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
Docker: stop JupyterLab from silently taking Studio's port, and correct three host-script messages (#10828)
a092906 oobabooga committed 2 months ago
Docker Studio: keep Studio's data on a volume without pinning its code (#10829)
6d7fd47 oobabooga committed 2 months ago
Fail closed when Studio exposure prompt is aborted (#10638)
c10bff8 Wasim Yousef Said committed 2 months ago
Skip the backgrounded-raw-bind prompt test on Windows (#10885)
98f9e68 Daniel Han committed 2 months ago
Studio: Stop training the word None into CSV rows with blank cells (#10813)
9cfaa9d Nilay committed 2 months ago
installer: the adaptive uv cache selection was unreachable on every writable install (#10626)
d728b5f Daniel Han committed 2 months ago
Docker: drop the baked uv cache from the Studio image and correct its labels and Hub page (#10827)
6b36c8e oobabooga committed 2 months ago
Fix Q-GaLore quantization of single-sign groups (#10841)
3bdb5a4 Vineeth Sai Varikuntla committed 2 months ago
Studio: Stop Deep Research deleting links from code in its report (#10814)
8aef4f9 Nilay committed 2 months ago
Replay tool-call arguments in the order the model generated (#10791) (#10845)
d85c2f0 jishnav committed 2 months ago
Name the method on the uploaded model card (#10816)
d4de144 Vineeth Sai Varikuntla committed 2 months ago
Size training VRAM with the model's attention kernel and 4-bit config (#10815)
6417d4b Nilay committed 2 months ago
fix(studio): avoid resident GGUF mappings for Windows no-reserve loads (#10618)
1592078 Lee Jackson committed 2 months ago
Studio: duplicate tool-call guard re-run fix (#10792) (#10846)
2d26c93 jishnav committed 2 months ago
Stop clean_text deleting every non-ASCII character (#10741)
d49efcf Vineeth Sai Varikuntla committed 2 months ago
Studio: Stop at the requested stop text on transformers models (#10812)
9360dcc Nilay committed 2 months ago
Studio: Let the agent run a command again after it edits a file (#10810)
701ba22 Nilay committed 2 months ago
Studio: Show an error when Anthropic fails a reply partway through (#10811)
a320992 Nilay committed 2 months ago
Studio: Export a full fine-tune as a real 16-bit model (#10808)
3a5a851 Nilay committed 2 months ago
Studio: Keep API keys from training with the server's Hugging Face login (#10809)
14dcd6c Nilay committed 2 months ago
Clear the metadata directory cache before scanning installed records (#10797)
f29beb2 小云 committed 2 months ago
Speed up the three slowest tests without changing what they check (#10867)
7bdace6 Daniel Han committed 2 months ago
Studio: add a Download all logs button to Settings > Logs (#10719)
a19c9bd Daniel Han committed 2 months ago
Pair the linearity measurements, so a busy runner stops reading as quadratic (#10868)
b3dfd08 Daniel Han committed 2 months ago
Studio: prefill MLX VLM prompts at mlx-vlm's default step, fix cold-path position ids (#10778)
ed27a3b Long Yixing committed 2 months ago
Ask the spoof-survival question in a subprocess, not in the worker (#10866)
1a6c33b Daniel Han committed 2 months ago
Studio: stop the log redactor backtracking quadratically on a cut ANSI sequence (#10721)
d7e5c6b Daniel Han committed 2 months ago
Studio: read the NVLink fabric from NVML instead of a 1.2s nvidia-smi shell-out (#10720)
0fe1e50 Daniel Han committed 2 months ago