nvidia-nemo/megatron-bridge

1

adding-model-support Guide for adding support for new LLM or VLM models in Megatron-Bridge. Covers bridge, provider, recipe, tests, docs, …

nvidia-nemo/megatron-bridge

2

2

nemo-mbridge-perf-tp-dp-comm-overlap Operational guide for enabling TP, DP, and PP communication overlap in Megatron-Bridge, including config knobs, code …

nvidia-nemo/megatron-bridge

2

3

parity-testing Structured framework for exact HF↔MCore weight verification, forward-pass logit correlation, and optional strict nume…

nvidia-nemo/megatron-bridge

2

4

testing Testing reference for Megatron Bridge — unit and functional test layout, tier semantics (L0/L1/L2/flaky), script conv…

nvidia-nemo/megatron-bridge

2

5

build-and-dependency Dev environment setup for Megatron Bridge — container-based development, uv package management, lockfile regeneration…

nvidia-nemo/megatron-bridge

1

6

bump-dependency Bump a pinned dependency (TransformerEngine, Megatron-LM, NRX, etc.), regenerate the lockfile, open a PR, and drive i…

nvidia-nemo/megatron-bridge

1

7

cicd CI/CD reference for Megatron Bridge — pipeline structure, commit and PR workflow, CI failure investigation, and commo…

nvidia-nemo/megatron-bridge

1

8

code-style Code style and quality guidelines for Megatron Bridge. Covers naming, type hints, ruff enforcement, keyword-arg safet…

nvidia-nemo/megatron-bridge

1

9

create-model-verification-card Create or update concise, agent-readable Megatron Bridge model verification cards. Use when adding a model support ca…

nvidia-nemo/megatron-bridge

1

10

developer-guide Developer environment setup, CI/CD workflows, and CI failure debugging for Megatron Bridge. Covers container-based de…

nvidia-nemo/megatron-bridge

1

11

linting-and-formatting Code style and quality rules for Megatron Bridge — ruff configuration, naming conventions, type hints, mypy rules, do…

nvidia-nemo/megatron-bridge

1

12

mlm-bridge-training Run Megatron-LM (MLM) and Megatron Bridge training with mock or real data. Covers correlation testing, available reci…

nvidia-nemo/megatron-bridge

1

13

multi-node-slurm Convert single-node scripts to multi-node Slurm sbatch jobs and debug common multi-node failures. Covers srun-native …

nvidia-nemo/megatron-bridge

1

14

nemo-mbridge-mlm-bridge-training Run Megatron-LM (MLM) and Megatron Bridge training with mock or real data. Covers correlation testing, available reci…

nvidia-nemo/megatron-bridge

1

15

nemo-mbridge-multi-node-slurm Convert single-node scripts to multi-node Slurm sbatch jobs and debug common multi-node failures. Covers srun-native …

nvidia-nemo/megatron-bridge

1

16

nemo-mbridge-perf-activation-recompute >- Validate and use selective and full activation recompute in Megatron Bridge to reduce GPU memory usage at the cost…

nvidia-nemo/megatron-bridge

1

17

nemo-mbridge-perf-cpu-offloading Validate and use CPU offloading in Megatron Bridge, including layer-level activation offloading and fractional optimi…

nvidia-nemo/megatron-bridge

1

18

nemo-mbridge-perf-cuda-graphs Validate and use CUDA graph capture in Megatron Bridge, including local full-iteration graphs and Transformer Engine …

nvidia-nemo/megatron-bridge

1

19

nemo-mbridge-perf-expert-parallel-overlap Validate and use MoE expert-parallel communication overlap in Megatron-Bridge, including overlap_moe_expert_parallel_…

nvidia-nemo/megatron-bridge

1

20

nemo-mbridge-perf-hierarchical-context-parallel Operational guide for enabling hierarchical context parallelism in Megatron-Bridge, including config knobs, code anch…

nvidia-nemo/megatron-bridge

1

21

nemo-mbridge-perf-megatron-fsdp Operational guide for enabling Megatron FSDP in Megatron-Bridge, including config knobs, code anchors, pitfalls, and …

nvidia-nemo/megatron-bridge

1

22

nemo-mbridge-perf-memory-tuning >- Techniques for reducing peak GPU memory in Megatron Bridge, including expandable segments, PEFT plus sequence-para…

nvidia-nemo/megatron-bridge

1

23

nemo-mbridge-perf-moe-comm-overlap MoE expert-parallel communication overlap in Megatron Bridge. Covers dispatch/combine overlap, flex dispatcher backen…

nvidia-nemo/megatron-bridge

1

24

nemo-mbridge-perf-moe-dispatcher-selection >- Select and validate an MoE token dispatcher (`alltoall`, DeepEP, or HybridEP) for a fixed workload and runtime. Co…

nvidia-nemo/megatron-bridge

1

25

nemo-mbridge-perf-moe-hardware-configs Representative, point-in-time MoE training playbooks by hardware and model family. Use them as candidate seeds, then …

nvidia-nemo/megatron-bridge

1

26

nemo-mbridge-perf-moe-long-context Long-context MoE training guidance for Megatron Bridge. Covers CP sizing, selective recompute, dispatcher choices, an…

nvidia-nemo/megatron-bridge

1

27

nemo-mbridge-perf-moe-optimization-workflow >- Evidence-gated workflow for MoE performance optimization in Megatron Bridge. Covers measurement contracts, the Thr…

nvidia-nemo/megatron-bridge

1

28

nemo-mbridge-perf-moe-vlm-training Practical guidance for training MoE VLMs in Megatron Bridge. Compares FSDP and 3D-parallel approaches, using rounded …

nvidia-nemo/megatron-bridge

1

29

nemo-mbridge-perf-parallelism-strategies Operational guide for choosing and combining parallelism strategies in Megatron Bridge, including sizing rules, hardw…

nvidia-nemo/megatron-bridge

1

30

nemo-mbridge-perf-sequence-packing Validate and use packed sequences and long-context training in Megatron-Bridge, including offline LLM packing, collat…

nvidia-nemo/megatron-bridge

1

31

nemo-mbridge-recipe-recommender Recommend and customize Megatron Bridge library and benchmark recipes for a user's model, GPU count, hardware, sequen…

nvidia-nemo/megatron-bridge

1

32

nemo-mbridge-resiliency Resiliency features in Megatron Bridge including fault tolerance, straggler detection, in-process restart, preemption…

nvidia-nemo/megatron-bridge

1

33

nemo-rl-e2e-testing External NeMo-RL end-to-end validation workflow for Megatron-Bridge model/provider changes, including downstream comp…

nvidia-nemo/megatron-bridge

1

34

recipe-recommender Recommend and customize Megatron Bridge recipes for a user's model, GPU count, and training goal. Indexes library rec…

nvidia-nemo/megatron-bridge

1

35

resiliency Resiliency features in Megatron Bridge including fault tolerance, straggler detection, in-process restart, preemption…

nvidia-nemo/megatron-bridge

1

36

review-pr Structured single-agent code review workflow for PRs, commits, and local diffs. Use when asked to review code, unders…

nvidia-nemo/megatron-bridge

1

37

verl-e2e-testing External verl end-to-end validation workflow for Megatron-Bridge changes.

nvidia-nemo/megatron-bridge

1