Commits

Branch selector

 master

User selector

All users

All time

Commit history

Commits on May 28, 2026

• fix: correct tae for models that use the flux2 vae (#1571)

  wbruna

  authored
  0e4ee04

  Copy full SHA for 0e4ee04

Commits on May 27, 2026

• fix: resolve LLM norm tensor names by architecture (#1570)

  leejet

  authored
  29ab511

  Copy full SHA for 29ab511

• refactor: simplify diffusion model runner params (#1569)

  leejet

  authored
  55c2aed

  Copy full SHA for 55c2aed

• fix: preserve frontend tooling in ROCm CI build (#1568)

  leejet

  authored
  8eded49

  Copy full SHA for 8eded49

Commits on May 26, 2026

• feat: add microsoft lens support (#1560)

  leejet

  authored
  92dc726

  Copy full SHA for 92dc726

• fix: skip permission denied errors in recursive_directory_iterator (#1564)

  Show description for 07b2b18
  

  schirik

  and

  Serge F. Chirik

  authored
  07b2b18

  Copy full SHA for 07b2b18

Commits on May 25, 2026

• fix: package ROCm BLAS runtime in Windows artifacts (#1562)

  leejet

  authored
  1ceb5bd

  Copy full SHA for 1ceb5bd

• fix: use flux flow prediction for LTXAV (#1561)

  leejet

  authored
  202c615

  Copy full SHA for 202c615

Commits on May 23, 2026

• feat: add Longcat-Image / Longcat-Image-Edit support (#1053)

  Show description for a397e03
  

  stduhpf

  and

  leejet

  authored
  a397e03

  Copy full SHA for a397e03

• fix: make macOS binaries use relocatable rpaths (#1552)

  leejet

  authored
  72e512a

  Copy full SHA for 72e512a

Commits on May 22, 2026

• feat: add LTX temporal latent upscaler support (#1551)

  leejet

  authored
  0baf721

  Copy full SHA for 0baf721

• feat: add LTX rational latent upscaler (#1549)

  leejet

  authored
  645e6e9

  Copy full SHA for 645e6e9

• fix: strip trailing latent channels for preview decode (#1548)

  stduhpf

  authored
  cbf9219

  Copy full SHA for cbf9219

• fix: load TAESD preview-only model correctly (#1547)

  stduhpf

  authored
  8cf55a3

  Copy full SHA for 8cf55a3

Commits on May 21, 2026

• refactor: unify extra argument parsing (#1540)

  leejet

  authored
  3a8788c

  Copy full SHA for 3a8788c

• feat: stream LTX VAE temporal tile decoding (#1539)

  leejet

  authored
  449165c

  Copy full SHA for 449165c

• Feat: Temporal tile custom size with overlap (#1510)

  Show description for adaa599
  

  stduhpf

  and

  leejet

  authored
  adaa599

  Copy full SHA for adaa599

• perf: run LTX audio VAE decode in one ggml graph (#1538)

  leejet

  authored
  2e35146

  Copy full SHA for 2e35146

• feat: add taeltx2_3_wide support (#1535)

  stduhpf

  authored
  47d8198

  Copy full SHA for 47d8198

Commits on May 20, 2026

• feat: add graph cut markers for LTXAV transformer (#1534)

  leejet

  authored
  ef92a00

  Copy full SHA for ef92a00

• feat: add LTX spatial latent upscale hires support (#1533)

  leejet

  authored
  b3374e6

  Copy full SHA for b3374e6

• feat: add taeltx2/taeltx2.3 support (#1531)

  stduhpf

  authored
  bdd937f

  Copy full SHA for bdd937f

• fix: always load runtimle lora params on runtime backend (#1532)

  stduhpf

  authored
  c51ec7c

  Copy full SHA for c51ec7c

Commits on May 19, 2026

• fix: avoid Vulkan f16 repeat in LTX audio VAE (#1528)

  leejet

  authored
  5b0267e

  Copy full SHA for 5b0267e

• fix: trigger ci for docker image changes (#1527)

  leejet

  authored
  0045a72

  Copy full SHA for 0045a72

• fix: update sycl docker image to oneapi 2025.3 (#1526)

  leejet

  authored
  99bd062

  Copy full SHA for 99bd062

• fix: build web UI for Windows ROCm server releases (#1525)

  leejet

  authored
  9d8c9e4

  Copy full SHA for 9d8c9e4

Commits on May 18, 2026

• ci: add RDNA1 + RDNA2 targets for ROCm 7.13 (#1511)

  sofiageo

  authored
  caa823a

  Copy full SHA for caa823a

• sync: update ggml (#1520)

  leejet

  authored
  22c8c40

  Copy full SHA for 22c8c40

• fix: restore singleton dims for LLM outputs (#1518)

  leejet

  authored
  b706d68

  Copy full SHA for b706d68

• fix: only enable TAE after successful load (#1517)

  leejet

  authored
  b758b7d

  Copy full SHA for b758b7d

• feat: make negative max_vram control the amount of spare vram (#1503)

  wbruna

  authored
  f683c88

  Copy full SHA for f683c88

• ci: add CUDA Docker image support for NVIDIA Spark GB10 (#1512)

  ich777

  authored
  21fd4e6

  Copy full SHA for 21fd4e6

Commits on May 17, 2026

• docs: update news

  leejet

  committed
  8308042

  Copy full SHA for 8308042

• ci: add inactive pr clean up workflow

  leejet

  committed
  82e03ef

  Copy full SHA for 82e03ef