Problemsby IbisTrouble with PyTorch worker
MLOps3 posts from 2 people2d active+22 posts in the last 7 days, 0 the 7 days before (rising)
Posts per day
The posts behind it
3, newest firsttorchrun exits 0 when a worker group in D-state runs out of restarts### 🐛 Describe the bug When the uninterruptible-sleep (D-state) detection marks the worker group UNHEALTHY and no restarts are left, the agent stops the workers and then reports success: `elastic_launch` returns an empty dict instead of raising `ChildFailedError`, and torchrun...pytorch/pytorch · digit50Open ↗TORCHELASTIC_ENABLE_FILE_TIMER: one expired worker timer SIGKILLs torchrun### 🐛 Describe the bug With `TORCHELASTIC_ENABLE_FILE_TIMER=1`, when a worker's `expires()` timer runs out, the `FileTimerServer` kills the worker and then sends the same signal (SIGKILL by default) to its own process. That server runs as a thread inside the torchrun agent, so...pytorch/pytorch · digit50Open ↗[docs] Worker.__init__ docstring documents a nonexistent `id` argument### 🐛 Describe the bug In `torch/distributed/elastic/agent/server/api.py`, the `Worker` class docstring lists `id (Any)` under its `Args:` block, but `Worker.__init__` does not accept an `id` argument. It only takes the rank and world-size ints, and `self.id` is set internally...pytorch/pytorch · alwaysprince05Open ↗
Companies and products named
About this problem
Evidence
3 posts from 2 people in 1 place, about 1 a week over 18 days. Mostly on pytorch/pytorch. Tools named alongside: GitHub.
Frustration Frustration 0 of 3· Seen on GitHub issues
How it was grouped
Posts about PyTorch with no known issue, sharing the word "worker". First post Sep 30, 2026, latest Oct 17, 2026. Corroborated: 3 or more posts from 2 or more people or places. Method
Other PyTorch problems
PyTorch users keep filing bugsTrouble with PyTorch inputTrouble with PyTorch truePyTorch deprecations force unplanned migrationsPyTorch upgrades break working setupsPyTorch builds are slow or flakyTrouble with PyTorch inductorPyTorch GPU access and drivers are painfulTrouble with PyTorch testPyTorch vulnerabilities are hard to triage and patchRunning PyTorch locally hits memory and speed limitsPyTorch error messages are cryptic
History
Trouble with PyTorch worker (3 posts)