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: train audio datasets on the columns the dataset check found (#11850)
ed55faa Nilay committed 1 month ago
Studio: fix document search for embedding models other than the default (#11853)
ff46f0f Nilay committed 1 month ago
Studio: stop inflating chat images by re-encoding every one to PNG (#11889)
996b6a1 oobabooga committed 1 month ago
Studio: initialize new chats before attaching documents (#11838)
43af73a Lee Jackson committed 1 month ago
SentenceTransformer: preserve masks for patched Gemma3 attention (#11866)
8d22913 Etherl committed 1 month ago
Studio: use the browser's timezone for today's date in chat (#11852)
6a16993 Nilay committed 1 month ago
Studio: stop opening a saved chat from re-running a Max Tokens reply (#11875)
8a738c1 Michael Han committed 1 month ago
Parallel-isolation guard: exempt the resolver's back-dated staleness precondition (#11894)
a0e21dd Daniel Han committed 1 month ago
studiobench: image_upload closes the menu it opened when it gives up (#11892)
944fbd0 Daniel Han committed 1 month ago
Studio: stop rescanning every model folder on each request that names a model not on disk (#11872)
9ff97b6 oobabooga committed 1 month ago
Studio: say the installer script is missing instead of showing the PowerShell logo (#11862)
f3a5c5e oobabooga committed 1 month ago
Studio: stop holding a passthrough response 5 s when a watcher swallows its cancel (#11859)
2806e93 oobabooga committed 1 month ago
Studio: stop unsloth start announcing a switch when --model names the file already loaded (#11873)
c2ac19d oobabooga committed 1 month ago
Give vLLM a valid top_k when fast_inference enables it in the GRPO trainer (#11675)
4c30b0d Daniel Han committed 1 month ago
Studio: give the Windows ROCm torchao stub a version transformers 5 can parse (#11640)
eb8165a Daniel Han committed 1 month ago
fix(studio): keep Stop generating visible while queuing (#9123)
e597898 Kenshi committed 1 month ago
Agent guides CI: pass optional installer flags only while the installer takes them (#11878)
0b99d88 Daniel Han committed 1 month ago
Studio: a sandbox read waits out a legacy move's staging window (#11876)
77ecdbb Daniel Han committed 1 month ago
Studio update launcher tests: give every test a private STUDIO_HOME (#11869)
1ecc01a Daniel Han committed 1 month ago
Eval step: restore UNSLOTH_RETURN_LOGITS even when evaluation raises (#11865)
5b5b78e Daniel Han committed 1 month ago
Padding-free gate tests: state that UNSLOTH_RETURN_LOGITS is unset (#11863)
e207e1e Daniel Han committed 1 month ago
Unsloth Studio (AMD/ROCm): don't turn on cudnn.benchmark for image and video generation (#11732)
f9bffe2 Leo Borcherding committed 1 month ago
Load-replacement contract: read the rollback guard, not its one-line spelling (#11817)
df946e0 Daniel Han committed 1 month ago
Prebuilt installers: retry a download the server drops (#11816)
2c629b1 Daniel Han committed 1 month ago
Replace an in-flight chat model load with the latest pick (#11697)
466d9c5 Wasim Yousef Said committed 1 month ago
Baseline the two huggingface_hub 1.33.0 / 2.0.0 findings after review (#11811)
f5a8a65 Daniel Han committed 1 month ago
Fix false ONNX rejection for models with native weights (#11695)
89ed32d Wasim Yousef Said committed 1 month ago
Studio: show the Custom size fields under Output size in unified Edit (#11690)
0cf8bbc oobabooga committed 1 month ago
Account matrix: cover the audio gallery move and add-to-project routes (#11789)
81e2d07 Daniel Han committed 1 month ago