Search

AI agents running research on single-GPU nanochat training automatically

Single-GPU LLM fine-tuning preflight: memory estimates, runnable LoRA/QLoRA recipes, and local measurements.

High-performance single-GPU inference for selected model checkpoints and GPUs.

C++/CUDA single-GPU inference engine for Qwen3.8 (incl. Qwen3.8-Flash-Next with offloaded experts) on the RTX 5090; grew from NInfer