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
Patch TRL's served-teacher, GOLD and SDFT distillation trainers (#13580)
a43cf77 Daniel Han committed 1 month ago
Pass str target_modules (regex or "all-linear") to PEFT on the Llama path (#13567)
b71d711 Daniel Han committed 1 month ago
Fit Gemma 4 E2B on 8 GB cards on Windows and in Studio chat (#13425)
8e2f6dc Daniel Han committed 1 month ago
perf(studio): enter Zoo's exact verify, NAX and packed gated-delta scopes on Unsloth Studio's own MLX paths (#13404)
937d54e Long Yixing committed 1 month ago
Parse each forward once in the per-step loss-kwargs walk (#13585)
447cb3c Daniel Han committed 1 month ago
Keep torch's bundled flash on Hopper for narrow models under UNSLOTH_FLASH_ATTENTION_IMPL=auto (#13544)
a96a4a7 Daniel Han committed 1 month ago
Studio: popups dim without blur, with a faint shadow; Settings a step darker (#13577)
b73110c Michael Han committed 1 month ago
Studio: fix compressed pages, router paths and early empty reads in the browser panel (#13532)
aea512a oobabooga committed 1 month ago
Studio: optional notification when a chat reply finishes in the background (#13554)
1e7d784 Daniel Han committed 1 month ago
Re-measure the Studio startup budget after the file history and agent merges (#13586)
90a968c Daniel Han committed 1 month ago
Studio: fp8 / mxfp8 image LoRA training on every aspect bucket, mxfp8 for biased linears, fp8 sm89 floor (#13317)
3deec0e Daniel Han committed 1 month ago
Studio: Run scripts setting and per-site Unlabelled clicks rule for the browser agent (#13529)
3d64021 Michael Han committed 1 month ago
Studio: translate the chat welcome greeting in every locale (#13549)
a7dc8a0 Daniel Han committed 1 month ago
Studio: rename sidebar chats in the Rename chat dialog (#13575)
b45c8ac Michael Han committed 1 month ago
Studio: list VS Code and Pi as coding agents and reorder the agent picker (#13547)
0b81fb8 Michael Han committed 1 month ago
Studio: browser agent can click at a point, move the mouse, drag, and run approved scripts (#13505)
2c1923e Michael Han committed 1 month ago
Studio: stop LTX-2 loads crashing Windows with an access violation on 32 GB machines (#13432)
65e03e7 Daniel Han committed 1 month ago
Studio: skip a DFlash drafter that an image would break under Auto (#13553)
579809c Daniel Han committed 1 month ago
Fix int8-int4 QAT LoRA at low ranks and on tied-embedding models (#13552)
0943697 Daniel Han committed 1 month ago
Studio: run LoftQ when the LoftQ variant is selected (#13535)
882293f Daniel Han committed 1 month ago
Studio: faster image LoRA training (fused AdamW, compile nf4 bases) (#13236)
9ef7ef3 Daniel Han committed 1 month ago
fix(studio): handle harmless Node helper stderr on Windows (#13530)
bb3e31e Lee Jackson committed 1 month ago
Studio: run AutoRound checkpoints on vLLM, and tidy the engine dialog (#13405)
50e5b8a Daniel Han committed 1 month ago
Studio: keep Hub Run settings open on a first Chat visit (#13524)
8d450cb oobabooga committed 1 month ago
Studio: explicit XPU GPU ids on single-tile Intel hosts (#13540)
b16b139 Daniel Han committed 1 month ago
Skip bitsandbytes' per-parameter device sync in non-paged optimizer steps (#13389)
ed8e5e7 Daniel Han committed 1 month ago