Personal fork for understanding DeepSpeed's megatron-trained model training.
active 2025-02-04 → 2026-05-28 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 479 days from 2025-02-04 to 2026-05-28. Pushes: 7 total, peak 3 in a day. Pull requests: 23 total, peak 5 in a day. Issues: 8 total, peak 1 in a day. Comments: 20 total, peak 3 in a day. Stars: 204 total, peak 7 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 |
|---|---|---|---|---|
| tjruwase | 11 | 3 | 3 | 2 |
| YJHMITWEB | 7 | 0 | 5 | 2 |
| loadams | 6 | 3 | 3 | 0 |
| 9LLPPLL6 | 5 | 0 | 0 | 3 |
| mngom2 | 4 | 0 | 4 | 0 |
| delock | 4 | 0 | 0 | 3 |
| ys950902 | 4 | 0 | 2 | 2 |
| Schwidola0607 | 3 | 0 | 3 | 0 |
| sfc-gh-truwase | 2 | 0 | 0 | 1 |
| SeekPoint | 2 | 0 | 0 | 1 |
| anasegor | 2 | 0 | 0 | 0 |
| GuanhuaWang | 2 | 0 | 0 | 1 |
| auggie246 | 1 | 0 | 0 | 0 |
| QuPengfei | 1 | 0 | 1 | 0 |
| zuoyanzhang | 1 | 0 | 0 | 1 |
| HenryTangIntel | 1 | 0 | 0 | 0 |
| lzyonetwothree | 1 | 0 | 0 | 1 |
| Jiminator | 1 | 0 | 0 | 1 |
| jomayeri | 1 | 1 | 0 | 0 |
| KeitaW | 1 | 0 | 1 | 0 |
Recent activity
Latest issues, pull requests and releases
- Pull request#485QuPengfei2026-01-22 01:56
- Pull request#484dalongbao2025-11-04 17:12
- Issue comment#456zuoyanzhang2025-10-03 05:21[BUG] AttributeError: 'list' object has no attribute 'to'
- Pull request#483KeitaW2025-09-27 23:59
- Issue comment#443lzyonetwothree2025-09-23 06:54llama3 and llama3.1 support
- Pull request#477tjruwase2025-08-14 15:04
- Issue comment#469YJHMITWEB2025-08-09 03:22[BUG]Fix the error issue for q/k/v stride is not match for non FPDT scenarios.
- Pull request#479YJHMITWEB2025-08-09 03:19
- Pull request#480YJHMITWEB2025-08-09 03:19
- Pull request#480YJHMITWEB2025-08-09 03:13
- Pull request#479YJHMITWEB2025-08-09 03:04
- Pull request#479YJHMITWEB2025-08-09 03:01
- Issue comment#469YJHMITWEB2025-08-06 20:52[BUG]Fix the error issue for q/k/v stride is not match for non FPDT scenarios.
- Issue comment#469sfc-gh-truwase2025-08-04 23:50[BUG]Fix the error issue for q/k/v stride is not match for non FPDT scenarios.
- Issue comment#469rraminen2025-08-04 23:44[BUG]Fix the error issue for q/k/v stride is not match for non FPDT scenarios.
- Pull request#468tjruwase2025-07-16 11:06
- Issue comment#468delock2025-07-16 01:39[Bug]Add sequence_parallel in layernorm init to enable 3D parallelism with DeepSpeed for non CUDA device.
- Pull request#478mngom22025-07-11 23:07
- Pull request#478mngom22025-07-11 23:07
- Issue comment#468delock2025-06-30 10:58[Bug]Add sequence_parallel in layernorm init to enable 3D parallelism with DeepSpeed for non CUDA device.
- Pull request#477Schwidola06072025-06-27 06:46
- Issue comment#338lzy-edu2025-05-22 07:442nodes, 4 gpu, tp=2,pp=2, timeout
- Issue comment#2399LLPPLL62025-05-08 06:10ZERO_STAGE doesn't seem to be working
- Issue comment#4569LLPPLL62025-05-08 05:09[BUG] AttributeError: 'list' object has no attribute 'to'
- Issue comment#278SeekPoint2025-05-06 08:40Details on ZeRO++ tutorials
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 204 stars here means stars gained during the window, not the repo's star count.