unslothai/unsloth
Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.
385 best points · 126 comments · 9 Hacker News posts · 79,135 GitHub stars · View on GitHub
Hacker News discussions
- Show HN: 80% faster, 50% less memory, 0% loss of accuracy Llama finetuning (385 points, 119 comments, by danielhanchen, 2023-12-01)
- Unsloth: 80% faster, 50% less memory, 0% accuracy loss Llama finetuning (132 points, 3 comments, by bratao, 2023-12-01)
- Unsloth Desktop: Open-Source App for Local Models (10 points, 2 comments, by thunderbong, 2026-08-11)
- Show HN: Unsloth Studio - Local Fine-tuning, Chat UI (8 points, 2 comments, by danielhanchen, 2026-03-17)
- Unsloth: Introducing AMD support (7 points, 0 comments, by wertyk, 2026-07-20)
- Show HN: Unsloth – finetune Llama 2x faster 50% less memory on your GPU (3 points, 0 comments, by unsloth, 2023-11-30)
- Unsloth Now Supports GRPO (3 points, 0 comments, by sadiq, 2025-02-06)
- Unsloth – Train LLMs 2x faster with 70% less VRAM (3 points, 0 comments, by jhack, 2025-12-10)
- Unsloth: Run and Train Local LLMs (2 points, 0 comments, by ungreased0675, 2026-08-03)