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 prompt/completion message lists as one conversation (#12910)
4fa4db5 Nilay committed 1 month ago
fix(studio): stop a managed runtime when the client drops its stream (#12266)
95529f1 Aleksei Boklag committed 1 month ago
Studio: pick the quant of a GGUF dictation model in Voice settings (#12900)
1cd08cb Etherl committed 1 month ago
Studio: support current native builds and preserve CPU asset selection (#12902)
5a3d892 oobabooga committed 1 month ago
Composer settings driver: poll the submitted list instead of reading it once after the key press (#12920)
a1b73ad Daniel Han committed 1 month ago
Run shell suites and Windows browser checks in parallel (#12899)
06b9015 oobabooga committed 1 month ago
Frontend test: give the cold Vite SSR render in reasoning-source-render room on a loaded runner (#12903)
3000d91 Daniel Han committed 1 month ago
Studio: list every Transcribe ASR model in Voice settings (#12898)
e7cc9d5 Etherl committed 1 month ago
Studio: continue searching past unusable results and add a Wikipedia fallback (#12892)
0a15408 oobabooga committed 1 month ago
Studio: keep browser panel tooltips, toasts and menus visible over desktop web pages (#12895)
d17af33 oobabooga committed 1 month ago
Sandbox test: wait for the cache scan workers before counting them (#12896)
344a2be Daniel Han committed 1 month ago
Support every PEFT init_lora_weights option, with fast PiSSA and MiCA init (#6879)
c17ca64 Suchitra Malimbada committed 1 month ago
Baseline the eight unsloth-zoo 2026.10.1 findings after review (#12884)
6c3d915 Daniel Han committed 1 month ago
Studio: fill the Hebrew and Swedish strings that left the strict i18n check red on main (#12881)
94b26b3 Daniel Han committed 1 month ago
Tests: give the setup.ps1 download progress pwsh its own startup cache, and catch argv0-less pwsh spawns (#12882)
07d9f88 Daniel Han committed 1 month ago
Studio: free PyAV's per-thread scalers before a fork so preexec_fn spawns still exec (#12863)
8435a74 Daniel Han committed 1 month ago
Studio: run ComfyUI-format video quants on the int8 / fp8 runtimes (#12851)
e4911f5 Daniel Han committed 1 month ago
Repair four checks that went red on main with the 10-06 Studio merges (#12868)
73b43f6 Daniel Han committed 1 month ago
Studio: list embeddinggemma-2 first in the embedding model picker (#12870)
2216ba6 Michael Han committed 1 month ago
Bump install.sh / install.ps1 pins to unsloth>=2026.10.1, unsloth-zoo>=2026.10.1 (#12869)
c7f7ff8 Daniel Han committed 1 month ago
Studio UI test: drive the sandbox level picker that replaced the menu switch (#12866)
3a8c92a Daniel Han committed 1 month ago
Studio: stop loading EmbeddingGemma in float16 in the RAG embedder (#12864)
62ea2d7 oobabooga committed 1 month ago
EmbeddingGemma 2 support and embedding improvements (#12865)
9d90136 Daniel Han committed 1 month ago
Compile the Laya encoder for decision model training (#12778)
cb78293 Daniel Han committed 1 month ago
Do not print the 16bit LoRA notice when a quantization_config still quantizes (#12848)
333ad5c Daniel Han committed 1 month ago
Studio: convert the decoded image to uint8 on the device (#12847)
25546d0 Daniel Han committed 1 month ago
Studio: read chat replies aloud in a saved Audio voice (#12819)
7da8ab5 Etherl committed 1 month ago