Plan and troubleshoot Nemotron retrieval recipes for embeddings, which turn text into searchable vectors, and reranking, which reorders search results. Review configurations, prepare training and evaluation commands, and check export or deployment requirements.
Add support for new model architectures in NeMo AutoModel, from inspecting configuration files to implementing layers, converting saved weights, registering models, and testing numerical agreement with Hugging Face implementations.
Troubleshoot applications where NeMo Relay is installed but runtime behavior is missing or incorrect. It helps isolate loading failures, inactive execution contexts, missing events, and plugin or export configuration issues before changing application settings.
Configure CPU offloading in Megatron Bridge to reduce GPU memory use by moving intermediate training data or optimizer states to CPU memory. Check compatibility, adjust the offloaded fraction, and investigate memory errors or slowdowns.
Migrate existing NeMo Flow code and documentation to NeMo Relay by updating explicit names, imports, paths, and configuration references. A bundled helper previews changes, while language-specific checks cover dependencies, generated files, and remaining legacy identifiers.
Run training with mock or real data in Megatron-LM and Megatron Bridge, match their settings, and compare loss curves, which show training error over time. Includes recipes and multi-GPU examples for checking training behavior.
Build, migrate, and review adapters that connect external agent runtimes to NVIDIA NeMo Fabric. It guides configuration translation, startup and shutdown, discovery metadata, and validation against the public contract, the rules defining how adapters interact with Fabric.
Deploy, run, and test NVIDIA speech recognition, speech synthesis, and translation services in the cloud or on your own infrastructure. Guidance covers model selection, environment checks, custom models, and adjustments to speech processing and pronunciation.
Write and review NeMo-RL documentation using its project conventions. Keep the documentation index aligned with added or renamed Markdown files, document new features in the appropriate location, and format function and class docstrings in Google style.
Diagnose GPU out-of-memory errors during Megatron Bridge training and choose memory reduction settings. Covers fragmentation, LoRA input storage, recomputation, parallelism changes, and CPU offloading, explaining compatibility limits and potential tradeoffs between memory use and training throughput.
Helps tune activation recomputation in Megatron Bridge, recalculating intermediate training values to reduce GPU memory use. Guides selective or full-layer settings, checks compatibility, and compares memory peaks with training speed to assess whether a configuration completes training steps.
Tune adaptive behavior in an instrumented NeMo Relay application using runtime data. Configure tool parallelism, prompt caching, and state storage, then compare changes against a baseline to check latency, correctness, and failure rates.
Choose how to split large-model training across GPUs in Megatron Bridge. This guide helps combine parallelism methods, estimate GPU requirements, match settings to hardware connections, and check memory use and performance after configuration changes.
Choose how to distribute tokens, the units a model processes, across MoE model experts during training. The guide compares alltoall, DeepEP, and HybridEP by GPU hardware, parallelism scale, memory needs, and optimization stage.
Configure observability in NeMo Relay 0.6 or 0.7 to inspect recorded activity. Choose raw events, execution histories, or traces that show how calls unfold, then verify the selected output before adding more destinations.
Search documents through NeMo Retriever MCP and use retrieved passages to support answers. Add missing documents when server access is confirmed, then search them while preserving available document identifiers and page numbers for citations.
Configure Megatron-Bridge to overlap communication with computation during distributed training. Review settings for tensor, data, and pipeline parallelism, check activation requirements, and use included tests to investigate configuration issues or throughput regressions.
Guides hierarchical context parallelism in Megatron-Bridge, which splits input sequences across nested GPU groups. Covers configuration, implementation locations, constraints, and log checks to help enable the feature and investigate memory errors or performance regressions.
Helps configure training across multiple GPUs or machines in NeMo AutoModel. Guides strategy selection, YAML settings, and parallelism choices that split model computation or data, with constraints for large models, long sequences, and expert models.
Guides typed data integration in NeMo Relay by choosing codecs, converters between application values and JSON. Helps preserve middleware edits and caller-visible responses while distinguishing application data conversion from provider-specific LLM request and response handling.
Convert single-node training or inference commands into multi-node Slurm batch jobs. Choose an appropriate process launch method, configure shared storage and containers, and investigate communication timeouts, launch errors, and GPU memory shortages.
Configure and validate communication overlap in Megatron-Bridge for MoE models, which distribute work among expert networks. Check compatibility, compare training speed, and investigate regressions across standard dispatch, DeepEP, HybridEP, and delayed weight-gradient computation.
Enable Megatron FSDP in Megatron-Bridge to distribute model training state across GPUs. Follow configuration and checkpoint requirements, investigate memory failures or performance regressions, and verify the setup with an existing two-GPU functional test.
Configure fault detection and recovery for distributed training in Megatron Bridge. The skill covers slow GPU detection, automatic restarts, job interruption handling, and checkpoint saving, with configuration examples and guidance for investigating hangs or failed recovery.
Is this your account? Link GitHub to bring it onto your creator page (Japanese)