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 created files outside the collapsed tool card (#10446)
2d9671d Chaitanya Laxman committed 2 months ago
Studio: keep the operator's package-manager policy in the installer (#10902)
6a4a041 Daniel Han committed 2 months ago
Docker: mount the host's LM Studio, Ollama and Hermes models into the container (#10936)
e1205f5 Nilay committed 2 months ago
studio: compact the chat composer and add animated expansion (#10863)
536e4a8 Maheswar Kumar committed 2 months ago
fix(studio): stop MLX native controls from duplicating in streamed replies (#10905)
de1f4a7 Long Yixing committed 2 months ago
Studio: build the blocked-command alternation once instead of per call (#10927)
c9e8c70 Daniel Han committed 2 months ago
fix(sentence_transformer): resolve device consistently with other branches (#10843)
c678258 li-lizhe committed 2 months ago
fix(chat_templates): remove CUDA hardcoded device from stopping criteria (#10684)
0f46188 li-lizhe committed 2 months ago
Studio: raise the micro-batch for the projectors that abort llama-server on an image (#10683)
c3329b9 Chakshu Dhannawat committed 2 months ago
Studio: a masked NVIDIA GPU must not install the CPU bundle over CUDA (#10909)
9e911ee Daniel Han committed 2 months ago
Studio: a Windows AMD host without usable ROCm takes Vulkan, not CPU (#10908)
8760d44 Daniel Han committed 2 months ago
Studio: stop a CPU torch handover from disarming the flavor invariant (#10906)
24f0627 Daniel Han committed 2 months ago
Studio: add Download Dataset button for Data Recipes (#10708)
d584692 Sourav Rajvi committed 2 months ago
Release GGUF context while tool approval is pending (#10673)
c2bfd21 Lee Jackson committed 2 months ago
Run the unsloth_zoo suite beside Core instead of inside it (#10919)
9395876 Daniel Han committed 2 months ago
Shard the Python 3.13 backend pytest leg three ways (#10920)
8018196 Daniel Han committed 2 months ago
Studio: Use saved Hugging Face login for model downloads (#10758)
5c04ec1 Nilay committed 2 months ago
Drop an MTP sidecar llama-server cannot load as --model-draft (#10197)
645a350 Long Yixing committed 2 months ago
feat(docker): add an AMD ROCm image, for RDNA2 through RDNA4 and CDNA (#10820)
28299ac Leo Borcherding committed 2 months ago
Use the dynamic cache for Gemma multimodal generation (#10903)
b58ea42 Daniel Han committed 2 months ago
Stop ten test assertions from measuring the runner instead of the code (#10897)
f82efb6 Daniel Han committed 2 months ago
fix(studio): share Run settings draft across sidebar and model picker (#10851)
03af220 Sourav Rajvi committed 2 months ago
Honor user tensor_split when auto tensor-parallel planner returns an even split (#10884)
a1598ff Chakshu Dhannawat committed 2 months ago
Studio: block sandbox bypass via command-position substitutions in terminal tool (#10860)
a09f376 veriepic committed 2 months ago
Studio: do not lose a DGX Spark's memory pool to the page cache when fitting context (#10704)
61eefe6 Daniel Han committed 2 months ago