Skip to content
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
216 commits
Select commit Hold shift + click to select a range
2b00fce
feat: support async ulysses cp for flux (#480)
DefTruth Nov 26, 2025
2cbb731
chore: update support matrix (#484)
DefTruth Nov 26, 2025
1e3c69e
chore: update async ulysses cp docs (#486)
DefTruth Nov 27, 2025
e61498e
chore: update async ulysses cp refs (#487)
DefTruth Nov 27, 2025
92db44a
feat: support FLUX.2-dev Tensor Parallelism (#485)
gameofdimension Nov 27, 2025
54461e5
feat: support Hybrid cache + TP for 🔥FLUX.2 (#489)
DefTruth Nov 27, 2025
4cae178
feat: Add seq offload for 🔥FLUX.2 w/o parallel (#490)
DefTruth Nov 27, 2025
b9fd68e
feat: support 🔥FLUX.2 context parallel (#492)
DefTruth Nov 27, 2025
1f68bc1
feat: support torch profiler in cache-dit (#491)
BBuf Nov 28, 2025
bae27d6
feat: support 🔥z-image tensor parallel (#494)
gameofdimension Nov 28, 2025
245c8ab
feat: support lumina2 tensor parallel (#495)
gameofdimension Nov 28, 2025
bd2f685
feat: support cache for 🔥z-image (#496)
DefTruth Nov 28, 2025
4abc829
feat: support context parallel for 🔥z-image (#497)
DefTruth Nov 28, 2025
73fbd44
fix: temp FnB(n>0) workaround for z-image cache w/ cp (#499)
DefTruth Nov 28, 2025
29a8345
add profiler for flux tp and cp example (#501)
BBuf Nov 29, 2025
879e85c
chore: Update README.md (#502)
DefTruth Nov 29, 2025
9a5cdd9
feat: support FnB0 for z-image w/ cp (#503)
DefTruth Nov 29, 2025
8bd89ba
feat: support _sdpa_cudnn backend for cp (#504)
DefTruth Dec 1, 2025
e0aabd7
feat: support async ulysses cp for z-image (#505)
DefTruth Dec 1, 2025
0f891ea
feat: add all_to_all_single v2 (#507)
DefTruth Dec 1, 2025
ed9a226
feat: support async ulysses cp for qwen-image (#508)
DefTruth Dec 1, 2025
6137662
feat: support all2all qkv per token fp8 (#509)
triple-mu Dec 2, 2025
ad1db02
chore: improve flux2 and qwen examples (#512)
BBuf Dec 2, 2025
88d7ae0
fix uaa-fp8 .view compile error (#514)
triple-mu Dec 2, 2025
60b3f2f
feat: relaxed transformer strict assert (#515)
DefTruth Dec 2, 2025
1dd51a2
feat: all2all qkv fp8 for ulysses (#516)
DefTruth Dec 2, 2025
6e24363
feat: support pre-defined step masks (#517)
DefTruth Dec 2, 2025
1e582fe
chore: separate chrono-edit and wan cp plan (#519)
DefTruth Dec 3, 2025
025c802
fix example utils.py uaa fp8 flag typo (#521)
DefTruth Dec 3, 2025
1253c6a
feat: extend predefined step masks for 4/6 steps (#523)
DefTruth Dec 3, 2025
2a39a20
misc: add z-image-turbo predefined step masks (#525)
DefTruth Dec 3, 2025
85549dc
feat: support per_token_quant_fp8 triton kernel (#524)
triple-mu Dec 3, 2025
4fd9984
feat: unified async ulysses fp8 (#526)
DefTruth Dec 4, 2025
a3b9540
feat: support serving for cache-dit (#522)
BBuf Dec 4, 2025
a66cd6f
fix get_model_info error when serving with tp/cp (#529)
BBuf Dec 5, 2025
814d792
feat: support cache for hunyuanvideo-1.5 (#528)
DefTruth Dec 5, 2025
6c21efc
feat: support cache for ovis-image (#530)
DefTruth Dec 5, 2025
6fc8489
fix: avoid nccl timeout when the service hangs (#531)
BBuf Dec 5, 2025
6d9f074
chore: remove un-needed cp imports (#532)
DefTruth Dec 5, 2025
8350be6
feat: pointer casting for fp8 all2all (#533)
triple-mu Dec 6, 2025
7a81c1e
chore: relax block adapter deps (#534)
DefTruth Dec 6, 2025
ce37923
Add request queue to limit concurrent generation requests (#535)
BBuf Dec 6, 2025
a14d4d8
News: 🔥🔥SGLang Diffusion x Cache-DiT ready!🔥🔥 (#536)
DefTruth Dec 7, 2025
fba8878
feat: optimize async fp8 ulysses attn (#537)
DefTruth Dec 8, 2025
dd83c0b
chore: delete useless code in serving (#539)
BBuf Dec 8, 2025
97648e1
feat: make all_to_all comm unified (#538)
DefTruth Dec 8, 2025
b583763
feat: add refresh cache context api (#542)
DefTruth Dec 9, 2025
6dc80a1
feat: support image edit model in serving (#541)
BBuf Dec 9, 2025
caf6a76
chore: simplify ulysses async flag (#543)
DefTruth Dec 9, 2025
5507473
chore: re-registered sage attn backend (#544)
DefTruth Dec 9, 2025
0846a8a
feat: support any head num for ulysses (#546)
DefTruth Dec 9, 2025
6ebf703
feat: support uneven heads in ulysses w/o padding (#547)
triple-mu Dec 10, 2025
9cdfc1a
chore: refactor FLUX.2 image editing tests (#548)
BBuf Dec 10, 2025
c6e51e3
feat: Add ulysses for any heads w/o padding (#549)
DefTruth Dec 10, 2025
da48aa3
feat: add envs manager for cache-dit (#550)
DefTruth Dec 10, 2025
d3e973f
fix ulysses fp8 and uneven head conflicts (#552)
DefTruth Dec 10, 2025
c9ac429
feat: UAA avoid extra memory IO access (#551)
triple-mu Dec 10, 2025
b0c10eb
chore: simplify quantize flags in example utils (#553)
DefTruth Dec 10, 2025
027d2bb
chore: fix quantize flags in example (#554)
DefTruth Dec 10, 2025
b69841a
chore: fix quantize & TP conflicts for wan (#556)
DefTruth Dec 10, 2025
a03bbfa
feat: support serving text2video (#555)
BBuf Dec 11, 2025
2d27b23
chore: Update SERVING Doc and FAQ Doc (#557)
BBuf Dec 11, 2025
21a62ea
chore: qwen edit lightning cp/tp examples (#559)
DefTruth Dec 11, 2025
4d0cfe4
feat: support ovis-image context parallel (#560)
DefTruth Dec 12, 2025
3e11fe9
feat: serving support image2video (#558)
BBuf Dec 12, 2025
21f2981
chore: add collect_env script (#562)
DefTruth Dec 12, 2025
7c4bc0e
Add pre-commit and GitHub Actions CI (#564)
DefTruth Dec 12, 2025
b1fdb6c
chore: refactor parallelism for better reusability (#565)
DefTruth Dec 15, 2025
3e6b6a5
chore: Update 🔥vLLM-Omni integration (#566)
SamitHuang Dec 15, 2025
92c24a3
feat: add pipe quant config for serving (#563)
nono-Sang Dec 15, 2025
9aac831
News: 🔥vLLM-Omni x Cache-DiT ready! (#567)
DefTruth Dec 15, 2025
df9d9f9
feat: enable custom attn backend for TP (#568)
DefTruth Dec 16, 2025
a821f80
feat: support TP for many text encoder (#569)
DefTruth Dec 16, 2025
6d77dec
fix qwen-edit-lightning examples (#571)
DefTruth Dec 16, 2025
2b5133c
fix get_text_encoder_from_pipe (#572)
DefTruth Dec 16, 2025
5f820d5
fix: add general compile options in example utils (#573)
DefTruth Dec 16, 2025
76b8940
chore: reduce un-popular examples (#574)
DefTruth Dec 16, 2025
2ee309e
feat: add text_encoder tp for serving (#570)
nono-Sang Dec 16, 2025
dd6edd5
chore: simplify example (#575)
DefTruth Dec 16, 2025
23f07f2
chore: make unified examples (#576)
DefTruth Dec 17, 2025
ae780ba
chore: fix vllm-omni docs link (#577)
DefTruth Dec 17, 2025
6dc1728
chore: optimize examples default path mapping (#579)
DefTruth Dec 18, 2025
490542f
chore: fix vllm-omni docs link (#580)
DefTruth Dec 18, 2025
a8c4945
feat: support Ovis-Image tensor parallel (#582)
DefTruth Dec 18, 2025
a37cbae
chore: fix typo in User_Guide.md (#583)
DefTruth Dec 18, 2025
de39534
chore: fail fast TP validation for attn heads (#581)
CPFLAME Dec 18, 2025
2c04959
fix patch functor for multi transformers (#586)
DefTruth Dec 19, 2025
cccd638
chore: add qwen image controlnet example (#588)
DefTruth Dec 19, 2025
339cafa
chore: update docs (#590)
DefTruth Dec 19, 2025
17c3a4f
feat: register fa3 backend for context parallel (#589)
nono-Sang Dec 19, 2025
1534d69
chore: support separate quant-type for text encoder (#591)
DefTruth Dec 19, 2025
2092bfe
hotfix for fa3 backend import error (#593)
DefTruth Dec 19, 2025
69e82bd
chore: fix typo in README.md (#594)
DefTruth Dec 20, 2025
7c9f450
chore: set save_ctx to False for inference (#596)
nono-Sang Dec 22, 2025
57a6f23
fix flux examples model path mismatch (#597)
DefTruth Dec 22, 2025
08253c7
Simplify CLI: Make task argument optional (#600)
BBuf Dec 22, 2025
b4c1c02
chore: fix extra path compare (#603)
DefTruth Dec 23, 2025
bab24b8
CI: Add build_wheel CI (#604)
DefTruth Dec 23, 2025
ec65b19
feat: support cache for LongCat-Image (#602)
e1ijah1 Dec 23, 2025
099d3ea
feat: Serving support LORA (#601)
BBuf Dec 23, 2025
27384b4
CI: Add Forward Pattern CPU CI Tests (#605)
DefTruth Dec 23, 2025
4b49748
chore: Update README.md (#606)
DefTruth Dec 23, 2025
f8981b5
Fix typo (#607)
BBuf Dec 23, 2025
fb5c737
feat: z-image-controlnet 🔥4x speedup! (#608)
DefTruth Dec 24, 2025
21e7ed6
fix lora path mismatch in examples (#609)
DefTruth Dec 24, 2025
9b4dac9
feat: support TP and CP for LongCat-Image (#610)
DefTruth Dec 24, 2025
5df6d80
misc: fix typo (#612)
DefTruth Dec 25, 2025
442ac7c
feat: support 🔥Qwen-Image-Edit-2511 (#614)
DefTruth Dec 25, 2025
d958b77
feat: support 🔥Qwen-Image-Layered (#615)
DefTruth Dec 25, 2025
ef5cb71
chore: simplify parallelism dispatch (#616)
DefTruth Dec 25, 2025
226ba52
chore: simplify quantize dispatch (#617)
DefTruth Dec 25, 2025
21936c6
chore: refactor kernels module (#618)
DefTruth Dec 25, 2025
15e62ed
ci: add refresh context ci tests (#619)
DefTruth Dec 26, 2025
8426e94
misc: add device info to example summary (#621)
DefTruth Dec 26, 2025
44ee074
feat: support ⚡️Z-Image-Turbo Nunchaku (#623)
DefTruth Dec 26, 2025
f912c36
[chore] Improve error_logging in serving tp_worker.py (#627)
BBuf Dec 26, 2025
8ac6ed8
chore: support more alias for quant types (#628)
DefTruth Dec 29, 2025
ca27b04
fix alias rev map for quant types (#629)
DefTruth Dec 29, 2025
c53c427
chore: lazy import check for quantize api (#630)
DefTruth Dec 30, 2025
9adc38e
chore: add more compile flags setting (#631)
DefTruth Dec 30, 2025
84a980d
[Bug] Apply --attn backend in single-GPU examples (#633)
BBuf Dec 31, 2025
e9b8d2f
Bump up to v1.1.10 (#634)
DefTruth Dec 31, 2025
e7f0430
[chore] serving support more quant (#635)
BBuf Jan 3, 2026
b40853d
feat: fused permute into ulysses qkv fp8 (#636)
triple-mu Jan 4, 2026
57da2a0
feat: support 🔥qwen-image-2512 (#637)
DefTruth Jan 4, 2026
c2e334c
[chore] Serving cient support return inference time (#638)
BBuf Jan 4, 2026
1bf6dec
fix wan 2.1 i2v cp and t5/umt5 te-p (#639)
DefTruth Jan 4, 2026
3c6a985
Serving: Align inference time format with LightX2V (#640)
BBuf Jan 4, 2026
4c53f10
[Test] Serving flux2_image_edit test add ulysses-anything test (#641)
BBuf Jan 5, 2026
8c331b8
chore: fix wan 2.1 t2v example path mapping (#642)
DefTruth Jan 5, 2026
5625c3a
feat: native attn support ring-attn (#643)
DefTruth Jan 5, 2026
09b70bb
feat: support 🔥vae parallelism (#645)
DefTruth Jan 6, 2026
17e5587
feat: support flux2 vae parallelism (#646)
DefTruth Jan 7, 2026
8480c05
fix wan vae parallel & format (#647)
DefTruth Jan 7, 2026
1e08777
update license (#648)
DefTruth Jan 7, 2026
32823f9
chore: fix video output path parse error (#649)
DefTruth Jan 7, 2026
cb14d5d
chore: fix load options utils (#650)
DefTruth Jan 7, 2026
47df7b9
feat: support ascend npu (#651)
gameofdimension Jan 7, 2026
a4371b3
chore: optimize qwen-image vae parallel (#652)
DefTruth Jan 7, 2026
d62b0aa
feat: add abstract platform (#653)
DefTruth Jan 7, 2026
3ee4547
chore: use unified platform for ci tests (#655)
DefTruth Jan 8, 2026
b4bfe63
chore: use unified platform for serving (#656)
DefTruth Jan 8, 2026
ac82f7d
[chore] Optional stats + path/base64 outputs (#658)
BBuf Jan 8, 2026
837ff50
feat: add online readthedocs.io (#659)
DefTruth Jan 8, 2026
9d6920a
hotfix for online readthedocs.io (#660)
DefTruth Jan 8, 2026
47d72ae
hotfix for online readthedocs.io (#661)
DefTruth Jan 8, 2026
2eab454
hotfix for online readthedocs.io (#662)
DefTruth Jan 8, 2026
3d9853e
chore: test online readthedocs.io ci (#663)
DefTruth Jan 8, 2026
0767200
feat: support ulysses npu attention (#654)
luren55 Jan 8, 2026
2713022
docs: update cache-dit docs (#664)
DefTruth Jan 8, 2026
ff544b7
docs: add example docs (#665)
DefTruth Jan 8, 2026
ece5890
docs: update docs (#666)
DefTruth Jan 8, 2026
b3f2d61
docs: fix typos (#667)
DefTruth Jan 8, 2026
9de66d1
fix typo
DefTruth Jan 8, 2026
b5dc90a
docs: optimize user docs (#668)
DefTruth Jan 9, 2026
0e12c7d
docs: add more image assets for docs (#670)
DefTruth Jan 9, 2026
a771043
docs: fix image rendering broken (#671)
DefTruth Jan 9, 2026
f3b78f3
feat: allow pass attn_mask for npu attention (#672)
DefTruth Jan 9, 2026
06c568e
[SERVING] Add FLUX.2 LoRA tests& support sigmas (#674)
BBuf Jan 9, 2026
98a4c12
feat: support load cache/parallel config from yaml (#675)
DefTruth Jan 9, 2026
c9e01b9
docs: add dbprune assets (#676)
DefTruth Jan 9, 2026
ccdd5a1
docs: fix dbprune image broken (#677)
DefTruth Jan 9, 2026
3703f77
docs: fix lemica link broken (#678)
DefTruth Jan 9, 2026
2003b53
[Doc] Profiling doc match new generate.py (#679)
BBuf Jan 12, 2026
c05baa9
CI: add check-mkdocs ci (#680)
DefTruth Jan 12, 2026
f51f19a
CI: add load configs ci tests (#681)
DefTruth Jan 12, 2026
45f12be
docs: refactor benchmark docs (#682)
DefTruth Jan 12, 2026
dbb3767
docs: update README.md (#683)
DefTruth Jan 12, 2026
2fb55aa
docs: update nvidia gpu benchmark (#684)
DefTruth Jan 12, 2026
98b12de
docs: move down serving deployment docs (#685)
DefTruth Jan 12, 2026
1aa2255
Add NPU support user guide (#687)
luren55 Jan 13, 2026
7a0cb71
chore: deprecated load_options in refresh_context (#689)
DefTruth Jan 13, 2026
052770d
feat: support ltx-2 (#691)
DefTruth Jan 13, 2026
df034d0
fix qwen-image cp error for latest diffusers (#692)
DefTruth Jan 13, 2026
e286533
chore: serving time support microsecond and uuid (#694)
BBuf Jan 14, 2026
e80cc8f
feat: improve stability by safe imports (#695)
DefTruth Jan 14, 2026
2059742
docs: add npu support matrix and bench (#693)
luren55 Jan 14, 2026
ef66037
docs: Add new model support guide (#696)
DefTruth Jan 14, 2026
2e80944
docs: update ascend npu docs (#697)
DefTruth Jan 14, 2026
ab4d082
chore: serving clean time record useless format (#698)
BBuf Jan 14, 2026
75d24eb
chore: change safe import logging to debug (#699)
DefTruth Jan 14, 2026
02a886b
npu docs update (#700)
luren55 Jan 14, 2026
56c235a
chore: update npu support matrix and user guide (#702)
DefTruth Jan 15, 2026
bb70174
fix README.md (#703)
DefTruth Jan 15, 2026
7a4363a
bugfix: some hotfix for examples (#705)
DefTruth Jan 15, 2026
e51dfca
docs: add attn backend to npu docs (#706)
DefTruth Jan 15, 2026
18152db
[Diffusion] Update h100 benchmark (#707)
BBuf Jan 15, 2026
f40e2ab
fix kwargs override error in load configs (#708)
DefTruth Jan 15, 2026
d808fff
CI: use torch-cpu for basic cpu tests (#709)
DefTruth Jan 15, 2026
cd60960
docs: update vllm-omni x cache-dit link (#710)
DefTruth Jan 15, 2026
c7398ea
chore: add flux fill examples (#711)
DefTruth Jan 15, 2026
f474da9
npu docs update (#712)
luren55 Jan 15, 2026
ce429ae
npu docs update (#713)
luren55 Jan 15, 2026
09bac0a
NPU support update (#715)
luren55 Jan 16, 2026
74d80b2
fix dim computation for FLUX.2 klein (#716)
gameofdimension Jan 16, 2026
128069f
feat: support 🚀flux2-klein series (#717)
DefTruth Jan 16, 2026
dfd4b2b
chore: use new logo (#718)
DefTruth Jan 16, 2026
be1bb83
fix logo link (#719)
DefTruth Jan 16, 2026
30ba424
chore: enable scm for examples (#720)
DefTruth Jan 16, 2026
8c047c3
chore: use new logo in mkdocs.yml (#721)
DefTruth Jan 16, 2026
853df89
docs: update installation user guide (#722)
DefTruth Jan 16, 2026
a75920d
[chore] Align torch generator with example (#723)
BBuf Jan 17, 2026
94b4e16
Fix generator bug in cache-dit (#724)
BBuf Jan 19, 2026
00a2264
examples: allow custom generator device (#726)
DefTruth Jan 19, 2026
95b32ff
examples: allow custom warmup-steps (#727)
DefTruth Jan 19, 2026
410cb87
docs: add latest news (#728)
DefTruth Jan 19, 2026
1585482
docs: fix docs format (#729)
DefTruth Jan 19, 2026
c4c0d82
fix selected metrics print (#730)
66RING Jan 19, 2026
328edea
docs: add flux examples to tp docs (#731)
DefTruth Jan 20, 2026
7c26b25
fix ltx-2 i2v example (#734)
DefTruth Jan 20, 2026
d130bc8
Update README.md (#735)
DefTruth Jan 20, 2026
a677008
chore: allow use default steps for scm (#736)
DefTruth Jan 21, 2026
4fd633b
chore: support gpu generator in server (#737)
BBuf Jan 21, 2026
38de92b
docs: update download badge (#738)
DefTruth Jan 21, 2026
e887a55
docs: refine profiler and serving docs (#739)
BBuf Jan 21, 2026
bb033b3
fix UAA broken while using joint attn
DefTruth Jan 21, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
67 changes: 67 additions & 0 deletions .github/workflows/build-wheel.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,67 @@
name: Build Wheel

on:
push:
branches:
- main
paths-ignore:
- 'docs/**'
- '.assets/**'
- '**.md'
- '.gitignore'
- '.gitattributes'
- 'LICENSE'
pull_request:
branches:
- main
paths-ignore:
- 'docs/**'
- '.assets/**'
- '**.md'
- '.gitignore'
- '.gitattributes'
- 'LICENSE'
workflow_dispatch:

jobs:
build:
runs-on: ubuntu-latest
strategy:
matrix:
python-version: ["3.12"]

steps:
- uses: actions/checkout@v4
with:
fetch-depth: 0
fetch-tags: true

- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
cache: 'pip'

- name: Install build
run: |
pip install -U pip && pip install build

- name: Check build configuration
run: |
if [ ! -f "pyproject.toml" ]; then
echo "Error: pyproject.toml not found! Cannot build wheel."
exit 1
fi
echo "✅ Build configuration file (pyproject.toml) found."

- name: Build Wheel
run: |
python -m build --wheel --outdir dist
ls -la dist/

- name: Upload Wheel Artifact
uses: actions/upload-artifact@v4
with:
name: cache-dist-wheel-py${{ matrix.python-version }}
path: dist/*.whl
retention-days: 15
70 changes: 70 additions & 0 deletions .github/workflows/check-mkdocs.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,70 @@
name: Check MkDocs Build

on:
push:
branches:
- main
paths-ignore:
- '.github/**'
- '.gitignore'
- '.gitattributes'
- 'LICENSE'
pull_request:
branches:
- main
paths-ignore:
- '.github/**'
- '.gitignore'
- '.gitattributes'
- 'LICENSE'
workflow_dispatch:

jobs:
build:
runs-on: ubuntu-latest
strategy:
matrix:
python-version: ["3.12"]

steps:
- uses: actions/checkout@v4
with:
fetch-depth: 0
fetch-tags: true

- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
cache: 'pip'

- name: Check build configuration
run: |
if [ ! -f "pyproject.toml" ]; then
echo "Error: pyproject.toml not found! Cannot build docs."
exit 1
fi
echo "✅ Build configuration file (pyproject.toml) found."

- name: Install Only MkDocs Dependencies
run: |
pip install mkdocs>=1.5.0 mkdocs-api-autonav mkdocs-material \
mkdocstrings-python mkdocs-gen-files mkdocs-awesome-nav \
mkdocs-glightbox mkdocs-git-revision-date-localized-plugin \
mkdocs-minify-plugin regex ruff pydantic

# Build MkDocs documentation with strict mode and fail it if the
# logs contains 'WARNING', 'aborted', 'ERROR', or not contains
# 'Documentation built'.
- name: Build MkDocs Documentation Strictly
run: |
mkdocs build --strict 2>&1 | tee build.log
if grep -E 'WARNING|aborted|abort|ERROR' build.log; then
echo "MkDocs build failed due to warnings or errors."
exit 1
elif ! grep -q 'Documentation built' build.log; then
echo "MkDocs build did not complete successfully."
exit 1
else
echo "MkDocs build completed successfully!"
fi
105 changes: 105 additions & 0 deletions .github/workflows/cpu-tests.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,105 @@
name: Run CPU Tests

on:
push:
branches:
- main
paths-ignore:
- 'docs/**'
- '.assets/**'
- '**.md'
- '.gitignore'
- '.gitattributes'
- 'LICENSE'
pull_request:
branches:
- main
paths-ignore:
- 'docs/**'
- '.assets/**'
- '**.md'
- '.gitignore'
- '.gitattributes'
- 'LICENSE'
workflow_run:
workflows: ["Build Wheel"]
branches: [main]
types: [completed]
workflow_dispatch:

jobs:
Basic_CPU_Tests:
if: >-
(github.event_name == 'workflow_run' && github.event.workflow_run.conclusion == 'success') ||
github.event_name == 'pull_request' ||
github.event_name == 'push' ||
github.event_name == 'workflow_dispatch'
runs-on: ubuntu-latest
strategy:
fail-fast: false
matrix:
python-version: ["3.12"]

steps:
- name: Checkout code (for test files)
uses: actions/checkout@v4
with:
fetch-depth: 0

- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
cache: 'pip'

- name: Download built wheel artifact (only for workflow_run)
if: github.event_name == 'workflow_run'
uses: dawidd6/action-download-artifact@v6
with:
workflow: build_wheel.yml
run_id: ${{ github.event.workflow_run.id }}
name: cache-dist-wheel-py${{ matrix.python-version }}
path: dist/

- name: Install dependencies
run: |
pip install -U pip && pip install torch==2.9.1 torchvision --index-url https://download.pytorch.org/whl/cpu
if [ -f "dist/*.whl" ]; then
pip install dist/*.whl
else
pip install -e .
fi
pip install pytest

- name: Run Forward Pattern CPU Tests (pytest)
run: |
cd tests
pytest api/test_forward_pattern.py -v -s -x

- name: Run Refresh Context CPU Tests (pytest)
run: |
cd tests
pytest api/test_refresh_context.py -v -s -x

- name: Run TaylorSeers CPU Tests (pytest)
run: |
cd tests
pytest api/test_taylorseers.py -v -s -x

- name: Run Load Configs CPU Tests (pytest)
run: |
cd tests
pytest api/test_load_configs.py -v -s -x

- name: Verify cache_dit installation
run: |
python -c "import cache_dit; print(f'✅ cache_dit imported successfully, version: {cache_dit.__version__}')"
continue-on-error: false

- name: Upload test logs (optional)
uses: actions/upload-artifact@v4
with:
name: test-logs-py${{ matrix.python-version }}
path: tests/basic-cpu-tests-results.txt
retention-days: 7
if: always()
26 changes: 26 additions & 0 deletions .github/workflows/pre-commit.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,26 @@
name: pre-commit

on:
pull_request:
branches: [main]
push:
branches: [main]
workflow_dispatch: # for manual trigger

concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: ${{ github.event_name == 'pull_request' }}

permissions:
contents: read

jobs:
pre-commit:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: actions/setup-python@v5
with:
python-version: "3.12"
cache: 'pip'
- uses: pre-commit/action@v3.0.1
2 changes: 1 addition & 1 deletion .pre-commit-config.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@ repos:
- id: check-docstring-first
- id: check-toml
- id: check-yaml
exclude: packaging/.*
exclude: ^(packaging/.*|mkdocs\.yml)$
args:
- --allow-multiple-documents
- id: mixed-line-ending
Expand Down
24 changes: 24 additions & 0 deletions .readthedocs.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,24 @@
# Read the Docs configuration file
# See https://docs.readthedocs.io/en/stable/config-file/v2.html for details

version: 2

build:
os: ubuntu-22.04
tools:
python: "3.12"
jobs:
post_checkout:
- git fetch --unshallow || true

mkdocs:
configuration: mkdocs.yml
fail_on_warning: true

# Optionally declare the Python requirements required to build your docs
python:
install:
- method: pip
path: .
extra_requirements:
- docs
2 changes: 1 addition & 1 deletion LICENSE
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved
Copyright (c) 2025 Cache-DiT Authors. All Rights Reserved

Apache License
Version 2.0, January 2004
Expand Down
Loading
Loading