Unsloth
unsloth.ai
A fine-tuning library that makes QLoRA and LoRA training 2x faster and use 50% less memory than standard implementations, enabling fine-tuning of large models on consumer GPUs.
Why it is useful
Makes fine-tuning models that previously required expensive A100 GPUs practical on a single RTX 3090 or 4090. The speed and memory improvements come from custom CUDA kernels rather than approximations, so accuracy is identical to standard training. For anyone wanting to fine-tune Llama, Mistral, or Qwen on their own data without renting cloud GPUs, this is the starting point.
More in Developer Utilities
Other developer utilities you might not have heard of.
- SSL LabsFree SSL/TLS configuration testing tool that grades your server security and identifies vulnerabilities.
- CyberChefThe local-first digital Swiss Army knife for analyzing, converting, and transforming data formats through simple drag-and-drop recipes.
- SquooshBrowser-based image compression tool from Google that supports WebP, AVIF, and advanced codec options.
- HoppscotchOpen-source API development ecosystem that runs entirely in your browser.
- RequestlyBrowser-level HTTP interception and mocking tool — redirect requests, modify headers, and mock API responses without setting up a proxy.
- DevhintsA collection of concise, well-organised cheatsheets for languages, tools and frameworks, from Bash and Git to CSS, React and SQL.