NVIDIA/Megatron-LM
View on GitHubOngoing research training transformer models at scale
- Stars
- 17.9k
- Forks
- 4.5k
- Open beginner issues
- 0
- Indexed issues
- 419
- Avg merge
- 4d 6h
- Merged PRs (30d)
- 271
- Dominant language
- Python
- License
- No license data
- Last GitHub push
- Sep 15, 2026
- Latest indexed
- Sep 16, 2026
- Contributing guide
- Contributing guide
- Code of conduct
- No code of conduct
- Beginner labels
- good first issue
-
bug
NVIDIA/Megatron-LM#453 · 6 comments · 0 reactions · 0 assignees ·
-
enhancement module: data pipeline
NVIDIA/Megatron-LM#698 · 5 comments · 3 reactions · 0 assignees ·
-
bug community-request waiting-on-customer
NVIDIA/Megatron-LM#737 · 2 comments · 0 reactions · 0 assignees ·
-
bug community-request waiting-on-customer
NVIDIA/Megatron-LM#957 · 3 comments · 2 reactions · 0 assignees ·
-
enhancement
NVIDIA/Megatron-LM#1135 · 1 comment · 0 reactions · 0 assignees ·
-
bug community-request waiting-on-maintainers
NVIDIA/Megatron-LM#1138 · 8 comments · 2 reactions · 1 assignee ·
-
community-request enhancement module: data pipeline
NVIDIA/Megatron-LM#1139 · 3 comments · 0 reactions · 0 assignees ·
-
enhancement
NVIDIA/Megatron-LM#1198 · 1 comment · 0 reactions · 0 assignees ·
-
question
NVIDIA/Megatron-LM#1283 · 1 comment · 0 reactions · 0 assignees ·
-
bug community-request
NVIDIA/Megatron-LM#1310 · 4 comments · 1 reaction · 0 assignees ·
-
question
NVIDIA/Megatron-LM#1335 · 1 comment · 0 reactions · 0 assignees ·
-
bug community-request module: moe
NVIDIA/Megatron-LM#1350 · 5 comments · 0 reactions · 0 assignees ·
-
enhancement
NVIDIA/Megatron-LM#1354 · 4 comments · 1 reaction · 0 assignees ·
-
community-request enhancement waiting-on-maintainers
NVIDIA/Megatron-LM#1375 · 4 comments · 0 reactions · 1 assignee ·
-
question
NVIDIA/Megatron-LM#1380 · 7 comments · 0 reactions · 0 assignees ·
-
bug community-request waiting-on-customer
NVIDIA/Megatron-LM#1384 · 2 comments · 0 reactions · 0 assignees ·
-
enhancement
NVIDIA/Megatron-LM#1385 · 1 comment · 0 reactions · 0 assignees ·
-
community-request enhancement waiting-on-customer
NVIDIA/Megatron-LM#1386 · 2 comments · 0 reactions · 0 assignees ·
-
[QUESTION] plan to implement zero bubble pipeline or dual pipeline and MoE comm-comp overlapping Opencommunity-request module: moe question
NVIDIA/Megatron-LM#1399 · 3 comments · 10 reactions · 0 assignees ·
-
bug module: training
NVIDIA/Megatron-LM#1402 · 5 comments · 0 reactions · 0 assignees ·
-
community-request module: moe question waiting-on-maintainers
NVIDIA/Megatron-LM#1406 · 7 comments · 0 reactions · 1 assignee ·
-
bug community-request waiting-on-customer
NVIDIA/Megatron-LM#1407 · 7 comments · 0 reactions · 0 assignees ·
-
question
NVIDIA/Megatron-LM#1413 · 1 comment · 0 reactions · 0 assignees ·
-
enhancement
NVIDIA/Megatron-LM#1423 · 1 comment · 0 reactions · 0 assignees ·
-
community-request enhancement waiting-on-customer
NVIDIA/Megatron-LM#1435 · 2 comments · 0 reactions · 0 assignees ·
-
community-request enhancement waiting-on-customer
NVIDIA/Megatron-LM#1437 · 5 comments · 3 reactions · 0 assignees ·
-
question
NVIDIA/Megatron-LM#1443 · 1 comment · 1 reaction · 0 assignees ·
-
bug community-request module: moe waiting-on-customer
NVIDIA/Megatron-LM#1444 · 6 comments · 0 reactions · 0 assignees ·
-
bug community-request module: moe
NVIDIA/Megatron-LM#1445 · 15 comments · 0 reactions · 0 assignees ·
-
question
NVIDIA/Megatron-LM#1447 · 1 comment · 0 reactions · 0 assignees ·
-
community-request question
NVIDIA/Megatron-LM#1448 · 5 comments · 0 reactions · 0 assignees ·
-
bug community-request module: distributed waiting-on-maintainers
NVIDIA/Megatron-LM#1450 · 2 comments · 0 reactions · 0 assignees ·
-
community-request question waiting-on-customer
NVIDIA/Megatron-LM#1455 · 2 comments · 0 reactions · 0 assignees ·
-
bug
NVIDIA/Megatron-LM#1468 · 4 comments · 0 reactions · 1 assignee ·
-
bug community-request module: moe
NVIDIA/Megatron-LM#1469 · 2 comments · 0 reactions · 0 assignees ·
-
bug community-request
NVIDIA/Megatron-LM#1477 · 3 comments · 2 reactions · 0 assignees ·
-
bug community-request module: transformer engine
NVIDIA/Megatron-LM#1481 · 2 comments · 0 reactions · 0 assignees ·
-
community-request enhancement question
NVIDIA/Megatron-LM#1489 · 2 comments · 0 reactions · 0 assignees ·
-
bug community-request waiting-on-customer
NVIDIA/Megatron-LM#1494 · 6 comments · 1 reaction · 0 assignees ·
-
bug community-request waiting-on-customer
NVIDIA/Megatron-LM#1497 · 4 comments · 0 reactions · 0 assignees ·
-
bug community-request module: transformer engine waiting-on-customer
NVIDIA/Megatron-LM#1503 · 4 comments · 0 reactions · 1 assignee ·
-
community-request question
NVIDIA/Megatron-LM#1504 · 2 comments · 0 reactions · 0 assignees ·
-
bug community-request waiting-on-customer
NVIDIA/Megatron-LM#1509 · 3 comments · 0 reactions · 0 assignees ·
-
enhancement
NVIDIA/Megatron-LM#1511 · 2 comments · 0 reactions · 1 assignee ·
-
bug community-request module: moe
NVIDIA/Megatron-LM#1513 · 3 comments · 2 reactions · 0 assignees ·
-
bug community-request
NVIDIA/Megatron-LM#1515 · 2 comments · 0 reactions · 0 assignees ·
-
[QUESTION]Question about packed_seq_params in decoder intermediate ranks when using sequence packing Openmodule: multimodal question
NVIDIA/Megatron-LM#1518 · 1 comment · 1 reaction · 0 assignees ·
-
community-request module: moe question
NVIDIA/Megatron-LM#1521 · 2 comments · 0 reactions · 0 assignees ·
-
bug module: transformer engine
NVIDIA/Megatron-LM#1528 · 1 comment · 0 reactions · 0 assignees ·
-
bug
NVIDIA/Megatron-LM#1533 · 2 comments · 0 reactions · 0 assignees ·