From 2c018f8be623994a853be5731b215a80e802d122 Mon Sep 17 00:00:00 2001 From: YiYi Xu Date: Sun, 30 Nov 2025 14:43:51 -1000 Subject: [PATCH 1/5] Update docs/source/en/api/pipelines/hunyuan_video15.md --- docs/source/en/api/pipelines/hunyuan_video15.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/source/en/api/pipelines/hunyuan_video15.md b/docs/source/en/api/pipelines/hunyuan_video15.md index 9a9bdcb352..033c8490a0 100644 --- a/docs/source/en/api/pipelines/hunyuan_video15.md +++ b/docs/source/en/api/pipelines/hunyuan_video15.md @@ -54,7 +54,7 @@ export_to_video(video, "output.mp4", fps=15) ## Notes -- HunyuanVideo1.5 use attention masks with avariable-length sequences. For best performance, we recommend using an attention backend that handles padding efficiently. +- HunyuanVideo1.5 use attention masks with variable-length sequences. For best performance, we recommend using an attention backend that handles padding efficiently. - **H100/H800:** `_flash_3_hub` or `_flash_varlen_3` - **A100/A800/RTX 4090:** `flash` or `flash_varlen` From c7154707097f78e66a8cf7390950d856c8e4f5f2 Mon Sep 17 00:00:00 2001 From: yiyixuxu Date: Mon, 1 Dec 2025 02:06:13 +0100 Subject: [PATCH 2/5] add a note on changing guidance_scale on doc --- .../en/api/pipelines/hunyuan_video15.md | 35 +++++++++++++++++++ 1 file changed, 35 insertions(+) diff --git a/docs/source/en/api/pipelines/hunyuan_video15.md b/docs/source/en/api/pipelines/hunyuan_video15.md index 033c8490a0..fcd75b2aa9 100644 --- a/docs/source/en/api/pipelines/hunyuan_video15.md +++ b/docs/source/en/api/pipelines/hunyuan_video15.md @@ -67,6 +67,41 @@ Refer to the [Attention backends](../../optimization/attention_backends) guide f pipe.transformer.set_attention_backend("flash_varlen") # or your preferred backend ``` +- [`HunyuanVideo15Pipeline`] use guider and does not take `guidance_scale` parameter at runtime. + +You can check the default guider configuration using `pipe.guider`: + +```py +>>> pipe.guider +ClassifierFreeGuidance { + "_class_name": "ClassifierFreeGuidance", + "_diffusers_version": "0.36.0.dev0", + "enabled": true, + "guidance_rescale": 0.0, + "guidance_scale": 6.0, + "start": 0.0, + "stop": 1.0, + "use_original_formulation": false +} + +State: + step: None + num_inference_steps: None + timestep: None + count_prepared: 0 + enabled: True + num_conditions: 2 +``` + +To update guider configuration, you can run `pipe.guider = pipe.guider.new(...)` + +```py +pipe.guider = pipe.guider.new(guidance_scale=5.0) +``` + +Read more on Guider [here](../../modular_diffusers/guider). + + ## HunyuanVideo15Pipeline From 0dae8f956de3c2d3c63a904f4f90342bc6e49769 Mon Sep 17 00:00:00 2001 From: YiYi Xu Date: Sun, 30 Nov 2025 15:10:16 -1000 Subject: [PATCH 3/5] Apply suggestions from code review --- docs/source/en/api/pipelines/hunyuan_video15.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/source/en/api/pipelines/hunyuan_video15.md b/docs/source/en/api/pipelines/hunyuan_video15.md index fcd75b2aa9..b99114d792 100644 --- a/docs/source/en/api/pipelines/hunyuan_video15.md +++ b/docs/source/en/api/pipelines/hunyuan_video15.md @@ -57,8 +57,8 @@ export_to_video(video, "output.mp4", fps=15) - HunyuanVideo1.5 use attention masks with variable-length sequences. For best performance, we recommend using an attention backend that handles padding efficiently. - **H100/H800:** `_flash_3_hub` or `_flash_varlen_3` - - **A100/A800/RTX 4090:** `flash` or `flash_varlen` - - **Other GPUs:** `sage` + - **A100/A800/RTX 4090:** `flash_hub` or `flash_varlen` + - **Other GPUs:** `sage_hub` Refer to the [Attention backends](../../optimization/attention_backends) guide for more details about using a different backend. From 5989014cfe7d5010eb39d99f4d1597c3cda27ae9 Mon Sep 17 00:00:00 2001 From: YiYi Xu Date: Sun, 30 Nov 2025 15:21:17 -1000 Subject: [PATCH 4/5] Update docs/source/en/api/pipelines/hunyuan_video15.md --- docs/source/en/api/pipelines/hunyuan_video15.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/source/en/api/pipelines/hunyuan_video15.md b/docs/source/en/api/pipelines/hunyuan_video15.md index b99114d792..4e35232211 100644 --- a/docs/source/en/api/pipelines/hunyuan_video15.md +++ b/docs/source/en/api/pipelines/hunyuan_video15.md @@ -64,7 +64,7 @@ Refer to the [Attention backends](../../optimization/attention_backends) guide f ```py -pipe.transformer.set_attention_backend("flash_varlen") # or your preferred backend +pipe.transformer.set_attention_backend("flash_hub") # or your preferred backend ``` - [`HunyuanVideo15Pipeline`] use guider and does not take `guidance_scale` parameter at runtime. From 404d3fa9a5b8b82d55ca45d7177fdefebefcfd26 Mon Sep 17 00:00:00 2001 From: YiYi Xu Date: Sun, 30 Nov 2025 15:24:18 -1000 Subject: [PATCH 5/5] Update docs/source/en/api/pipelines/hunyuan_video15.md --- docs/source/en/api/pipelines/hunyuan_video15.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/source/en/api/pipelines/hunyuan_video15.md b/docs/source/en/api/pipelines/hunyuan_video15.md index 4e35232211..d86b9f37b2 100644 --- a/docs/source/en/api/pipelines/hunyuan_video15.md +++ b/docs/source/en/api/pipelines/hunyuan_video15.md @@ -99,7 +99,7 @@ To update guider configuration, you can run `pipe.guider = pipe.guider.new(...)` pipe.guider = pipe.guider.new(guidance_scale=5.0) ``` -Read more on Guider [here](../../modular_diffusers/guider). +Read more on Guider [here](../../modular_diffusers/guiders).