Show HN: Profine – Profile and rewrite your PyTorch training loop on real GPUs
Profine is a CLI tool that profiles PyTorch code on real GPUs, suggests optimizations, and automatically rewrites the training loop. It uses a six-step pipeline (read, profile, interpret, suggest, edit, benchmark) and achieved a 3.11x speedup and 68.7% memory reduction on minGPT. Requires a Modal account and an LLM provider.
Show HN: Profine – Profile and rewrite your PyTorch training loop on real GPUs
Report an error
The correction channel is not available yet. You can copy the article reference below for later.
Correction instructionsSummary only
Under the source content permission, this page provides an editorial summary, key points, and a link to the original article.