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
Unsloth Studio / Desktop: log what an image load and each generation actually resolved to (#11994)
2c18308 Leo Borcherding committed 1 month ago
Studio: stop MiniMax-H3 recompiling on the second caption and the first i2v (#11880)
6c19cf0 Daniel Han committed 1 month ago
Studio: size the image memory plan at the dtype the pipeline loads in (SDXL resident on 24 GB) (#11922)
9cf9a34 Daniel Han committed 1 month ago
Studio: stop the second prompt length recompiling Qwen-Image on the max tier and with int8 (#11842)
3cb9dd7 Daniel Han committed 1 month ago
Unsloth Studio: refuse picking a hosted FP8/INT8 checkpoint repo as a pipeline before it downloads (#11829)
c7ea18c Leo Borcherding committed 1 month ago
Studio: decode Wan video in fp16 with channels_last_3d convs (#11999)
c8d146f Daniel Han committed 1 month ago
Studio: stop edit_file writing a compacted-argument placeholder into files (#11950)
4489b5f oobabooga committed 1 month ago
Unsloth Desktop: show Stopping… after Stop so a pending cancel doesn't look ignored (#11976)
06c60d0 Leo Borcherding committed 1 month ago
Keep the conversion backfill's donor stub off the transformers package (#12034)
96b6316 Daniel Han committed 1 month ago
Unsloth Studio: list a GGUF with no header metadata under On Device on the Images page (#11830)
4c45445 Leo Borcherding committed 1 month ago
Studio: center collapsed sidebar icons and tighten the rail (#12031)
90b2c60 Michael Han committed 1 month ago
Studio: stop Python tool network calls from skipping the host allowlist (#11172)
032e43d oobabooga committed 1 month ago
Unsloth Studio: a companion fetch with no denoiser no longer blocks deleting its base (#11828)
5559ac9 Leo Borcherding committed 1 month ago
Read the hot-path I/O cost at its steady minimum across repeats (#12029)
e19760a Daniel Han committed 1 month ago
Unsloth Studio (AMD): keep export off a GPU PyTorch has no kernels for, like the iGPU (#11946)
52f2e77 oobabooga committed 1 month ago
Keep more VRAM headroom on Windows CUDA, and say when a hand-set context does not fit (#11368)
dac0dd0 Daniel Han committed 1 month ago
Studio: name the Git Bash MXC incompatibility and stop re-probing it (#12021)
22b80c2 Daniel Han committed 1 month ago
Studio: shadow the dark mode composer so it stands off the chat (#12023)
c27da94 Michael Han committed 1 month ago
Studio media viewer: Zoom to fit at the bottom of the scale menu, below a separator (#12028)
2f647e8 Michael Han committed 1 month ago
Studio: keep a CSV seed's values as written when dropping its index column (#11945)
9910194 Lukas committed 1 month ago
Unsloth Studio installer (AMD/Linux): send RDNA 4 cards (RX 9000, R9700) to AMD's gfx120X wheels (#11935)
cc44b7d Daniel Han committed 1 month ago
Give the real-host NVIDIA probe test a budget that a busy driver can meet (#12026)
88507f8 Daniel Han committed 1 month ago
Studio: keep a working GPU when one probe fails, and name the card behind no_gpu (#11944)
d2804cc Daniel Han committed 1 month ago
Studio: keep a 0 label and blank a NaN cell when mapping columns to chat roles (#11895)
f276fbd breken committed 1 month ago
Studio: return an error when embeddings dimensions can't be honored (#11968)
7959a29 Nilay committed 1 month ago
Studio: run image and video denoises on one render thread so cuDNN caches are reused (#11843)
6e046f9 Daniel Han committed 1 month ago
Studio: do not reapply ROCR_VISIBLE_DEVICES when picking the AMD card in setup.sh and the llama.cpp prebuilt probe (#11965)
5b28886 Daniel Han committed 1 month ago
Studio CLI: stop `unsloth run` re-exec'ing itself forever when the Studio venv is a symlink (#11788)
69a8fa3 Daniel Han committed 1 month ago