All skills
training-system-planning
Estimates memory, step time, checkpoints, failure rework and deadlines for training or RL jobs.
- Part of bojieli/ai-infra-book
- Machine Learning · Training frameworks
More Training frameworks skills
See all- dali-dynamic-mode2.1kWrites and migrates NVIDIA DALI dynamic-mode data pipeline code for you.
- cuml30Writes, runs, and debugs GPU machine learning code with cuML, including zero-change scikit-learn acceleration.
- pytorch-fsdp22Trains large models with fully sharded data parallelism in PyTorch.
- pytorch-lightning20Writes clean PyTorch training loops with built-in distributed support.
Also in this pack
- accelerator-selectionCompares accelerators on capacity, effective compute, memory bandwidth, power and cost for a given model workload.
- agent-runtime-capacityPlans how agent model calls and tool environments are created, kept, released and restored for capacity and latency.
- cluster-network-analysisEstimates cross-server traffic, network bottlenecks and sync waits for multi-node training or inference.
- distributed-inferenceCompares cross-instance inference options like replicas, prefill/decode split, expert placement and shared KV cache.