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: read the NVIDIA inventory from NVML or the CUDA driver when nvidia-smi cannot answer (#10985)
1e07231 Daniel Han committed 2 months ago
Studio: keep a working GPU llama.cpp prebuilt over a CPU-only source fallback, and say so in the footer (#10975)
6f64398 Daniel Han committed 2 months ago
Keep a passthrough SSE stream alive while llama-server prefills (#10911)
91b5adb Daniel Han committed 2 months ago
Studio: a Windows host whose NVIDIA GPU is hidden by CUDA_VISIBLE_DEVICES selects the CUDA bundle, not windows-cpu (#10976)
3712a3c Daniel Han committed 2 months ago
Studio: fix unsloth start when port 8888 is already in use (#10942)
676332d Nilay committed 2 months ago
fix: materialize strided LayerNorm inputs and gradients (#10675)
66a38c7 Chenghao Liu committed 2 months ago
Take the wall clock out of two backend guards that fail on a loaded runner (#10952)
4938d2b Daniel Han committed 2 months ago
A first tool call must not answer from inside a legacy move's staging window (#10953)
0bb3c06 Daniel Han committed 2 months ago
Name the Ascend NPU as its own tests expect, and close the spoofable skip-guard hole (#10968)
3fe93e0 Daniel Han committed 2 months ago
studio: centre the chat column by reserving the scrollbar gutter on both edges (#10966)
1811677 Michael Han committed 2 months ago
Studio: make studio/setup.sh read the recorded uv cache, and probe the cache the CLI hands it (#10647)
a3dbab3 Daniel Han committed 2 months ago
Studio: show the right reasoning effort and image options for every API model (#10933)
02751d2 Nilay committed 2 months ago
Studio: leave top_p out of external requests when set to Off (#10934)
54c76e1 Nilay committed 2 months ago
Studio: keep chat titles clear of the options hover circle (#10969)
f93c523 Michael Han committed 2 months ago
Studio: keep nested tool call fields llama.cpp drops when written out of order (#10935)
b3fbc5e Nilay committed 2 months ago
Studio: return string error codes from streaming /v1/responses (#10938)
feb3335 Nilay committed 2 months ago
Studio: fix Push to Hub for base and quantized merged exports on transformers 5 (#10943)
b83a7f5 Nilay committed 2 months ago
Studio: tell the model when a tool returns audio or a file (#10940)
cccb7aa Nilay committed 2 months ago
studio: configurable HuggingFace endpoint via HF_ENDPOINT env var (#4989)
20f1846 AdamPlatin123 committed 2 months ago
Studio: reinstall a missing PyTorch when unsloth studio update finds the package current (#10956)
9f8e13c oobabooga committed 2 months ago
Studio: make the model use the tool the client asks for (#10939)
f3cc1d7 Nilay committed 2 months ago
Tests: stop the live no-torch server test from uninstalling torch from the Studio venv (#10955)
f1ff13e oobabooga committed 2 months ago