DarthAffe
1 Followers · 0 Following
Repositories
16
 Projects Packages Public Activity Starred Repositories
1
DarthAffe synced new reference perf/reduce-sampling-cpu-overhead to DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:39 +00:00
DarthAffe synced commits to refs/tags/master-879-d32b4e8 at DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:39 +00:00
DarthAffe synced new reference refs/tags/master-879-d32b4e8 to DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:39 +00:00
DarthAffe synced commits to refs/tags/master-881-17860c0 at DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:39 +00:00
DarthAffe synced new reference refs/tags/master-881-17860c0 to DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:39 +00:00
DarthAffe synced commits to master at DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:39 +00:00
1330cebae8 feat: support building with upstream ggml (#1999)
17860c0e45 perf: parallelize host tensor elementwise and broadcast ops (#1998)
275ab58e01 perf: reduce CPU overhead in graph execution and sampling (#1997)
d32b4e893b fix: prevent clip_preprocess center crop from exceeding the resized image (#1995)
9982c9caae fix: propagate CUDA driver dependency to shared library consumers
Compare 6 commits »
DarthAffe synced commits to feat/support-upstream-ggml at DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:38 +00:00
DarthAffe synced new reference feat/support-upstream-ggml to DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 17:50:38 +00:00
DarthAffe synced commits to perf/ggml-implicit-gemm-conv3d at DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 09:40:37 +00:00
DarthAffe synced new reference perf/ggml-implicit-gemm-conv3d to DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 09:40:37 +00:00
DarthAffe synced commits to refs/tags/master-874-656a135 at DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 01:30:37 +00:00
DarthAffe synced new reference refs/tags/master-874-656a135 to DarthAffe/stable-diffusion.cpp from mirror 2026-09-19 01:30:37 +00:00
DarthAffe synced commits to feat/qwen-image-2-1 at DarthAffe/stable-diffusion.cpp from mirror 2026-09-18 17:20:38 +00:00
DarthAffe synced new reference feat/qwen-image-2-1 to DarthAffe/stable-diffusion.cpp from mirror 2026-09-18 17:20:38 +00:00
DarthAffe synced commits to master at DarthAffe/stable-diffusion.cpp from mirror 2026-09-18 17:20:38 +00:00
2ea8aff7ef perf: update ggml for faster direct convolutions (#1993)
adcac69650 perf: pad small attention heads to 64 for MMA Flash Attention (#1992)
656a1354c3 refactor: remove obsolete unused tensor filtering (#1984)
269e726015 fix: honor flash attention flag in LLM text encoder attention (#1987)
Compare 4 commits »
DarthAffe synced commits to perf/cuda-fa-head-padding at DarthAffe/stable-diffusion.cpp from mirror 2026-09-18 17:20:38 +00:00
DarthAffe synced new reference perf/cuda-fa-head-padding to DarthAffe/stable-diffusion.cpp from mirror 2026-09-18 17:20:38 +00:00
DarthAffe synced commits to perf/ggml-implicit-gemm-conv2d at DarthAffe/stable-diffusion.cpp from mirror 2026-09-18 17:20:38 +00:00
DarthAffe synced new reference perf/ggml-implicit-gemm-conv2d to DarthAffe/stable-diffusion.cpp from mirror 2026-09-18 17:20:38 +00:00
DarthAffe synced commits to refactor/remove-unused-tensor-filter at DarthAffe/stable-diffusion.cpp from mirror 2026-09-17 00:30:38 +00:00