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: add chat on the AMD Ryzen AI NPU through Lemonade and FastFlowLM (#11743)
fd6f72f oobabooga committed 1 month ago
Studio: return freed host memory to the OS after diffusion and video unload (#11795)
3d2e576 Daniel Han committed 1 month ago
Prevent setup-size flash during desktop startup (#11910)
ec67ce3 Wasim Yousef Said committed 1 month ago
Studio: propagate the Ollama CUDA runtime to llama-server (replacement for #7563) (#11666)
828234f Wasim Yousef Said committed 1 month ago
Studio: skip xFormers when its torch requirement is unmet (#11847)
14ceddf oobabooga committed 1 month ago
Installer: stop picking cu126 when a slow NVIDIA driver times out CUDA detection (#11916)
5665b5a Daniel Han committed 1 month ago
Plan a device map for text_only loads of vision-language models (#11584)
4f10215 Daniel Han committed 1 month ago
Rebuild byte-level tokenizers that transformers v5 loads as LlamaTokenizer (#11686)
a46a993 Daniel Han committed 1 month ago
Studio: show the full URL while a web fetch waits for approval (#11694)
46a468d Wasim Yousef Said committed 1 month ago
Studio: draw one drop line per gap when dragging sidebar rows (#11942)
901d9fd Michael Han committed 1 month ago
Studio: reuse unchanged files from an older snapshot instead of re-downloading them (#11796)
5da3bb2 Daniel Han committed 1 month ago
Studio: list the Images workflows in the phone sidebar again (#11936)
35ee14e oobabooga committed 1 month ago
Studio: stop rewriting the compile-cache bundle on every warm start, and bound its disk use (#11874)
06ed937 Daniel Han committed 1 month ago
Studio: convert decoded video frames to uint8 on the GPU before the mp4 encode (#11879)
ff8b04f Daniel Han committed 1 month ago
Accept block_sequence_ids in chunked causal masks on transformers 5.17 (#11693)
4202d7e Daniel Han committed 1 month ago
Studio: translate the inline Read aloud and Edit response settings (#11933)
e8e44f2 Michael Han committed 1 month ago
Cast fp16 leftovers to the requested dtype after a text_only pre-quantized load (#11692)
23ab9ed Daniel Han committed 1 month ago
Studio: drag to reorder pinned models in the model picker (#11941)
454ecb8 Michael Han committed 1 month ago
Studio: show context checkpoints apart from the KV cache in the memory estimate (#11581)
6748f09 oobabooga committed 1 month ago
Studio: load the hosted INT8 pre-quant checkpoints on torchao 0.18 and later (#11884)
4786aff Daniel Han committed 1 month ago
Studio: show when a response was written, in its More menu (#11928)
680a6ba Michael Han committed 1 month ago
Studio: keep all text when adding Word files to a knowledge base (#11725)
86edc79 Nilay committed 1 month ago
Studio: build the ConvRot rotation from its definition (#11807)
3ed8f68 Daniel Han committed 1 month ago
Embed server tests: intercept only the server's own Popen (#11934)
e4939fe Daniel Han committed 1 month ago
Studio: stream durable chat runs at display frame rate (#11900)
a8d998f Michael Han committed 1 month ago
Studio: add a multiline send shortcut and spell out what each one does (#11927)
967b8ce Michael Han committed 1 month ago
Studio: show example prompts as placeholder hints (#11931)
e5fcd4c Michael Han committed 1 month ago