• Joined on 2024-08-30
wangbo synced commits to refs/pull/12765/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:35 +08:00
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
fabed694a2 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 10 commits »
wangbo synced commits to refs/pull/12757/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:32 +08:00
5ad45ab3e0 Merge 040e2199ce9ebc389bddd837fb59f439ad02b854 into 8458ae2686
8458ae2686 Revert "fix: run text encoders on MPS GPU instead of CPU for Apple Silicon (#…" (#13070)
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 11 commits »
wangbo synced commits to refs/pull/12685/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:31 +08:00
8458ae2686 Revert "fix: run text encoders on MPS GPU instead of CPU for Apple Silicon (#…" (#13070)
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 11 commits »
wangbo synced commits to refs/pull/12647/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:28 +08:00
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
fabed694a2 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
f6b869d7d3 fp16 intermediates doen't work for some text enc models. (#13056)
Compare 10 commits »
wangbo synced commits to refs/pull/12535/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:25 +08:00
8458ae2686 Revert "fix: run text encoders on MPS GPU instead of CPU for Apple Silicon (#…" (#13070)
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 6 commits »
wangbo synced commits to refs/pull/12342/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:21 +08:00
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
fabed694a2 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
f6b869d7d3 fp16 intermediates doen't work for some text enc models. (#13056)
Compare 9 commits »
wangbo synced commits to refs/pull/12333/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:17 +08:00
8458ae2686 Revert "fix: run text encoders on MPS GPU instead of CPU for Apple Silicon (#…" (#13070)
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 12 commits »
wangbo synced commits to refs/pull/12170/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:14 +08:00
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
fabed694a2 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 11 commits »
wangbo synced commits to refs/pull/12043/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:13 +08:00
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
fabed694a2 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
f6b869d7d3 fp16 intermediates doen't work for some text enc models. (#13056)
Compare 9 commits »
wangbo synced commits to refs/pull/11294/merge at wangbo/ComfyUI from mirror 2026-03-20 08:34:11 +08:00
8458ae2686 Revert "fix: run text encoders on MPS GPU instead of CPU for Apple Silicon (#…" (#13070)
af6c5d6de9 tested with a modern version of comfyui
fd0261d2bc Reduce tiled decode peak memory (#13050)
Compare 8 commits »
wangbo synced commits to refs/pull/11294/head at wangbo/ComfyUI from mirror 2026-03-20 08:34:10 +08:00
af6c5d6de9 tested with a modern version of comfyui
Compare 2 commits »
wangbo synced commits to master at wangbo/ComfyUI from mirror 2026-03-20 08:33:55 +08:00
8458ae2686 Revert "fix: run text encoders on MPS GPU instead of CPU for Apple Silicon (#…" (#13070)
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
fabed694a2 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 5 commits »
wangbo synced commits to deepme987/add-execution-environment-api at wangbo/ComfyUI from mirror 2026-03-20 08:33:50 +08:00
cb1ca33ac0 Merge branch 'master' into deepme987/add-execution-environment-api
8458ae2686 Revert "fix: run text encoders on MPS GPU instead of CPU for Apple Silicon (#…" (#13070)
fd0261d2bc Reduce tiled decode peak memory (#13050)
ab14541ef7 memory: Add more exclusion criteria to pinned read (#13067)
6589562ae3 ltx: vae: implement chunked encoder + CPU IO chunking (Big VRAM reductions) (#13062)
Compare 6 commits »
wangbo synced and deleted reference refs/tags/refs/pull/13067/merge at wangbo/ComfyUI from mirror 2026-03-20 08:33:48 +08:00
wangbo synced and deleted reference refs/tags/refs/pull/13050/merge at wangbo/ComfyUI from mirror 2026-03-20 08:33:46 +08:00
wangbo synced and deleted reference refs/tags/refs/pull/13062/merge at wangbo/ComfyUI from mirror 2026-03-20 08:33:46 +08:00
wangbo synced and deleted reference refs/tags/comfyanonymous-patch-1 at wangbo/ComfyUI from mirror 2026-03-20 08:33:45 +08:00
wangbo synced commits to refs/pull/7701/merge at wangbo/ComfyUI from mirror 2026-03-20 00:25:22 +08:00
f6b869d7d3 fp16 intermediates doen't work for some text enc models. (#13056)
56ff88f951 Fix regression. (#13053)
9fff091f35 Further Reduce LTX VAE decode peak RAM usage (#13052)
dcd659590f Make more intermediate values follow the intermediate dtype. (#13051)
Compare 29 commits »
wangbo synced commits to refs/pull/6277/merge at wangbo/ComfyUI from mirror 2026-03-20 00:25:21 +08:00
f6b869d7d3 fp16 intermediates doen't work for some text enc models. (#13056)
56ff88f951 Fix regression. (#13053)
9fff091f35 Further Reduce LTX VAE decode peak RAM usage (#13052)
dcd659590f Make more intermediate values follow the intermediate dtype. (#13051)
Compare 24 commits »
wangbo synced commits to refs/pull/13062/merge at wangbo/ComfyUI from mirror 2026-03-20 00:25:20 +08:00
ec60b6969b ltx: vae-encode: round chunk sizes more strictly
Compare 2 commits »