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
Give this work a before-and-after number, and find out the hardening already worked (#10996)
f13208b Daniel Han committed 2 months ago
Studio runtime keeps the GPU when nvidia-smi cannot answer, and prefers a GPU-capable llama-server beside a CPU build (#11005)
c829b33 Daniel Han committed 2 months ago
Studio installers read the NVIDIA driver library when nvidia-smi cannot answer, and setup.ps1 keeps a working GPU prebuilt over a source build (#11007)
3a0dae2 Daniel Han committed 2 months ago
Studio: gate tool file access outside the session sandbox (#10907)
a76c38c Daniel Han committed 2 months ago
Studio: stop a short chat video attachment from taking minutes to prefill (#10973)
4f56de1 oobabooga committed 2 months ago
Measure whether the host property can replace the native console thunk (#10984)
9bef060 Daniel Han committed 2 months ago
Format main to its own pinned hook, so pre-commit autofix can land again (#11019)
bda17fd Daniel Han committed 2 months ago
Capture which antivirus product is actually blocking installs, and stop discarding the false-positive submission packets (#10986)
7e128c0 Daniel Han committed 2 months ago
Studio: stop iOS Safari zooming into chat text fields (#10972)
9bf25a1 oobabooga committed 2 months ago
Fix diffusion load cancellation during construction (#10780)
93b3ded Michael Han committed 2 months ago
Studio Linux update: torch repair and the source build read the NVIDIA driver library, a CPU wheel recorded as CUDA is repaired, a satisfied pin keeps the GPU prebuilt (#11008)
1ee14f0 Daniel Han committed 2 months ago
Studio: Make installed Ollama models available through the API (#10763)
1286590 Nilay committed 2 months ago
fix(amd): let a pinned torch index carry a package extra, so RDNA1 can reach TheRock (#9138)
52fb7b0 Leo Borcherding committed 2 months ago
fix(AMD): recognise the RX 6950 XT, 6850M XT and 6550M instead of leaving them on CPU torch (#10746)
2e0340a Leo Borcherding committed 2 months ago
Linux: index AMD visible-device masks by device, and floor gfx1102/RDNA4 to rocm6.4 (#9672)
838f4bd Leo Borcherding committed 2 months ago
tests: pin the amd/huggingfacenotorch extras and name the audit failure (#7536)
d601e5c Leo Borcherding committed 2 months ago
fix(amd): turn on PyTorch's fast ROCm attention kernels on import unsloth, and drop the installers' host-wide export (#8821)
164d154 Leo Borcherding committed 2 months ago
Studio: download image model assets without loading (#10789)
217614f Michael Han committed 2 months ago
Pre-import diffusers in the post-warm worker so the first image load does not (#10989)
405b2a5 Daniel Han committed 2 months ago
Stop committing the two fixture archives that antivirus vendors flag (#10979)
0a12e13 Daniel Han committed 2 months ago
Studio prebuilt selection: Windows CUDA lines ordered by detected DLLs rather than filtered, portable fallback attempt, keep paths require SM coverage (#11006)
23146c6 Daniel Han committed 2 months ago
Close the partially initialized torch._dynamo window on the diffusion load path (#10981)
5fbbd61 Daniel Han committed 2 months ago
Studio: load the user's own Pi extensions and packages in unsloth start pi (#10971)
8fb25b7 oobabooga committed 2 months ago
Stop the video gate test failing on a rename it should not care about (#11003)
429f5b3 Daniel Han committed 2 months ago
Run the chat width presets in CI, and two guards main is currently red on (#11001)
9bdc7c9 Daniel Han committed 2 months ago
Desktop: extend the NVIDIA X11 WebKit fence fix to native packages (#10964)
b944ebf oobabooga committed 2 months ago
Say which unsloth_zoo revision Core tested, and warn when it is not main (#10988)
ae295db Daniel Han committed 2 months ago
Time the stall window test instead of counting polls (#10987)
e60f48c Daniel Han committed 2 months ago