZATION, MACOS CLEARS WARNINGS

By RepoJournal · Filed · About PyTorch · Composed from the cited sources · methodology

choijon5 fulvius31 SherlockNoMad
3 people shipped this
choijon5 @choijon5 3 cited fulvius31 @fulvius31 1 cited SherlockNoMad @SherlockNoMad 1 cited

PyTorch tightened flex_attention validation and shipped the missing vmap rule for count_nonzero, while macOS builds finally silence the deprecated declarations cascade.

The flex_attention operator now explicitly rejects NestedTensor inputs [1] instead of falling through to compiler errors, fixing #177377 with a regression test on the compiled fullgraph path. In parallel, count_nonzero gained its missing batching rule [2], eliminating the performance warning and enabling vectorized execution under torch.vmap. The macOS backend suppressed the flood of -Wdeprecated-declarations from Apple framework includes [3] by wrapping Foundation, Metal, MPS, and MPSGraph headers, clearing noise from recent SDK upgrades. On the infrastructure side, inductor.yml migrated to OSDC with the dial-up pattern [4], plumbing ARC inputs through CUDA and CPU build/test pairs, while _FastCudaLauncher now silently handles oversized kernels [5] instead of throwing unexpected ValueError. Helion's autotuner hardened LLM search to fail loudly on errors [6] and wired cache backends to RemoteAutotuneCache for warm-start enrichment [7], plus added H100 sm90 pretuned heuristics [8]. The LLM search stack shipped Opus 4.6/4.7 fast mode [9] and an effort_level knob spanning none/low/medium/high/max [10] with Anthropic adaptive thinking support. TorchTitan pinned GITHUB_TOKEN to contents: read [11] in response to CVE-2025-30066, and skipped dense numerics tests [12] due to an upstream DTensor regression with mixed-dtype sharding propagation.

Action items

References

  1. [1] Reject NestedTensor inputs in flex_attention (#183516) pytorch/pytorch ↗
  2. [2] Add batching rule for count_nonzero (#183860) pytorch/pytorch ↗
  3. [3] [BE][MacOS] Suppress deprecated declarations warnings (#183927) pytorch/pytorch ↗
  4. [4] [OSDC] Migrate inductor.yml to OSDC (ARC) via dial-up pattern (#183646) pytorch/pytorch ↗
  5. [5] [inductor] Silence _FastCudaLauncher ValueError on oversized kernels (#183967) pytorch/pytorch ↗
  6. [6] [Autotuner] LLM search: fail loudly + mTLS gateway compatibility (#2448) ↗ pytorch/helion
  7. [7] [cache] Wire from_best_available / from_cache to RemoteCacheBackend ↗ pytorch/helion
  8. [8] Add H100 (sm90) pretuned heuristics and perf gates ↗ pytorch/helion
  9. [9] [Autotuner] LLM search: Anthropic Opus 4.6/4.7 fast mode ↗ pytorch/helion
  10. [10] [Autotuner] LLM search: effort_level knob + Anthropic adaptive thinking + OpenAI xhigh ↗ pytorch/helion
  11. [11] ci: declare workflow-level `contents: read` on 2 workflows (#3367) ↗ pytorch/torchtitan
  12. [12] [graph_trainer] Skip dense numerics tests due to upstream DTensor regression ↗ pytorch/torchtitan

$ ls pytorch/reviews/ # the step-back reads

the week in review torch.compile dtype fix, two Arm CVEs patched the month in review PyTorch 2.14.0 ships, TorchRL 0.14 lands
Full archive → Weekly reviews →

Quick answers

What shipped in PyTorch on May 17, 2026?
PyTorch tightened flex_attention validation and shipped the missing vmap rule for count_nonzero, while macOS builds finally silence the deprecated declarations cascade. In total, 42 commits and 14 pull requests landed.
Who contributed to PyTorch on May 17, 2026?
3 developers shipped this update, including fulvius31, choijon5, and SherlockNoMad.
What were the notable PyTorch updates?
Reject NestedTensor inputs in flex_attention (#183516), Add batching rule for count_nonzero (#183860), and [BE][MacOS] Suppress deprecated declarations warnings (#183927).
♥

$ log @pytorch --recent

Tensordict compiles masked indexing under fullgraph; ExecuTorch links WebGPU into the Wasm build 138 shipped TensorDict 0.15.0 goes pure Python, 1.77x faster 183 shipped Inductor hands Triton kernels to the worker pool, RL rollout KV cache persists across turns 219 shipped ExecuTorch drops Python 3.10, TensorRT trims compile startup, tensordict fixes silent h5 loss 164 shipped

$ log --all # elsewhere on the wire

Alignment & Post-training verl rejects conflicting vLLM engine_kwargs, open-instruct adds MILES OPD behind a Megatron learner Interpretability nnsight 0.8.0 cuts import time from 7.6s to 1.6s

$ ls showcases/

Safety Evals AI Red Teaming Agent Security Interpretability Alignment & Post-training

$ init your-desk

RepoJournal can write one of these for any repo you care about. Sign in with GitHub, pick your repos and sources, and your desk gets a wire like this one every morning.