Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
377 commits
Select commit Hold shift + click to select a range
81abfb2
chore: rename and reformat gits_noise.inl (#1617)
wbruna Jun 7, 2026
2a07540
refactor: move photomaker into generation extension (#1618)
leejet Jun 7, 2026
b3d56d0
refactor: split model loader from model definitions (#1619)
leejet Jun 7, 2026
17a2b4a
perf: cap planner budget when model dwarfs the streaming budget (#1612)
fszontagh Jun 8, 2026
138da14
apg: normalize diff_norm calculation by tensor size (#1620)
stduhpf Jun 8, 2026
19bdfe2
feat: set tensor names on block params (#1622)
RapidMark Jun 8, 2026
1b702a5
fix: correct mask shape for masked flash attention (#1625)
RapidMark Jun 13, 2026
c20769b
feat: add circular RoPE support for ideogram4 (#1627)
stduhpf Jun 13, 2026
1fb6b22
feat: add free_sd_images function to manage memory for C API (#1633)
Cyberhan123 Jun 13, 2026
1365008
chore: add script for automatic code formatting (#1636)
Cyberhan123 Jun 13, 2026
3a54597
fix: SD3 conditioning crash when clip_l text encoder is missing (#1638)
Detensable Jun 13, 2026
563137a
refactor: centralize runner weight staging and cleanup (#1644)
leejet Jun 13, 2026
9b0fceb
refactor: manage upscaler params through model manager (#1645)
leejet Jun 13, 2026
8d4c7af
refactor: route all runner params through model manager (#1649)
leejet Jun 13, 2026
276025e
fix: mark LoKR w2_a tensor as applied (#1650)
leejet Jun 13, 2026
bdb431a
feat: support disk params backend (#1651)
leejet Jun 14, 2026
749186c
refactor: remove vae_decode_only context flag (#1653)
leejet Jun 14, 2026
5db680c
refactor: route cpu placement through backend specs (#1654)
leejet Jun 14, 2026
17d70b9
docs: replace example option lists with help commands
leejet Jun 14, 2026
9838264
refactor: simplify ControlNet output caching (#1655)
leejet Jun 14, 2026
c2df4e1
feat: add RPC support (#1629)
stduhpf Jun 14, 2026
6f00939
docs: refresh README guide links
leejet Jun 14, 2026
517abc7
sync: update ggml (#1656)
leejet Jun 14, 2026
bb90bfa
feat: support backend-specific max-vram budgets
leejet Jun 14, 2026
6e66a1a
fix: allow oversized Vulkan parameter tensors (#1662)
leejet Jun 15, 2026
93527fd
feat: add PuLID-Flux identity-injection support (#1595)
RapidMark Jun 15, 2026
146b6cc
fix: simplify PuLID ID extraction setup (#1664)
leejet Jun 15, 2026
5a34bc7
feat: support for cancelling generations (#1124)
wbruna Jun 15, 2026
710bc91
fix: correct conversion from sd_type_t to ggml_type (#1519)
wbruna Jun 16, 2026
92a3b73
sync: update sdcpp-webui (#1668)
leejet Jun 16, 2026
7f0e728
fix: normalize CLIP prompts before special-token splitting (#1670)
leejet Jun 16, 2026
e8e012e
fix: workaround for Anima with Vulkan and Flash Attention (#1678)
wbruna Jun 21, 2026
e9e9524
fix: workaround for Ernie with Vulkan and Flash Attention (#1680)
daniandtheweb Jun 21, 2026
2bd249c
feat: concatenate repeated cli arg strings (#1686)
stduhpf Jun 21, 2026
b12098f
feat: add boogu image support (#1688)
leejet Jun 21, 2026
787d229
perf: --eager-load to pre-load params at model-load time (#1687)
fszontagh Jun 22, 2026
854bebf
feat: add --prompt-file and --negative-prompt-file flags (#1693)
grencez Jun 22, 2026
b395a69
refactor: add Flux VAE version helper (#1696)
leejet Jun 22, 2026
41f7acb
feat: support guidance_schedule (#1684)
stduhpf Jun 22, 2026
f440ad9
fix: avoid writable mmap for read-only weights (#1698)
leejet Jun 22, 2026
2938272
feat: add logit-normal scheduler (#1669)
stduhpf Jun 24, 2026
8caa3f9
feat: add krea2 support (#1705)
leejet Jun 24, 2026
39f7962
ci: adopt dynamic cpu backends on released binaries (#1704)
wbruna Jun 26, 2026
9ee77fc
fix(ci): disable dynamic CPU backends for arm64 CUDA image (#1709)
leejet Jun 26, 2026
3973015
sync: update ggml and revert vulkan workarounds for Anima and Ernie (…
daniandtheweb Jun 26, 2026
ec4cb81
fix: correct TAEHV encoding for image models (#1711)
wbruna Jun 26, 2026
9956436
refactor: consolidate WAN VAE version checks (#1712)
leejet Jun 26, 2026
f54e45e
fix: correct sycl ci (#1716)
leejet Jun 28, 2026
03e9a22
feat: add SeFi-Image support (#1707)
fszontagh Jun 28, 2026
d77b8f5
feat: support Qwen-Image/Wan VAE with diffusers naming (#1713)
stduhpf Jun 28, 2026
7b5f34d
feat: support Qwen2D VAE (#1714)
stduhpf Jun 28, 2026
9f855c9
chore: silence narrowing conversion warnings (#1717)
leejet Jun 28, 2026
c179075
feat: enhanced third-party integrations (#1632)
Cyberhan123 Jun 28, 2026
0484600
fix: add zip library back to cli target link libraries (#1719)
stduhpf Jun 29, 2026
3ec374a
fix: avoid crash and warn when using Qwen 2D VAE for Wan video (#1721)
stduhpf Jun 29, 2026
61a637b
feat: add Flux2 scheduler (#1722)
leejet Jun 29, 2026
57e19fa
feat: add Flux scheduler (#1723)
leejet Jun 29, 2026
3b6c9ca
feat: add normal alias for discrete scheduler (#1724)
leejet Jun 29, 2026
f027107
chore: strip UTF-8 BOMs and add cleanup script (#1726)
leejet Jun 30, 2026
ccda89e
feat: make ffi for same shape (#1635)
Cyberhan123 Jun 30, 2026
2bb0389
refactor: return bool from image and upscale APIs (#1728)
leejet Jun 30, 2026
484baa4
feat: add beta scheduler (#811)
Green-Sky Jun 30, 2026
1a13107
feat: add imatrix support (#633)
stduhpf Jul 1, 2026
3590aa8
feat: add MiniT2I support (#1683)
KenForever1 Jul 1, 2026
556f04b
feat: add qwen image layered support (#1119)
leejet Jul 2, 2026
7dab366
fix: enable Wan/TAEHV video generation on the Metal backend (#850) (#…
mmandelker-code Jul 2, 2026
2574f59
fix: fallback when backend rejects flash attention (#1732)
leejet Jul 2, 2026
7bcd189
feat: add multi-device layer split (--backend "diffusion=cuda0&cuda1"…
pwilkin Jul 4, 2026
68f3d6d
feat: support for cross-device row split (#1735)
pwilkin Jul 4, 2026
b11c95a
feat: auto fit tensors across devices to guarantee optimal load (#1736)
pwilkin Jul 4, 2026
e790073
fix: avoid layer splitting unet block paths (#1741)
leejet Jul 5, 2026
45714b1
feat(sdapi): report generation parameters through the info field (#1426)
wbruna Jul 5, 2026
38a51f8
feat: add DPM++ 2M SDE sampler (#1742)
fszontagh Jul 5, 2026
c60b36a
feat: denoise strength as starting noise level (#1738)
stduhpf Jul 5, 2026
e9dee54
feat: add DPM++ 2M SDE (Brownian tree) sampler (#1743)
fszontagh Jul 5, 2026
da6db07
feat: stream model conversion (#1581)
shikaku2 Jul 5, 2026
2abbc77
fix: use larger image VAE encode tiles (#1744)
leejet Jul 5, 2026
0321ce1
docs: unify model path
leejet Jul 5, 2026
c674225
chore: move Dockerfiles into docker directory (#1745)
leejet Jul 5, 2026
dff0e88
chore: move utility scripts under scripts (#1746)
leejet Jul 5, 2026
e071aa3
feat: move circular padding from context to per-generation params (#1…
fszontagh Jul 6, 2026
8b135b5
docs: explain CPU streaming combo (--offload-to-cpu, --max-vram, --st…
fszontagh Jul 6, 2026
9e1055d
fix: SDXL ControlNet (diffusers naming + graph size) (#1752)
fszontagh Jul 6, 2026
4fcc6fe
fix: reject a repeated entry with an inconsistent value count in load…
professor-moody Jul 6, 2026
e22272e
fix: validate safetensors data offsets (#1754)
leejet Jul 6, 2026
bb84971
refactor: move model-specific args into model parsers (#1757)
leejet Jul 6, 2026
9ef6e73
feat: drive layer split from graph-cut segments (#1762)
leejet Jul 7, 2026
885f01a
chore: close inactive issues automatically
leejet Jul 7, 2026
6314af4
docs: add shared agent instructions
leejet Jul 7, 2026
cc73429
chore: close inactive issues as completed
leejet Jul 8, 2026
12b6fbf
feat: hot-reload ControlNet - swap without rebuilding the context (#1…
fszontagh Jul 10, 2026
9beb6ac
fix: avoid f16 overflow in Z-Image quantized matmuls on ROCm (#1771)
pwilkin Jul 10, 2026
ead6bf5
fix: extend f32 matmul precision to ROCm for Qwen-Image, Krea2 and Bo…
pwilkin Jul 10, 2026
1b04283
feat: support safetensors index loading (#1769)
leejet Jul 10, 2026
c79d24b
feat: add Krea2OstrisEdit support (#1775)
stduhpf Jul 11, 2026
b5d8120
feat: add lingbot video support (#1770)
leejet Jul 11, 2026
74bce04
feat: add configurable reference image processing for edit models (#1…
stduhpf Jul 14, 2026
833369d
fix: protect cross_attn and output_proj tokens for Anima LoRAs (#1786)
stduhpf Jul 14, 2026
c00a9e9
feat: AnimateDiff SD 1.5 motion modules (v2 + v3) (#1784)
fszontagh Jul 14, 2026
a8a91b2
feat: add ADetailer support (#1785)
leejet Jul 14, 2026
fafe8e6
docs: remove star history
leejet Jul 16, 2026
7717e82
feat(animatediff): support img2video via --init-img (#1789)
fszontagh Jul 16, 2026
b290693
feat: add PiD 1.5 support (#1790)
leejet Jul 16, 2026
ea4e566
feat: add hunyuan video 1.5 support (#1795)
leejet Jul 18, 2026
2961182
chore: add missing override declarations (#1800)
wbruna Jul 21, 2026
5e4e03c
docs: fix links to sd 1.5 and sd 2.1 (#1798)
Project516 Jul 21, 2026
cfd4cff
fix: Dockerfile.vulkan add missing libraries for nvidia support (#1805)
somewhatfrog Jul 22, 2026
35fb21f
fix: avoid structured binding capture in Hunyuan config (#1809)
leejet Jul 22, 2026
8a51eb9
feat: add Mage-Flow support (#1808)
leejet Jul 22, 2026
5114672
fix: detect vision patch size for unsplit (HF-format) Qwen3-VL (#1811)
fszontagh Jul 23, 2026
8d37707
feat: add IP-Adapter support for SD 1.5 and SDXL (#1803)
fszontagh Jul 24, 2026
b8bf676
fix: correct dangling pointer to empty image reference vector (#1813)
wbruna Jul 24, 2026
b338b4b
docs: add Gimp plugins to UIs section (#1799)
themanyone Jul 24, 2026
78124b6
ci: update ROCm releases to 7.14.0 (#1802)
superm1 Jul 24, 2026
b0f8568
fix: correct IP-Adapter CFG conditioning and defaults (#1815)
leejet Jul 24, 2026
87a0177
fix: add frame dimension for Hunyuan IMG2VID encoding (#1816)
leejet Jul 24, 2026
2d0385b
fix: add missing sampler names (#1819)
vmobilis Jul 26, 2026
5ef4a75
feat: expose IP-Adapter in server request schema and capabilities (#1…
fszontagh Jul 27, 2026
2251699
fix: skip incompatible LoRA weights (#1825)
leejet Jul 27, 2026
53856e7
fix: null pointer dereference when loading malformed LoHa file (#1826)
nmouha Jul 29, 2026
2993b7f
fix: make parameter loading backend-aware (#1828)
happyyzy Jul 29, 2026
9cfe2af
feat: display number of tokens for SD models (#1831)
vmobilis Jul 29, 2026
e92e86f
fix: prevent torch checkpoint offset overflow (#1832)
leejet Jul 29, 2026
af92790
feat: allow customizing the alpha and beta parameters of the beta sch…
wbruna Jul 30, 2026
735a4ef
fix(PhotoMaker): avoid GGML_ASSERT if trigger word 'img' was not foun…
akleine Jul 30, 2026
e31a86c
refactor: centralize CLIP prefix conversion (#1837)
leejet Jul 30, 2026
10378f4
fix: lora with split qkv compatibility check at runtime (#1836)
stduhpf Aug 2, 2026
8457624
feat: support more LoRA models (Kroma-v0.1 support) (#1842)
stduhpf Aug 2, 2026
50062a4
feat: add IP-Adapter Plus (Resampler image projection) support (#1839)
fszontagh Aug 2, 2026
eb7f35c
feat: add linear multi-step sampling method (#1843)
vmobilis Aug 2, 2026
db99efd
refactor: extract model loader initialization (#1844)
leejet Aug 2, 2026
b4e67d1
fix(cmake): only apply /MP to the MSVC compiler, not icx (#1846)
Stanley5249 Aug 4, 2026
ea7f0c8
feat: add minimax-h3 support (#1854)
leejet Aug 4, 2026
bfbef5b
feat: trained Minimax VAE Latent2rgb proj (#1856)
stduhpf Aug 5, 2026
c6beeef
fix: map Qwen3-VL DeepStack GGUF tensor names (#1858)
leejet Aug 5, 2026
b4f1fd6
fix: preserve "token_refiner" token for MiniMax H3 LoRAs (#1864)
stduhpf Aug 11, 2026
487de75
fix: fail with a message when MiniMax-H3 is run in img_gen mode (#1863)
danielhanchen Aug 11, 2026
bcc7e29
feat: support INT8 ConvRot safetensors (#1857)
leejet Aug 11, 2026
06c359f
fix: replace free_compute_buffer with runner_done in vae (#1872)
LostRuins Aug 12, 2026
fabe481
sync: update ggml (#1873)
leejet Aug 12, 2026
de298c2
fix(ci): trigger builds for ggml updates
leejet Aug 12, 2026
6100d83
feat: add taeh3 support (#1874)
stduhpf Aug 19, 2026
58b6cb6
fix: prevent gallocr hash overflow in tiny graph-cut segments (#1880)
fszontagh Aug 19, 2026
1706b32
fix: re-clamp streaming VRAM budget to currently free memory (#1878)
fszontagh Aug 19, 2026
88b044b
fix: mark graph cuts with both a prefix and a suffix (#1883)
wbruna Aug 19, 2026
760717a
fix: make max_order of lms sampler configurable (#1885)
vmobilis Aug 19, 2026
16304cc
fix: guard against missing sampler/scheduler names (#1887)
vmobilis Aug 19, 2026
97d2990
chore: format code
leejet Aug 19, 2026
12ee60d
fix: use sd_get_preview_interval() (#1907)
vmobilis Aug 25, 2026
0a565f2
feat: configurable image / video compression (#1909)
vmobilis Aug 25, 2026
50d6405
feat: support standard Qwen3-VL weights for MiniMax-H3 (#1910)
leejet Aug 25, 2026
be0e344
feat: load scaled FP8 weights without upfront conversion (#1913)
leejet Aug 27, 2026
2c92949
fix: match exact weights in LLM config detection (#1923)
leejet Aug 30, 2026
afd5306
feat: add LTX-2.5 support (#1893)
pwilkin Aug 30, 2026
c797899
fix: correct MiniMax H3 reference audio encoding (#1886)
jk212h20 Aug 30, 2026
dc4000d
fix: correct MiniMax H3 audio Euler steps (#1908)
jk212h20 Aug 30, 2026
2540a4f
feat: use backend-native FP8 matmul when supported (#1916)
leejet Aug 30, 2026
134c821
sync: update ggml
leejet Aug 30, 2026
d9b6e27
feat: additional `--preview-interval` values (#1915)
vmobilis Aug 30, 2026
9029655
feat: support numbering for preview images (#1895)
vmobilis Aug 30, 2026
40e605f
fix: use carrier sampling for MiniMax H3 audio (#1924)
leejet Aug 30, 2026
6b3edaa
feat: generalize temporal tiling across video VAEs (#1926)
leejet Aug 30, 2026
6c57cc3
feat: prefetch streamed layers during compute (#1905)
assouan Sep 6, 2026
462d675
refactor: unify runner lifecycles and weight residency (#1940)
leejet Sep 6, 2026
dbb6112
feat: add verbose logging and log-level selection (#1941)
leejet Sep 6, 2026
80bac2d
feat: enable single-GPU auto-fit with tiered parameter placement (#1942)
leejet Sep 6, 2026
d8fb10c
fix: reuse graph cut plans across CFG passes (#1943)
leejet Sep 6, 2026
31ab2b2
refactor: split ggml extensions and move implementations to cpp files…
leejet Sep 7, 2026
9cdb6b6
fix: preserve K-quantized embedding weights (#1936)
zjn20030811 Sep 7, 2026
d04e895
fix: correct SDXL embeddings loading (#1939)
wbruna Sep 7, 2026
6b47fec
refactor: unify model source and weight lifecycle management (#1956)
leejet Sep 10, 2026
469fc49
docs: reflect GGML_MAX_NAME value change in rpc docs (and in ggml_ext…
stduhpf Sep 10, 2026
14eddb3
refactor: split generation pipeline out of stable-diffusion.cpp (#1957)
leejet Sep 10, 2026
b68d586
fix: enable VAE decode tiling fallback without auto-fit (#1932)
Hmission Sep 10, 2026
e95ab96
fix: preserve BF16 embedding weights for get_rows (#1959)
xledx Sep 11, 2026
3191b23
fix: handle invalid option numbers (#1961)
vmobilis Sep 11, 2026
e06b205
feat: expose the loaded model version name through the public API (#1…
fszontagh Sep 11, 2026
7f986a9
feat: add SenseNova U1.5 support (#1935)
Maphist0 Sep 11, 2026
5ebce93
fix: reuse graph plans when scale parameters change (#1963)
leejet Sep 11, 2026
7f410a3
feat: add linear and attention scale overrides (#1964)
leejet Sep 11, 2026
44dd137
feat: preserve explicit backend assignments during auto-fit (#1967)
leejet Sep 13, 2026
9a97738
fix: guard GPU memory capacity and propagate encoding failures (#1958)
leejet Sep 13, 2026
4a7da26
fix: bound plain-text runs in parse_prompt_attention regex (#1919)
fszontagh Sep 13, 2026
0bd72f0
feat: add Wan2.2 S2V (audio+img-to-video) support (#1925)
noctrex Sep 13, 2026
ca37fad
fix: validate vision projector output dim against LLM hidden size (#1…
fszontagh Sep 13, 2026
5a5400b
fix: resolve MSVC narrowing conversion warnings (#1969)
leejet Sep 13, 2026
42d6c0a
feat: Add generation parameters into video metadata (#1901)
CAHbKA-IV Sep 13, 2026
4964abd
feat: support external Hugging Face tokenizer JSON files (#1973)
leejet Sep 14, 2026
f9ddc0f
refactor: require external Gemma 2 and GPT-OSS tokenizers (#1974)
leejet Sep 14, 2026
07a85c7
feat: support Brownian tree noise in all noise injection samplers (#1…
wbruna Sep 14, 2026
59c23bc
fix: use tokenizer-specific pre-tokenization rules (#1975)
leejet Sep 14, 2026
3161505
fix: remove vision_model. from ununsed tensors (#1983)
leejet Sep 16, 2026
cc515a0
perf: eliminate temporary allocations in Philox rounds (#1982)
leejet Sep 16, 2026
269e726
fix: honor flash attention flag in LLM text encoder attention (#1987)
linxuhao Sep 18, 2026
656a135
refactor: remove obsolete unused tensor filtering (#1984)
leejet Sep 18, 2026
adcac69
perf: pad small attention heads to 64 for MMA Flash Attention (#1992)
leejet Sep 18, 2026
2ea8aff
perf: update ggml for faster direct convolutions (#1993)
leejet Sep 18, 2026
3e037a8
perf: accelerate VAE direct 3D convolutions (#1996)
leejet Sep 19, 2026
9982c9c
fix: propagate CUDA driver dependency to shared library consumers
leejet Sep 19, 2026
d32b4e8
fix: prevent clip_preprocess center crop from exceeding the resized i…
akhenakh Sep 19, 2026
275ab58
perf: reduce CPU overhead in graph execution and sampling (#1997)
leejet Sep 19, 2026
17860c0
perf: parallelize host tensor elementwise and broadcast ops (#1998)
leejet Sep 19, 2026
1330ceb
feat: support building with upstream ggml (#1999)
leejet Sep 19, 2026
137f740
feat: add Qwen Image 2.1 support (#1994)
leejet Sep 20, 2026
008ca5b
feat: restore legacy fp8 handling when building with upstream ggml (#…
wbruna Sep 20, 2026
b8248a8
fix: avoid passing ggml logs as format strings (#2002)
wbruna Sep 20, 2026
15f335d
feat: add LLaDA-Image support (#1968)
fszontagh Sep 20, 2026
187b256
feat: add native CUDA SageAttention support (#2005)
leejet Sep 20, 2026
b56c686
fix: avoid narrowing conversion in SigVQ patch embedding and format code
leejet Sep 20, 2026
c678dfe
docs: update CONTRIBUTING.md
leejet Sep 20, 2026
74988b2
fix: reject video models in image generation (#2017)
leejet Sep 21, 2026
2726dd3
fix: update ggml to prevent permute metadata truncation (#2018)
leejet Sep 21, 2026
78557f8
fix: honor reference image resize opt-out in server requests (#2011)
AzizMuminov Sep 21, 2026
97d932b
fix: restrict VAE tiling retries to allocation failures (#2019)
leejet Sep 21, 2026
6dcb5bb
fix: handle GPU memory reports and LLM encoding failures (#2020)
leejet Sep 21, 2026
e012065
fix: preserve reference image dimensions in server requests (#2007)
mikemikimike Sep 22, 2026
e112ab5
fix: add alpha channel input for Qwen Image 2.1 and relative docs (#2…
CarlGao4 Sep 22, 2026
ac45422
fix: honor reference image resize settings in OpenAI edits (#2025)
leejet Sep 22, 2026
2bb7294
perf: cache MiniMax H3 text conditioning (#1966)
xledx Sep 22, 2026
28b454b
feat: add configurable image input preprocessing (#2028)
leejet Sep 22, 2026
c92d73c
fix: preserve alpha when upscaling RGBA images with ESRGAN (#2029)
leejet Sep 22, 2026
241518b
feat: add latent2rgba preview for Qwen-Image 2.1 (#2032)
stduhpf Sep 23, 2026
e6281b6
feat: add configurable conditioning cache for all models (#2034)
leejet Sep 23, 2026
2dc7f54
feat: add Qwen Image 2.1 prefix KV cache (#2035)
leejet Sep 23, 2026
3674693
fix: add graph cuts for MiniMax-H3 text conditioning (#1900)
assouan Sep 23, 2026
70c1dbc
perf: run one-frame Wan VAE convolutions as 2D convolutions (#2038)
nanguoyu Sep 23, 2026
2a4ebba
ci: automatically close PRs from organization-owned forks
leejet Sep 23, 2026
500ef5f
fix: map mmapped weights through Metal buffers instead of CPU buffers…
nanguoyu Sep 23, 2026
88411ef
refactor: centralize circular RoPE and extend image model support (#2…
leejet Sep 23, 2026
caa111a
feat: optimize cfg special cases with guidance schdeule (#2033)
stduhpf Sep 24, 2026
4dfe8f5
feat: add a stand-alone upscale endpoint to the server (#2026)
nbeerbower Sep 24, 2026
740c7ae
feat: add configurable Qwen cache types and early cache scheduling (#…
leejet Sep 24, 2026
1a2330d
fix: reserve 128 MiB headroom when selecting monolithic execution (#2…
leejet Sep 24, 2026
b167b94
fix: align Qwen Image 2.1 flow schedule with official defaults (#2048)
leejet Sep 24, 2026
0a9340c
fix: scale Qwen Image 2.1 VAE convolutions (#2054)
leejet Sep 25, 2026
5e7b291
fix: keep the server frontend install out of a parent pnpm workspace …
Yi-111-a Sep 25, 2026
510bccf
fix: map Qwen Image 2.1 LoRAs to fused MLP weights (#2057)
leejet Sep 25, 2026
4c3cf75
fix: load safetensors index shards without recursion (#2058)
leejet Sep 25, 2026
39ada08
feat: add PixArt model family support (#2047)
losewayy Sep 25, 2026
19bbbca
refactor: define VAE tile dimensions in image pixels (#2059)
leejet Sep 25, 2026
2f88688
refactor: align PixArt weights with upstream layout (#2061)
leejet Sep 25, 2026
168f7b8
fix: improve GGUF metadata parsing and reader selection (#2062)
losewayy Sep 27, 2026
9947eeb
fix: handle filesystem errors during lora and upscaler cache scan (#2…
Yi-111-a Sep 27, 2026
61a83b8
docs: update hugging face links for MageFlow diffusion and VAE models…
akleine Sep 27, 2026
f8890b9
feat: add Ming-Image Design support (#2063)
leejet Sep 27, 2026
47e83d7
sync: update ggml
leejet Sep 27, 2026
ede32a6
fix: load INT8 convrot LLM embeddings correctly (#2067)
leejet Sep 27, 2026
27f7c43
fix: update ggml to fix INT8 convrot backend fallback (#2069)
leejet Sep 27, 2026
42ab1c1
fix: keep scaled INT8 convrot matmuls on Vulkan (#2070)
leejet Sep 27, 2026
3f8527a
feat: enable hip INT8 tensorwise matmul and convrot (#2071)
leejet Sep 27, 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
  •  
  •  
  •  
4 changes: 3 additions & 1 deletion .github/ISSUE_TEMPLATE/bug_report.yml
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,9 @@ body:
- type: markdown
attributes:
value: |
Please use this template and include as many details as possible to help us reproduce and fix the issue.
Before submitting a bug report, please read the [Troubleshooting guide](https://github.com/leejet/stable-diffusion.cpp/blob/master/docs/troubleshooting.md) and try the steps relevant to your problem.

If the problem persists, complete this form and include what you tried and the results, along with enough details to help us reproduce and fix the issue.
- type: textarea
id: commit
attributes:
Expand Down
4 changes: 4 additions & 0 deletions .github/ISSUE_TEMPLATE/config.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,4 @@
contact_links:
- name: Troubleshooting
url: https://github.com/leejet/stable-diffusion.cpp/blob/master/docs/troubleshooting.md
about: Read the troubleshooting guide first. If the problem persists, submit a bug report.
15 changes: 15 additions & 0 deletions .github/pull_request_template.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
## Summary

<!-- Describe what changed and why. Keep the PR focused on one clear change. -->

## Related Issue / Discussion

<!-- Link related issues, discussions, or previous PRs if applicable. -->

## Additional Information

<!-- Add verification notes, screenshots, sample output, or other context when applicable. -->

## Checklist

- [ ] I have read and confirmed this PR follows the [contribution guidelines](https://github.com/leejet/stable-diffusion.cpp/blob/master/CONTRIBUTING.md).
331 changes: 179 additions & 152 deletions .github/workflows/build.yml

Large diffs are not rendered by default.

48 changes: 48 additions & 0 deletions .github/workflows/close-inactive-issues.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,48 @@
name: Close inactive issues

on:
schedule:
# Run daily. GitHub cron schedules use UTC.
- cron: "30 1 * * *"
workflow_dispatch:
inputs:
debug_only:
description: "Dry run: log intended actions without changing issues"
required: false
default: false
type: boolean

permissions:
issues: write

concurrency:
group: ${{ github.workflow }}
cancel-in-progress: false

jobs:
close-inactive-issues:
runs-on: ubuntu-latest
steps:
- name: Comment and close inactive issues
uses: actions/stale@v10
with:
days-before-issue-stale: 365
days-before-issue-close: 0

days-before-pr-stale: -1
days-before-pr-close: -1

stale-issue-label: issue:inactive
close-issue-label: issue:auto-closed
close-issue-reason: completed
stale-issue-message: ""
close-issue-message: >
This issue has had no activity for one year. The latest version of
the code may already have fixed the problem.

If the issue still exists in the latest version, you can reopen
this issue at any time with updated reproduction details.

remove-issue-stale-when-updated: true
operations-per-run: 1000
debug-only: ${{ github.event_name == 'workflow_dispatch' && inputs.debug_only || false }}
61 changes: 61 additions & 0 deletions .github/workflows/close-organization-fork-prs.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,61 @@
name: Close PRs from organization forks

on:
pull_request_target:
types: [opened, reopened]

permissions:
pull-requests: write

concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number }}
cancel-in-progress: false

jobs:
close-organization-fork-pr:
if: >-
github.event.pull_request.head.repo.owner.type == 'Organization' &&
github.event.pull_request.head.repo.id != github.event.pull_request.base.repo.id
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- name: Explain the contribution policy and close the PR
uses: actions/github-script@v9
with:
script: |
const { data: pr } = await github.rest.pulls.get({
...context.repo,
pull_number: context.issue.number,
});
const headRepo = pr.head.repo;
if (pr.state !== 'open' || !headRepo ||
headRepo.id === pr.base.repo.id || headRepo.owner.type !== 'Organization') {
return;
}

const marker = '<!-- organization-fork-policy -->';
const comments = await github.paginate(github.rest.issues.listComments, {
...context.repo,
issue_number: pr.number,
per_page: 100,
});
const alreadyExplained = comments.some(comment =>
comment.user?.login === 'github-actions[bot]' && comment.body?.includes(marker));
if (!alreadyExplained) {
await github.rest.issues.createComment({
...context.repo,
issue_number: pr.number,
body: [
marker,
'This repository requires contributions from forks to use a personal fork with **Allow edits from maintainers** enabled.',
'GitHub does not support this option for organization-owned forks, so this PR is being closed automatically.',
'Please open a new PR from a fork in your personal GitHub account and enable **Allow edits from maintainers** so maintainers can help update the branch.',
'See [the GitHub documentation](https://docs.github.com/en/pull-requests/how-tos/work-with-forks/allowing-changes-to-a-pull-request-branch-created-from-a-fork).',
].join('\n\n'),
});
}
await github.rest.pulls.update({
...context.repo,
pull_number: pr.number,
state: 'closed',
});
55 changes: 55 additions & 0 deletions .github/workflows/stale-prs.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,55 @@
name: Close inactive PRs

on:
schedule:
# Run daily. GitHub cron schedules use UTC.
- cron: "30 1 * * *"
workflow_dispatch:
inputs:
debug_only:
description: "Dry run: log intended actions without changing PRs"
required: false
default: false
type: boolean

permissions:
issues: write
pull-requests: write

concurrency:
group: ${{ github.workflow }}
cancel-in-progress: false

jobs:
stale-prs:
runs-on: ubuntu-latest
steps:
- name: Mark and close inactive PRs
uses: actions/stale@v10
with:
days-before-issue-stale: -1
days-before-issue-close: -1

days-before-pr-stale: 365
days-before-pr-close: 7

stale-pr-label: pr:inactive
close-pr-label: pr:auto-closed
exempt-pr-labels: pr:keep-open

stale-pr-message: >
This PR has been inactive for 365 days. If there is no new activity
within 7 days, it will be closed automatically. Comment, push new
commits, or remove the pr:inactive label to keep it open. Add
pr:keep-open to exempt it from future inactive PR cleanup.

close-pr-message: >
Closing this PR because it has had no activity for 7 days after
being marked inactive. If this is still useful or ready to move
forward, feel free to reopen it with fresh context or updated
details. Sorry for any inconvenience.

remove-pr-stale-when-updated: true
delete-branch: false
operations-per-run: 100
debug-only: ${{ github.event_name == 'workflow_dispatch' && inputs.debug_only || false }}
5 changes: 5 additions & 0 deletions .gitignore
Original file line number Diff line number Diff line change
@@ -1,6 +1,7 @@
build*/
cmake-build-*/
test/
tests/
.vscode/
.idea/
.cache/
Expand All @@ -13,3 +14,7 @@ output*.png
models*
*.log
preview.png
.claude/
CLAUDE.local.md
.agents/
.codex/
7 changes: 5 additions & 2 deletions .gitmodules
Original file line number Diff line number Diff line change
@@ -1,9 +1,12 @@
[submodule "ggml"]
path = ggml
url = https://github.com/ggml-org/ggml.git
url = https://github.com/leejet/ggml.git
[submodule "examples/server/frontend"]
path = examples/server/frontend
url = https://github.com/leejet/stable-ui.git
url = https://github.com/leejet/sdcpp-webui.git
[submodule "thirdparty/libwebp"]
path = thirdparty/libwebp
url = https://github.com/webmproject/libwebp.git
[submodule "thirdparty/libwebm"]
path = thirdparty/libwebm
url = https://github.com/webmproject/libwebm.git
Loading