Painless distributed training with torch
active 2024-10-25 → 2026-04-28 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 551 days from 2024-10-25 to 2026-04-28. Pushes: 26 total, peak 6 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 22 total, peak 14 in a day. Comments: 5 total, peak 2 in a day. Stars: 3 total, peak 1 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| sehoffmann | 51 | 24 | 0 | 5 |
| vitusbenson | 2 | 2 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#46sehoffmann2025-01-25 16:55Refactor checkpoint_dir
- Issue#46sehoffmann2025-01-25 16:55Refactor checkpoint_dir
- Issue comment#47sehoffmann2025-01-25 16:52Random test failure
- Issue#52sehoffmann2025-01-25 13:44I/O operation on closed file
- Issue#54sehoffmann2025-01-25 13:24`Stage.current_epoch` is inconsistent
- Issue#53sehoffmann2025-01-25 13:15`monitored_barrier` not working with dummy initialization
- Issue#30sehoffmann2025-01-05 16:48Tensorboard support
- Issue#35sehoffmann2025-01-05 16:47Test coverage
- Issue comment#14sehoffmann2025-01-05 16:35Distributed carbon and power tracking
- Issue#34sehoffmann2025-01-05 16:33Meta data
- Issue#33sehoffmann2025-01-05 16:22Track GPU memory
- Issue#32sehoffmann2025-01-05 16:15Write all GPUs to checkpoint dir
- Issue#31sehoffmann2025-01-05 16:14Stage finishing line printed before table is closed
- Issue#30sehoffmann2025-01-05 16:11Tensorboard support
- Issue#29sehoffmann2025-01-05 15:15Control characters in logfile
- Issue#28sehoffmann2025-01-05 15:11Elasticlaunch support
- Issue comment#8sehoffmann2025-01-05 15:10Git diffs
- Issue#8sehoffmann2025-01-05 15:10Git diffs
- Issue#27sehoffmann2025-01-05 15:09time.monotonic_ns() causes segmentation faults in multiprocessing torch.Dataloader
- Issue#26sehoffmann2025-01-05 14:05Ensrue dmlcloud runs without a git repo
- Issue#25sehoffmann2025-01-05 12:29Save python and conda environment
- Issue#24sehoffmann2025-01-05 12:29Check cuda_visisble_devices for device placement
- Issue#15sehoffmann2024-12-26 18:02Save metrics as csv
- Issue#23sehoffmann2024-12-26 17:53Distributed seeding
- Issue#9sehoffmann2024-12-26 13:01Learning rate scaling
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 3 stars here means stars gained during the window, not the repo's star count.