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 an eagerly decoded image VAE contiguous on NVIDIA (#12035)
186b486 Daniel Han committed 1 month ago
Studio: chat attachment cards, chips and the Library viewer (#12017)
0db1cc5 Michael Han committed 1 month ago
Studio: document viewer for PDF, Word, Excel and PowerPoint, with origin links in Library (#12001)
23a87c0 Michael Han committed 1 month ago
Studio: honor response_format on the MLX backend with grammar-constrained decoding (#10180)
a00fcbb Long Yixing committed 1 month ago
Studio: stop generating when the client disconnects on a non-streaming request (#11961)
b3f8ffe Nilay committed 1 month ago
Fix the prebuilt wheel publish step and shorten the release notes to one sentence (#12068)
bbe44ab Daniel Han committed 1 month ago
Studio: reject echo, suffix and best_of on /v1/completions that llama-server ignores (#11964)
4697865 Nilay committed 1 month ago
Studio: stop the dense-quant probes pinning a CUDA context on every card of a multi-GPU host (#11954)
5592fc5 Leo Borcherding committed 1 month ago
Studio: don't stop long running tool calls that print nothing (#11969)
1d8c891 Nilay committed 1 month ago
Unsloth Studio installer (AMD/Windows): don't mistake ZLUDA for an NVIDIA GPU (#11736)
ecbb0d9 Leo Borcherding committed 1 month ago
Studio: find models in a Hugging Face cache folder added as a location (#11970)
49d6c01 Nilay committed 1 month ago
Studio: tell Safari which account the Studio password belongs to (#11858)
ced1753 Nilay committed 1 month ago
Stop unsloth chat and unsloth inference switching a loaded GGUF to a different quant (#11855)
1f186c6 Nilay committed 1 month ago
Studio: keep MLX elapsed time and duration when the final eval event omits them (#11967)
305b055 Nilay committed 1 month ago
Studio: keep the Images Cancel load pill off the model panel divider (#12079)
db60370 Leo Borcherding committed 1 month ago
Studio: put Reapply next to Generate on the image and video pages (#11974)
5d566d1 Leo Borcherding committed 1 month ago
Keep the padding-free column test off the datasets numpy formatter (#12082)
08f3dff Daniel Han committed 1 month ago
Studio: quieter settings headings and one section gap on every page (#12064)
d54b42a Michael Han committed 1 month ago
Studio: rework Appearance settings and add flavor color themes (#12030)
73858d5 Michael Han committed 1 month ago
Give the killed formatter grandchild as long to die as it had to start (#12086)
d4d224e Daniel Han committed 1 month ago
CI: CodeQL advanced setup that analyses only the languages a PR touches (#11998)
fdbc867 Daniel Han committed 1 month ago
Unsloth Studio (AMD/Windows): report an iGPU's used VRAM when a discrete card sits beside it (#11871)
79856a6 Leo Borcherding committed 1 month ago
Keep TRL's own RL config defaults and clamp preference max_length (#12069)
ec9cc2c Daniel Han committed 1 month ago
Back off and retry the update banner navigation on ERR_NO_BUFFER_SPACE (#12081)
d27631e Daniel Han committed 1 month ago
Studio: train a SQuAD answer's text, not the answers dict, when mapping columns to chat roles (#12056)
d013a3c breken committed 1 month ago
Studio: read personalization only once a first sign-in has changed its password (#12071)
def1b94 Daniel Han committed 1 month ago
Load #11526's text-core refusal in the save_method routing harness (#12077)
f367878 Daniel Han committed 1 month ago