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: keep a Codex chat working after a tool returns an image on a text-only model (#11476)
59ba458 Nilay committed 1 month ago
Studio: guard export operations when the Hub is unreachable (#11466)
8bb4322 Lee Jackson committed 1 month ago
studio: allow local imatrix files for gguf export (#11350)
e3cd413 Maheswar Kumar committed 1 month ago
UI scale contract: count the save-temporary-chat button among the chat header's 30px controls (#11932)
fa4a485 Daniel Han committed 1 month ago
Studio: make the composer the same width as the chat column (#11926)
953e64a Michael Han committed 1 month ago
Studio: move Read aloud and Edit response into the More menu (#11920)
2d219b0 Michael Han committed 1 month ago
Studio: clicking a gallery item no longer replaces the typed prompt (#11930)
d9b5f99 Michael Han committed 1 month ago
Stop the Starling, Yi-chat and LFM2 templates leaking whitespace (#11779)
96efda5 Abhishek B R committed 1 month ago
Studio: show Theme first in Settings > Appearance (#11921)
726f736 Michael Han committed 1 month ago
Studio: video auto precision keeps a resident bf16 DiT (#11831)
0349d9b Daniel Han committed 1 month ago
Unsloth Studio (AMD/ROCm): refuse a GPU the installed PyTorch has no kernels for (#11571)
4c17ada Leo Borcherding committed 1 month ago
Studio: style the chat scrollbar like Run settings (#11925)
f4ffddd Michael Han committed 1 month ago
Studio: align the model selector label and truncate long project names (#11917)
6b55ba7 Michael Han committed 1 month ago
Studio: start settings labels with the setting, not "Show" (#11924)
0cce5ac Michael Han committed 1 month ago
Studio: read DOCX content controls, tracked insertions and text boxes (#11803)
7c97396 Lukas committed 1 month ago
Unsloth Studio (AMD): floor torch at 2.11 on the gfx103X-all and gfx110X-all families too (#11834)
4ff930a Leo Borcherding committed 1 month ago
Load a VLM through its native image-text class when the repo's auto_map class is untrusted (#11613)
6eef21f Daniel Han committed 1 month ago
Studio: build Qwen-Image-2.1's token layout once per render instead of every step (#11887)
59b68c4 Daniel Han committed 1 month ago
Split the flash-attn prebuilt wheel build across parallel ccache jobs (#11812)
baa1da0 oobabooga committed 1 month ago
Studio: keep the model's own chat template when the Unsloth one can't render the rows (#11487)
3329218 oobabooga committed 1 month ago
Studio: add File and View menu items to the macOS desktop app (#11902)
2f68ec3 Michael Han committed 1 month ago
Fix text_only 4-bit load and generate for Gemma-4 and other VLM text configs (#11684)
ed18d9c Daniel Han committed 1 month ago
Studio: restyle the save temporary chat popup as a standard dialog (#11914)
5e73505 Michael Han committed 1 month ago
Studio: verify the Diffusers main zip against a pinned SHA-256 (#11908)
4af7e84 Daniel Han committed 1 month ago
Accept the transformers 5.0 ignore_keys argument in validate_rope (#11560)
de80174 Daniel Han committed 1 month ago
Studio Hub: show a running download as Downloading, not as a paused partial (#11797)
c71427c Daniel Han committed 1 month ago
Studio: follow the JSON format a client asks for on /v1/messages (#11854)
34992c3 Nilay committed 1 month ago