TuneSalon AI makes fine-tuning open source LLMs accessible to everyone. No Python, no GPU setup, just describe what you want and start training. What you can do: - Generate training datasets from a description - Fine-tune Qwen, Mistral, Phi on cloud GPUs or locally - Chat with your fine-tuned model - Export as GGUF for TuneSalon desktop or any compatible tool - Share or sell adapters on the marketplace Free/private desktop app(Windows only) + cloud web platform.
Hey everyone! I'm Ambla, who just likes building stuffs based in Sydney, Australia. And I'm very excited to show my first baby, TuneSalon AI.
I built TuneSalon AI because of the huge gap. Fine tuning LLMs has a lot of potential, especially in this era where new AI features you can play with come out everyday. But it's locked behind the wall of technology, python, CUDA, GPU, code knowledge, etc. Most people who would benefit from custom AI, such as small business owners/domain experts/creators, don't have this technical skills to get started.
I'm not a developer or techy myself. I started this journey as non techy person and taught myself to build this, one step at a time. If I can fine tune LLMs, anyone can. and that's exactly the point of TuneSalon AI.
Desktop app is fully free&private, and it's also released as open source in Github. You can test that out as well!(Windows only for now)
I'm going to expand scope that Tunesalon AI covers, such as MoE LLMs and agentic specific flow finetuning not only for text only LLMs. It won't be easy journey but I'm so excited to dive into this.
I believe TuneSalon is not perfect, and has a lot of flaws. So, I'd love your honest feedback. Either good or bad, any feedback or question, I would appreciate it!
Report
No reviews yetBe the first to leave a review for TuneSalon AI