liminfei-amd 70724ddf43 pixeldit: sanitize and warn on NaN PiD output on AMD fp16/bf16 ... On AMD ROCm (gfx11xx APUs/GPUs), the PiD PidNet forward in fp16/bf16 can return an all-NaN tensor through the ROCm/AOTriton attention path, which then corrupts the decoded image. The default (non-AOTriton) path stays clean, so the bad values come from an AOTriton attention miscompilation on gfx11xx rather than the PiD math itself (see ROCm/triton#909 and ROCm/aotriton#179). Guard the PidNet output on AMD: when it is fp16/bf16 and actually contains NaN/Inf, log a one-time warning that points at --use-split-cross-attention and clamp the values with nan_to_num before decode (the same pattern already used for flux/lumina). Non-AMD devices and fp32 paths are unaffected; finite outputs only pay a single isfinite() check. Fixes #14249 Signed-off-by: liminfei-amd <91481003+liminfei-amd@users.noreply.github.com>		2026-06-07 12:19:19 +08:00
..
ace	Support Ace Step 1.5 XL model. (#13317)	2026-04-07 03:13:47 -04:00
anima	Fix anima LLM adapter forward when manual cast (#12504)	2026-02-17 07:56:44 -08:00
audio	Disable sage attention in stable audio dit and VAE. (#14148)	2026-05-27 20:35:03 -04:00
aura	Enable Runtime Selection of Attention Functions (#9639)	2025-09-12 18:07:38 -04:00
cascade	cascade: remove dead weight init code (#13026)	2026-03-17 20:59:10 -04:00
chroma	Implement NAG on all the models based on the Flux code. (#12500)	2026-02-16 23:30:34 -05:00
chroma_radiance	Radiance: support variant with nonzero txt_ids (#14206)	2026-06-01 22:07:48 -07:00
cogvideo	Cogvideox (#13402)	2026-04-29 19:30:08 -04:00
cosmos	Speed up ernie model by a bit on nvidia and use higher quality rope. (#14192)	2026-05-30 17:53:37 -07:00
ernie	Speed up ernie model by a bit on nvidia and use higher quality rope. (#14192)	2026-05-30 17:53:37 -07:00
flux	Remove old useless no comfy kitchen fallback. (#14245)	2026-06-02 17:52:41 -07:00
genmo	Enable Runtime Selection of Attention Functions (#9639)	2025-09-12 18:07:38 -04:00
hidream	Enable Runtime Selection of Attention Functions (#9639)	2025-09-12 18:07:38 -04:00
hidream_o1	feat: Support HiDream-O1-Image (CORE-187) (#13817)	2026-05-11 20:35:53 -07:00
hunyuan3d	Enable Runtime Selection of Attention Functions (#9639)	2025-09-12 18:07:38 -04:00
hunyuan3dv2_1	MultiGPU Work Units For Accelerated Sampling (CORE-184) (#7063)	2026-05-25 18:26:40 -07:00
hunyuan_video	Implement NAG on all the models based on the Flux code. (#12500)	2026-02-16 23:30:34 -05:00
hydit	Change cosmos and hydit models to use the native RMSNorm. (#7934)	2025-05-04 06:26:20 -04:00
ideogram4	Fix ideogram if model dtype gets set to fp8. (#14291)	2026-06-04 19:20:22 -07:00
kandinsky5	Fix qwen scaled fp8 not working with kandinsky. Make basic t2i wf work. (#11162)	2025-12-06 17:50:10 -08:00
lens	Lens: some cleanup (#14112)	2026-05-26 10:32:53 +03:00
lightricks	fix: cross-attention AdaLN scale, shift, sigma parameters calculation (#14097)	2026-05-25 20:07:09 -07:00
lumina	Remove useless annotations imports. (#14105)	2026-05-25 19:23:29 -07:00
mmaudio/vae	Implement the mmaudio VAE. (#10300)	2025-10-11 22:57:23 -04:00
models	Add support for small flux.2 decoder (#13314)	2026-04-07 03:44:18 -04:00
modules	feat: Support NVIDIA PixelDiT and PiD (CORE-201) (#14103)	2026-05-26 17:50:14 -07:00
moge	Remove useless annotations imports. (#14105)	2026-05-25 19:23:29 -07:00
omnigen	Enable Runtime Selection of Attention Functions (#9639)	2025-09-12 18:07:38 -04:00
pixart	Remove windows line endings. (#8866)	2025-07-11 02:37:51 -04:00
pixeldit	pixeldit: sanitize and warn on NaN PiD output on AMD fp16/bf16	2026-06-07 12:19:19 +08:00
qwen_image	Remove useless code. (#14178)	2026-05-29 16:26:46 -07:00
rt_detr	CORE-13 feat: Support RT-DETRv4 detection model (#12748)	2026-03-28 23:34:10 -04:00
sam3	Improve SAM3 large input handling (#13767)	2026-05-07 17:18:28 -07:00
supir	feat: SUPIR model support (CORE-17) (#13250)	2026-04-18 23:02:01 -04:00
triposplat	feat: Add TripoSplat support (#14210)	2026-06-01 07:01:50 -07:00
wan	Support Wan-Dancer (#13813)	2026-05-09 14:02:56 -07:00
common_dit.py	add RMSNorm to comfy.ops	2025-04-14 18:00:33 -04:00
util.py	New Year ruff cleanup. (#11595)	2026-01-01 22:06:14 -05:00