-
Notifications
You must be signed in to change notification settings - Fork 391
- #3754 · anwithk opened
on May 8, 2026
Issues
is:issue state:open
is:issue state:open
Issue creation is restricted in this repository
Search results
[bug] CheckpointConfig.async_strategy is silently ignored on the sync save path
area:ckptCheckpoint conversion, loading, export, and save pathsCheckpoint conversion, loading, export, and save pathsbugSomething isn't workingSomething isn't workingneeds-triageNew item needs classification and ownershipNew item needs classification and ownershipStatus: Open.#4639 In NVIDIA-NeMo/Megatron-Bridge;[bug] Dropless DeepEP makes GPT-OSS 20B logits depend on other EP-rank samples
area:trainingTraining loop, callbacks, and runtime integrationTraining loop, callbacks, and runtime integrationbugSomething isn't workingSomething isn't workingneeds-triageNew item needs classification and ownershipNew item needs classification and ownershipStatus: Open.#4635 In NVIDIA-NeMo/Megatron-Bridge;PoR: Deepseek v4 Roadmap
area:perfPerformance optimizations and benchmarkingPerformance optimizations and benchmarkingfeatureNew capabilities, enhancements, or enablement workNew capabilities, enhancements, or enablement workPoRPlan of record item for roadmap and release trackingPlan of record item for roadmap and release trackingtrackingTracking issue for an ongoing project with smaller stepsTracking issue for an ongoing project with smaller stepsStatus: Open.#4633 In NVIDIA-NeMo/Megatron-Bridge;[bug] Gemma4 Dense model lacks final logit softcapping
area:modelModel implementations and HF bridge logicModel implementations and HF bridge logicbugSomething isn't workingSomething isn't workingwaiting-on-maintainersWaiting on maintainers to respondWaiting on maintainers to respondStatus: Open.#4610 In NVIDIA-NeMo/Megatron-Bridge;[feature] intra-microbatch reordering for MegatronMIMO (+ sequence packing, scalable DP)
area:dataDataset builders, preprocessing, and samplersDataset builders, preprocessing, and samplersfeatureNew capabilities, enhancements, or enablement workNew capabilities, enhancements, or enablement workwaiting-on-maintainersWaiting on maintainers to respondWaiting on maintainers to respondStatus: Open.#4609 In NVIDIA-NeMo/Megatron-Bridge;[tracking] Long-context (128k) extension example for Megatron-Bridge
area:trainingTraining loop, callbacks, and runtime integrationTraining loop, callbacks, and runtime integrationfeatureNew capabilities, enhancements, or enablement workNew capabilities, enhancements, or enablement workPoRPlan of record item for roadmap and release trackingPlan of record item for roadmap and release trackingtrackingTracking issue for an ongoing project with smaller stepsTracking issue for an ongoing project with smaller stepsStatus: Open.#4597 In NVIDIA-NeMo/Megatron-Bridge;[tracking] SFT dataset unification / upstreaming with Megatron Core
area:dataDataset builders, preprocessing, and samplersDataset builders, preprocessing, and samplersfeatureNew capabilities, enhancements, or enablement workNew capabilities, enhancements, or enablement workmlm-syncRequires API/behavior sync with upstream Megatron-LM changesRequires API/behavior sync with upstream Megatron-LM changesPoRPlan of record item for roadmap and release trackingPlan of record item for roadmap and release trackingtrackingTracking issue for an ongoing project with smaller stepsTracking issue for an ongoing project with smaller stepsStatus: Open.#4596 In NVIDIA-NeMo/Megatron-Bridge;[tracking] Training loop upstreaming from Megatron-Bridge to Megatron Core
area:trainingTraining loop, callbacks, and runtime integrationTraining loop, callbacks, and runtime integrationfeatureNew capabilities, enhancements, or enablement workNew capabilities, enhancements, or enablement workmlm-syncRequires API/behavior sync with upstream Megatron-LM changesRequires API/behavior sync with upstream Megatron-LM changesPoRPlan of record item for roadmap and release trackingPlan of record item for roadmap and release trackingtrackingTracking issue for an ongoing project with smaller stepsTracking issue for an ongoing project with smaller stepsStatus: Open.#4595 In NVIDIA-NeMo/Megatron-Bridge;PoR: GLM-5.2 packed SFT Bridge data/config blockers
area:dataDataset builders, preprocessing, and samplersDataset builders, preprocessing, and samplersbugSomething isn't workingSomething isn't workingPoRPlan of record item for roadmap and release trackingPlan of record item for roadmap and release trackingtrackingTracking issue for an ongoing project with smaller stepsTracking issue for an ongoing project with smaller stepsx-perplexityExternal request: PerplexityExternal request: PerplexityStatus: Open.#4593 In NVIDIA-NeMo/Megatron-Bridge;[bug] Qwen3-30B-A3B fp8_mx pretrain OOMs with official defaults on 8x B300, while bf16 and nvfp4 run successfully
area:perfPerformance optimizations and benchmarkingPerformance optimizations and benchmarkingbugSomething isn't workingSomething isn't workingwaiting-on-customerWaiting on the original author to respondWaiting on the original author to respondStatus: Open.#4587 In NVIDIA-NeMo/Megatron-Bridge;[support] Dynamic (Hybrid) Context Parallelism: enablement status and end-to-end verification in Megatron-Bridge
area:trainingTraining loop, callbacks, and runtime integrationTraining loop, callbacks, and runtime integrationfeatureNew capabilities, enhancements, or enablement workNew capabilities, enhancements, or enablement workPoRPlan of record item for roadmap and release trackingPlan of record item for roadmap and release trackingtrackingTracking issue for an ongoing project with smaller stepsTracking issue for an ongoing project with smaller stepsStatus: Open.#4586 In NVIDIA-NeMo/Megatron-Bridge;[model] Add Megatron-Bridge support for MiniMax M3
area:modelModel implementations and HF bridge logicModel implementations and HF bridge logicfeatureNew capabilities, enhancements, or enablement workNew capabilities, enhancements, or enablement workPoRPlan of record item for roadmap and release trackingPlan of record item for roadmap and release trackingtrackingTracking issue for an ongoing project with smaller stepsTracking issue for an ongoing project with smaller stepsStatus: Open.#4585 In NVIDIA-NeMo/Megatron-Bridge;