-
-
Notifications
You must be signed in to change notification settings - Fork 6.1k
Pull requests: unslothai/unsloth
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Auto-enable grouped MoE on loaded / PEFT'd models via loader hook
#6727
opened Jun 28, 2026 by
danielhanchen
Member
Loading…
Respect requested 4bit/8bit for local '-bf16' directories
#6726
opened Jun 28, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 16): route no-GPU loads to the native sd.cpp engine
#6724
opened Jun 28, 2026 by
danielhanchen
Member
Loading…
Studio: cascade user message deletion to include assistant reply
#6720
opened Jun 27, 2026 by
NilayYadav
Contributor
Loading…
Studio: size separate-drafter MTP draft KV at full context (swa_full)
#6719
opened Jun 27, 2026 by
danielhanchen
Member
Loading…
Studio: reserve CUDA context and mmproj/MTP soft overhead in the GGUF fit budget
#6718
opened Jun 27, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 15): build int8 pre-quantized checkpoints (skip M=1 modulation linears)
#6717
opened Jun 27, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 14): fix int8 dense quant on Flux / Qwen (skip M=1 modulation linears)
#6716
opened Jun 27, 2026 by
danielhanchen
Member
Loading…
fix(studio): show local file path tooltip for Hub-tab local models
#6715
opened Jun 27, 2026 by
mvanhorn
Contributor
Loading…
perf(dataprep): cache regex and field lists, fix typos
#6714
opened Jun 27, 2026 by
Muhammad-Ikhwan-Fathulloh
Loading…
Fix gpt-oss detection in save: config.architectures is a list, not a string
#6711
opened Jun 26, 2026 by
vineethsaivs
Contributor
Loading…
Add MLX backend support for CLI unsloth train
#6709
opened Jun 26, 2026 by
Lyxot
Contributor
Loading…
Add FP8/FP4 compressed export to save_pretrained_merged
#6706
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
feat(i18n): add Japanese translation support for Studio UI
#6705
opened Jun 26, 2026 by
DovahkiinYuzuko
Loading…
Studio diffusion (Phase 12): First-Block-Cache step caching for many-step DiT
#6703
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 11): prefer int8 on consumer GPUs in the auto ladder
#6702
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 10): attention-backend selection
#6701
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 9): pre-quantized transformer loading
#6700
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 8): opt-in fast transformer (torchao int8/fp8/fp4 on a dense source)
#6694
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
Whole-document context for RAG chat attachments
#6693
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
install: let UNSLOTH_TORCH_INDEX_FAMILY / _URL override CUDA wheel detection
#6692
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
Studio diffusion (Phase 7): accuracy-preserving speed pass (2.2x via GGUF compile)
#6690
opened Jun 26, 2026 by
danielhanchen
Member
Loading…
fix: reduce VRAM headroom fractions to prevent auto-context size spill (#6682)
#6688
opened Jun 25, 2026 by
anmolxlight
Contributor
Loading…
Studio: add OpenAI-compatible /v1/images/generations endpoint
#6686
opened Jun 25, 2026 by
oobabooga
Member
Loading…
Previous Next
ProTip!
Adding no:label will show everything without a label.