
Guides an agent through building and changing NeMo AutoModel training and eval recipes
- Open source
- 4K GitHub stars
- AI models and LLM tooling


Guides an agent through building and changing NeMo AutoModel training and eval recipes

Guides an agent through enabling Megatron FSDP sharded training in Megatron-Bridge

Guides an agent through adding new model architectures to NeMo AutoModel

Guides choosing and combining TP, DP, PP, CP and EP degrees for Megatron Bridge training

Sets up NeMo AutoModel job launches on interactive, Slurm and SkyPilot runs

Guidance for choosing FSDP or 3D parallelism when training MoE vision-language models

Hardware-specific MoE training playbooks and candidate parallelism configs for Megatron

Guides choice of MoE token dispatcher (alltoall, DeepEP, HybridEP) by hardware and EP size

Picks Megatron Bridge training recipes and resizes parallelism for your GPU allocation

Guides long-context MoE training in Megatron Bridge: CP sizing, recompute and dispatcher

Guides an agent through reducing peak GPU memory in Megatron Bridge training runs

Runs iterative NeMo-RL experiments with git branches and TSV logs as a research ledger
The closest alternatives to Nemo automodel distributed training are Nemo automodel recipe development, Nemo mbridge perf megatron fsdp, Nemo automodel model onboarding, Nemo mbridge perf parallelism strategies and Nemo automodel launcher config.
Yes. Nemo automodel recipe development, Nemo mbridge perf megatron fsdp, Nemo automodel model onboarding have a free or open-source option.