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.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
Commits on main
Studio: let a finished reply be edited when its settle write was refused (#14131)
ea71dc5 Daniel Han committed 1 month ago
Studio: let the browser panel and web fetch work behind fake-IP DNS proxies (#14127)
14e050e Daniel Han committed 1 month ago
Studio: name unknown GGUF quant types below the bundled type count (#14124)
1e2cedd Daniel Han committed 1 month ago
Studio: keep a dataset's thinking text when training Qwen models (#14016)
2682857 Nilay committed 1 month ago
Studio: push an export to the Hub with the Hugging Face CLI login (#14047)
8f7ff42 Nilay committed 1 month ago
Studio: read uploaded CSV cells as written for embedding training (#14014)
ff68ecc Nilay committed 1 month ago
fix(studio): guard missing scrollbar-gutter on Mac desktop startup (#14111)
157b855 Sourav Rajvi committed 1 month ago
Studio: keep a chat's files in the Library after it moves into a project (#14044)
caffebd Nilay committed 1 month ago
Studio: fit knowledge base chunks to the embedding model's input limit (#14045)
0b75841 Nilay committed 1 month ago
Studio: keep skills created in Docker on the unsloth-studio volume (#13912)
cf0b66c Nilay committed 1 month ago
Create the branch before pushing a LoRA GGUF to a revision (#13915)
6f645a1 Nilay committed 1 month ago
Let the Gemma-4 chat template format tool-calling conversations (#14015)
32d5383 Nilay committed 1 month ago
Studio: tell Gemma 4 with Think off and GLM today's date (#14013)
254b253 Nilay committed 1 month ago
Studio: automatic GPU placement caches failed probes, skips MIG winners, never blocks the event loop (#14118)
a411258 Daniel Han committed 1 month ago
Tests: make the --help, media scan, installer comparer and harness resolver checks hold on every host (#14116)
4c5fb40 Daniel Han committed 1 month ago
CI: pin the Arrow and sqlite-vec tags the Windows ARM wheelhouse signs to commits (#14092)
c5c62c3 Daniel Han committed 1 month ago
Studio: a managed account's Hub dataset card may not make the server dial an http(s) URL (#14091)
ce3b7a5 Daniel Han committed 1 month ago
Studio: browser panel follows the system proxy on macOS, Windows and Linux (#14029)
b4a4855 Long Yixing committed 1 month ago
Studio: use the inline markdown table style in the Hub, file previews and the composer draft (#14087)
72d0d2d Michael Han committed 1 month ago
Studio: slide the selected pill on every segmented toggle (#14030)
aead524 Etherl committed 1 month ago
Studio Hub: On Device filters and sort, Exports & Fine-tunes, unsloth.ai links (#14083)
2b0ad0c Michael Han committed 1 month ago
Studio: align toast icons with the first line on multi-line toasts (#14082)
eaa9c85 Michael Han committed 1 month ago
Notebooks smoke: run the install cell the way Colab does (#14058)
e795149 Daniel Han committed 1 month ago
CUDA spoof: cover CUDA wheels on GPU-less runners, not only CPU wheels (#14056)
e0dc284 Daniel Han committed 1 month ago
Studio: keep automatic GPU placement inert on ordinary hosts (#14020)
2f63c76 Daniel Han committed 1 month ago
Studio: read whole-GPU uuid masks from nvidia-smi rows and cache the MIG set (#14017)
eb0617b Daniel Han committed 1 month ago
Studio: API page toolbar gets clearer icons and hover tooltips (#14048)
a83fedd Michael Han committed 1 month ago
Studio Hub: sidebar-matched pill icons, drop the Memory stat (#14049)
9f10d63 Michael Han committed 1 month ago
Studio: show a finished chat's sidebar dot in the accent colour (#14042)
13d7ac8 Michael Han committed 1 month ago