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: stop a Git Bash nul file from blocking every Windows MXC tool call (#13037)
a13d3a5 Daniel Han committed 1 month ago
Studio: retry web search over HTTP/1.1 when the connection is reset (#13032)
96f333e Daniel Han committed 1 month ago
Studio: avoid a TypeError after a tools module reload (#11506)
06c2ef1 Jiang Wu committed 1 month ago
Studio: stop MXC read grants looping on a Microsoft Store Python (#13025)
a72be55 Daniel Han committed 1 month ago
Studio: skip the Xet probe's GPU-init-off zoo retry on GPU hosts (#12478)
33515a2 Srijan Keshri committed 1 month ago
Fix notebook failures on Kaggle T4x2 and duplicate import warnings (#13024)
cd905b3 Daniel Han committed 1 month ago
Studio: show the real cause when a desktop install runs out of disk space (#12919)
d2b4139 Hunter Gordon committed 1 month ago
Load one tensor at a time when quantizing a 16-bit checkpoint on the fly (#12997)
3e78986 Leo Borcherding committed 1 month ago
Unsloth Studio / Desktop: show a cached image GGUF as Partial until Run has its text encoder and VAE (#12557)
8e2ab33 Leo Borcherding committed 1 month ago
Studio: duplicate a past training run into a new Configure draft (#12979)
f51a3ff Padias committed 1 month ago
Studio: keep browsing from a temporary chat out of browser history (#12984)
a1b00a3 Nilay committed 1 month ago
Studio: close refused tool calls under tool_choice none (#12627)
8706c6e BevKim committed 1 month ago
Studio: put the skill row chevron next to the skill name (#13033)
f8fc752 Michael Han committed 1 month ago
Studio: make the browser's right-click downloads work on macOS (#13003)
3afc66d Michael Han committed 1 month ago
Unsloth Studio (AMD): show each GPU's live VRAM and utilization under its own HIP id (#12961)
154a5ea Daniel Han committed 1 month ago
FastModel: fall back to Unsloth inference on GPUs older than Volta (#12959)
e9c8662 Daniel Han committed 1 month ago
install.sh: install for the discrete AMD GPU when an iGPU is listed first (#12963)
d346d8c Daniel Han committed 1 month ago
Studio: keep new image sets from merging into an existing one (#12987)
acf89c2 Nilay committed 1 month ago
Studio: run llama-server with a per-launch API key by default (#13011)
e0fa1a4 Daniel Han committed 1 month ago
Studio: keep float64 VAE tile blend weights off MPS so tiled encode and decode work (#12937)
ce62144 GokayAI committed 1 month ago
Studio: leave a bracketed IPv6 host unchanged in dial_host (#12733)
adf23cc Younes Beriane committed 1 month ago
Studio: make llama-fit-params executable after the macOS prebuilt install (#12917)
6a25e67 Jay Zhou committed 1 month ago
Studio: find nvidia-smi under WSL on every query, count Core Ultra Arc iGPUs as XPU, cut GPU masks at an invalid index (#12962)
cf01874 Daniel Han committed 1 month ago
Studio: serve whisper-server under a random per-launch request path (#13002)
8ff8ad4 Daniel Han committed 1 month ago
Update the flex large head dim mask tests to the unpadded causal-mask skip (#13018)
637fd54 Daniel Han committed 1 month ago
Studio: parse Yahoo's newer result layout in web search (#12980)
8fa3fd8 oobabooga committed 1 month ago
Studio: ask before numpy pickle loads, keep audio tags inline, cap the variable-prose regex (#13001)
aa860c8 Daniel Han committed 1 month ago
Studio: prefer llama-server for EmbeddingGemma in the embedding model picker (#13006)
d632c82 oobabooga committed 1 month ago