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: answer current llama.cpp, whisper.cpp and Node installs from their markers (#10648)
a1ca9ec Daniel Han committed 2 months ago
Docker Studio: keep unsloth-studio-update --ref from leaving Studio dead (#10826)
3277f0d oobabooga committed 2 months ago
use the bundled icon first, and retry the cloudflared download (#10680)
332e258 Nilay committed 2 months ago
Fix MediaPageLink tooltip blocking Windows titlebar controls (#10253)
a197ed9 Sourav Rajvi committed 2 months ago
Prevent agent startup from waiting on model catalog scans (#10728)
c8ed02f Lee Jackson committed 2 months ago
Studio: adapt macOS tray artwork to the menu bar appearance (#10880)
dabb59f Wasim Yousef Said committed 2 months ago
Studio: keep a verified install when PyPI is unreachable and UV_OFFLINE is set (#10651)
c504802 Daniel Han committed 2 months ago
Studio: two edge failures the dependency-pass change left behind (#10915)
1336bd3 Daniel Han committed 2 months ago
Move the two highest-fire workflows off the ubuntu-latest label onto the same image (#10918)
fa16bf7 Daniel Han committed 2 months ago
Studio: cap accelerate below 1.15 on Windows, where 1.15 breaks all ROCm training (#10819)
910847e Leo Borcherding committed 2 months ago
Measure whether the runner label actually changes the queue wait (#10914)
a57f9fc Daniel Han committed 2 months ago
The install suite spends 1940s resolving the same few thousand paths over and over (#10916)
e29303b Daniel Han committed 2 months ago
Unbreak main: three files drifted out of the formatter, and one reads a file with the platform encoding (#10913)
85d0255 Daniel Han committed 2 months ago
Studio: skip dependency-pass steps whose evidence holds, and keep that evidence on Windows (#10649)
c38bb13 Daniel Han committed 2 months ago
perf(studio): integrate optional MLX MoE and decode optimizations (#10733)
2fec56f Long Yixing committed 2 months ago
installer: stop condemning a warm uv cache for a directory uv never writes (#10899)
a94ef51 Daniel Han committed 2 months ago
Share the setups and the child imports two test files were paying for twice (#10893)
d2b256a Daniel Han committed 2 months ago
Route every test that starts PowerShell through the shared pwsh runner (#10900)
b876d50 Daniel Han committed 2 months ago
Stop three studio meta-tests re-cutting and re-walking the same sources (#10886)
8bcf160 Daniel Han committed 2 months ago
Split Repo tests (CPU) into three shards, and take the shell suite off its critical path (#10882)
7f6219e Daniel Han committed 2 months ago
Pass Q-GaLore optimizer options to bitsandbytes by name (#10874)
9f18247 Vineeth Sai Varikuntla committed 2 months ago
Fix three CI failures that were not the PRs they failed on (#10896)
ec3dd7f Daniel Han committed 2 months ago
Accept the model transformers 5.x passes to create_optimizer (#10887)
47417fe Daniel Han committed 2 months ago
Run the Q-GaLore tests from the normal pytest job, not a workflow of their own (#10898)
7ef4ab4 Daniel Han committed 2 months ago
Advance the Q-GaLore Adam step counter once per update (#10842)
677a988 Vineeth Sai Varikuntla committed 2 months ago