From 2dad9a319c49c865b0cfb211372c55d3a770403c Mon Sep 17 00:00:00 2001
From: "lzw478614@alibaba-inc.com" <lzw478614@alibaba-inc.com>
Date: Wed, 20 Aug 2025 20:17:41 +0800
Subject: [PATCH 01/10] update wan2.2-fun

---
 diffsynth/configs/model_config.py    |  3 ++-
 diffsynth/models/wan_video_dit.py    | 36 ++++++++++++++++++++++++++++
 diffsynth/pipelines/wan_video_new.py | 31 +++++++++++++++++-------
 3 files changed, 60 insertions(+), 10 deletions(-)

diff --git a/diffsynth/configs/model_config.py b/diffsynth/configs/model_config.py
index 9de5bfb..4731fee 100644
--- a/diffsynth/configs/model_config.py
+++ b/diffsynth/configs/model_config.py
@@ -150,6 +150,8 @@ model_loader_configs = [
     (None, "b61c605c2adbd23124d152ed28e049ae", ["wan_video_dit"], [WanModel], "civitai"), 
     (None, "1f5ab7703c6fc803fdded85ff040c316", ["wan_video_dit"], [WanModel], "civitai"),
     (None, "5b013604280dd715f8457c6ed6d6a626", ["wan_video_dit"], [WanModel], "civitai"),
+    (None, "2267d489f0ceb9f21836532952852ee5", ["wan_video_dit"], [WanModel], "civitai"),
+    (None, "47dbeab5e560db3180adf51dc0232fb1", ["wan_video_dit"], [WanModel], "civitai"),
     (None, "a61453409b67cd3246cf0c3bebad47ba", ["wan_video_dit", "wan_video_vace"], [WanModel, VaceWanModel], "civitai"),
     (None, "7a513e1f257a861512b1afd387a8ecd9", ["wan_video_dit", "wan_video_vace"], [WanModel, VaceWanModel], "civitai"),
     (None, "cb104773c6c2cb6df4f9529ad5c60d0b", ["wan_video_dit"], [WanModel], "diffusers"),
@@ -169,7 +171,6 @@ model_loader_configs = [
     (None, "8004730443f55db63092006dd9f7110e", ["qwen_image_text_encoder"], [QwenImageTextEncoder], "diffusers"),
     (None, "ed4ea5824d55ec3107b09815e318123a", ["qwen_image_vae"], [QwenImageVAE], "diffusers"),
     (None, "073bce9cf969e317e5662cd570c3e79c", ["qwen_image_blockwise_controlnet"], [QwenImageBlockWiseControlNet], "civitai"),
-    (None, "a9e54e480a628f0b956a688a81c33bab", ["qwen_image_blockwise_controlnet"], [QwenImageBlockWiseControlNet], "civitai"),
 ]
 huggingface_model_loader_configs = [
     # These configs are provided for detecting model type automatically.
diff --git a/diffsynth/models/wan_video_dit.py b/diffsynth/models/wan_video_dit.py
index 419f8cf..ba28dcb 100644
--- a/diffsynth/models/wan_video_dit.py
+++ b/diffsynth/models/wan_video_dit.py
@@ -713,6 +713,42 @@ class WanModelStateDictConverter:
                 "eps": 1e-6,
                 "require_clip_embedding": False,
             }
+        elif hash_state_dict_keys(state_dict) == "2267d489f0ceb9f21836532952852ee5":
+            # Wan2.2-Fun-A14B-Control
+            config = {
+                "has_image_input": False,
+                "patch_size": [1, 2, 2],
+                "in_dim": 52,
+                "dim": 5120,
+                "ffn_dim": 13824,
+                "freq_dim": 256,
+                "text_dim": 4096,
+                "out_dim": 16,
+                "num_heads": 40,
+                "num_layers": 40,
+                "eps": 1e-6,
+                "has_ref_conv": True,
+                "require_clip_embedding": False,
+            }
+        elif hash_state_dict_keys(state_dict) == "47dbeab5e560db3180adf51dc0232fb1":
+            # Wan2.2-Fun-A14B-Control-Camera
+            config = {
+                "has_image_input": False,
+                "patch_size": [1, 2, 2],
+                "in_dim": 36,
+                "dim": 5120,
+                "ffn_dim": 13824,
+                "freq_dim": 256,
+                "text_dim": 4096,
+                "out_dim": 16,
+                "num_heads": 40,
+                "num_layers": 40,
+                "eps": 1e-6,
+                "has_ref_conv": False,
+                "add_control_adapter": True,
+                "in_dim_control_adapter": 24,
+                "require_clip_embedding": False,
+            }
         else:
             config = {}
         return state_dict, config
diff --git a/diffsynth/pipelines/wan_video_new.py b/diffsynth/pipelines/wan_video_new.py
index 89adbdf..9e83282 100644
--- a/diffsynth/pipelines/wan_video_new.py
+++ b/diffsynth/pipelines/wan_video_new.py
@@ -677,8 +677,11 @@ class WanVideoUnit_FunControl(PipelineUnit):
         if clip_feature is None or y is None:
             clip_feature = torch.zeros((1, 257, 1280), dtype=pipe.torch_dtype, device=pipe.device)
             y = torch.zeros((1, 16, (num_frames - 1) // 4 + 1, height//8, width//8), dtype=pipe.torch_dtype, device=pipe.device)
+            if pipe.dit2 is not None:
+                y = torch.zeros((1, 20, (num_frames - 1) // 4 + 1, height//8, width//8), dtype=pipe.torch_dtype, device=pipe.device)
         else:
-            y = y[:, -16:]
+            if pipe.dit2 is None:
+                y = y[:, -16:]
         y = torch.concat([control_latents, y], dim=1)
         return {"clip_feature": clip_feature, "y": y}
     
@@ -698,6 +701,8 @@ class WanVideoUnit_FunReference(PipelineUnit):
         reference_image = reference_image.resize((width, height))
         reference_latents = pipe.preprocess_video([reference_image])
         reference_latents = pipe.vae.encode(reference_latents, device=pipe.device)
+        if pipe.image_encoder is None:
+            return {"reference_latents": reference_latents}
         clip_feature = pipe.preprocess_image(reference_image)
         clip_feature = pipe.image_encoder.encode_image([clip_feature])
         return {"reference_latents": reference_latents, "clip_feature": clip_feature}
@@ -707,13 +712,14 @@ class WanVideoUnit_FunReference(PipelineUnit):
 class WanVideoUnit_FunCameraControl(PipelineUnit):
     def __init__(self):
         super().__init__(
-            input_params=("height", "width", "num_frames", "camera_control_direction", "camera_control_speed", "camera_control_origin", "latents", "input_image"),
+            input_params=("height", "width", "num_frames", "camera_control_direction", "camera_control_speed", "camera_control_origin", "latents", "input_image", "tiled", "tile_size", "tile_stride"),
             onload_model_names=("vae",)
         )
 
-    def process(self, pipe: WanVideoPipeline, height, width, num_frames, camera_control_direction, camera_control_speed, camera_control_origin, latents, input_image):
+    def process(self, pipe: WanVideoPipeline, height, width, num_frames, camera_control_direction, camera_control_speed, camera_control_origin, latents, input_image, tiled, tile_size, tile_stride):
         if camera_control_direction is None:
             return {}
+        pipe.load_models_to_device(self.onload_model_names)
         camera_control_plucker_embedding = pipe.dit.control_adapter.process_camera_coordinates(
             camera_control_direction, num_frames, height, width, camera_control_speed, camera_control_origin)
         
@@ -729,13 +735,20 @@ class WanVideoUnit_FunCameraControl(PipelineUnit):
         control_camera_latents = control_camera_latents.contiguous().view(b, f // 4, c * 4, h, w).transpose(1, 2)
         control_camera_latents_input = control_camera_latents.to(device=pipe.device, dtype=pipe.torch_dtype)
 
-        input_image = input_image.resize((width, height))
-        input_latents = pipe.preprocess_video([input_image])
-        pipe.load_models_to_device(self.onload_model_names)
-        input_latents = pipe.vae.encode(input_latents, device=pipe.device)
-        y = torch.zeros_like(latents).to(pipe.device)
-        y[:, :, :1] = input_latents
+        image = pipe.preprocess_image(input_image.resize((width, height))).to(pipe.device)
+        vae_input = torch.concat([image.transpose(0, 1), torch.zeros(3, num_frames-1, height, width).to(image.device)], dim=1)
+        y = pipe.vae.encode([vae_input.to(dtype=pipe.torch_dtype, device=pipe.device)], device=pipe.device, tiled=tiled, tile_size=tile_size, tile_stride=tile_stride)[0]
         y = y.to(dtype=pipe.torch_dtype, device=pipe.device)
+
+        if pipe.dit2 is not None:
+            msk = torch.ones(1, num_frames, height//8, width//8, device=pipe.device)
+            msk[:, 1:] = 0
+            msk = torch.concat([torch.repeat_interleave(msk[:, 0:1], repeats=4, dim=1), msk[:, 1:]], dim=1)
+            msk = msk.view(1, msk.shape[1] // 4, 4, height//8, width//8)
+            msk = msk.transpose(1, 2)[0]
+            y = torch.cat([msk,y])
+            y = y.unsqueeze(0)
+            y = y.to(dtype=pipe.torch_dtype, device=pipe.device)
         return {"control_camera_latents_input": control_camera_latents_input, "y": y}
 
 

From c0172e70b19d39458eaceafce813c47d38970a4a Mon Sep 17 00:00:00 2001
From: "lzw478614@alibaba-inc.com" <lzw478614@alibaba-inc.com>
Date: Thu, 21 Aug 2025 12:59:41 +0800
Subject: [PATCH 02/10] update wan2.2-fun

---
 diffsynth/models/wan_video_camera_controller.py | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/diffsynth/models/wan_video_camera_controller.py b/diffsynth/models/wan_video_camera_controller.py
index 026b558..45a44ee 100644
--- a/diffsynth/models/wan_video_camera_controller.py
+++ b/diffsynth/models/wan_video_camera_controller.py
@@ -182,7 +182,7 @@ def process_pose_file(cam_params, width=672, height=384, original_pose_width=128
 
 
 def generate_camera_coordinates(
-    direction: Literal["Left", "Right", "Up", "Down", "LeftUp", "LeftDown", "RightUp", "RightDown"],
+    direction: Literal["Left", "Right", "Up", "Down", "LeftUp", "LeftDown", "RightUp", "RightDown", "In", "Out"],
     length: int,
     speed: float = 1/54,
     origin=(0, 0.532139961, 0.946026558, 0.5, 0.5, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0)
@@ -198,5 +198,9 @@ def generate_camera_coordinates(
             coor[13] += speed
         if "Down" in direction:
             coor[13] -= speed
+        if "In" in direction:
+            coor[18] -= speed
+        if "Out" in direction:
+            coor[18] += speed
         coordinates.append(coor)
     return coordinates

From 952933eeb1dc5f77dd0a4510a7b31c2a11a15fdd Mon Sep 17 00:00:00 2001
From: "lzw478614@alibaba-inc.com" <lzw478614@alibaba-inc.com>
Date: Thu, 21 Aug 2025 13:34:09 +0800
Subject: [PATCH 03/10] update wan2.2-fun

---
 diffsynth/configs/model_config.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/diffsynth/configs/model_config.py b/diffsynth/configs/model_config.py
index 4731fee..b4b847f 100644
--- a/diffsynth/configs/model_config.py
+++ b/diffsynth/configs/model_config.py
@@ -171,6 +171,7 @@ model_loader_configs = [
     (None, "8004730443f55db63092006dd9f7110e", ["qwen_image_text_encoder"], [QwenImageTextEncoder], "diffusers"),
     (None, "ed4ea5824d55ec3107b09815e318123a", ["qwen_image_vae"], [QwenImageVAE], "diffusers"),
     (None, "073bce9cf969e317e5662cd570c3e79c", ["qwen_image_blockwise_controlnet"], [QwenImageBlockWiseControlNet], "civitai"),
+    (None, "a9e54e480a628f0b956a688a81c33bab", ["qwen_image_blockwise_controlnet"], [QwenImageBlockWiseControlNet], "civitai"),
 ]
 huggingface_model_loader_configs = [
     # These configs are provided for detecting model type automatically.

From 46bd05b54d8aef87f81d70c4714c8b68f0d4c388 Mon Sep 17 00:00:00 2001
From: "lzw478614@alibaba-inc.com" <lzw478614@alibaba-inc.com>
Date: Thu, 21 Aug 2025 13:41:07 +0800
Subject: [PATCH 04/10] add examples

---
 .../Wan2.2-Fun-A14B-Control-Camera.py         | 43 +++++++++++++++++++
 .../Wan2.2-Fun-A14B-Control.py                | 35 +++++++++++++++
 .../model_inference/Wan2.2-Fun-A14B-Inp.py    | 35 +++++++++++++++
 3 files changed, 113 insertions(+)
 create mode 100644 examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py
 create mode 100644 examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py
 create mode 100644 examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Inp.py

diff --git a/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py
new file mode 100644
index 0000000..27cda27
--- /dev/null
+++ b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py
@@ -0,0 +1,43 @@
+import torch
+from diffsynth import save_video,VideoData
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from PIL import Image
+from modelscope import dataset_snapshot_download
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.enable_vram_management()
+
+
+dataset_snapshot_download(
+    dataset_id="DiffSynth-Studio/examples_in_diffsynth",
+    local_dir="./",
+    allow_file_pattern=f"data/examples/wan/input_image.jpg"
+)
+input_image = Image.open("data/examples/wan/input_image.jpg")
+
+video = pipe(
+    prompt="一艘小船正勇敢地乘风破浪前行。蔚蓝的大海波涛汹涌，白色的浪花拍打着船身，但小船毫不畏惧，坚定地驶向远方。阳光洒在水面上，闪烁着金色的光芒，为这壮丽的场景增添了一抹温暖。镜头拉近，可以看到船上的旗帜迎风飘扬，象征着不屈的精神与冒险的勇气。这段画面充满力量，激励人心，展现了面对挑战时的无畏与执着。",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    seed=0, tiled=True,
+    input_image=input_image,
+    camera_control_direction="Left", camera_control_speed=0.01,
+)
+save_video(video, "video_left.mp4", fps=15, quality=5)
+
+video = pipe(
+    prompt="一艘小船正勇敢地乘风破浪前行。蔚蓝的大海波涛汹涌，白色的浪花拍打着船身，但小船毫不畏惧，坚定地驶向远方。阳光洒在水面上，闪烁着金色的光芒，为这壮丽的场景增添了一抹温暖。镜头拉近，可以看到船上的旗帜迎风飘扬，象征着不屈的精神与冒险的勇气。这段画面充满力量，激励人心，展现了面对挑战时的无畏与执着。",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    seed=0, tiled=True,
+    input_image=input_image,
+    camera_control_direction="Up", camera_control_speed=0.01,
+)
+save_video(video, "video_up.mp4", fps=15, quality=5)
\ No newline at end of file
diff --git a/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py
new file mode 100644
index 0000000..2941422
--- /dev/null
+++ b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py
@@ -0,0 +1,35 @@
+import torch
+from diffsynth import save_video,VideoData
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from PIL import Image
+from modelscope import dataset_snapshot_download
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.enable_vram_management()
+
+dataset_snapshot_download(
+    dataset_id="DiffSynth-Studio/examples_in_diffsynth",
+    local_dir="./",
+    allow_file_pattern=["data/examples/wan/control_video.mp4", "data/examples/wan/reference_image_girl.png"]
+)
+
+# Control video
+control_video = VideoData("data/examples/wan/control_video.mp4", height=832, width=576)
+reference_image = Image.open("data/examples/wan/reference_image_girl.png").resize((576, 832))
+video = pipe(
+    prompt="扁平风格动漫，一位长发少女优雅起舞。她五官精致，大眼睛明亮有神，黑色长发柔顺光泽。身穿淡蓝色T恤和深蓝色牛仔短裤。背景是粉色。",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    control_video=control_video, reference_image=reference_image,
+    height=832, width=576, num_frames=49,
+    seed=1, tiled=True
+)
+save_video(video, "video.mp4", fps=15, quality=5)
\ No newline at end of file
diff --git a/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Inp.py b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Inp.py
new file mode 100644
index 0000000..87c8488
--- /dev/null
+++ b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Inp.py
@@ -0,0 +1,35 @@
+import torch
+from diffsynth import save_video
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from PIL import Image
+from modelscope import dataset_snapshot_download
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.enable_vram_management()
+
+dataset_snapshot_download(
+    dataset_id="DiffSynth-Studio/examples_in_diffsynth",
+    local_dir="./",
+    allow_file_pattern=f"data/examples/wan/input_image.jpg"
+)
+image = Image.open("data/examples/wan/input_image.jpg")
+
+# First and last frame to video
+video = pipe(
+    prompt="一艘小船正勇敢地乘风破浪前行。蔚蓝的大海波涛汹涌，白色的浪花拍打着船身，但小船毫不畏惧，坚定地驶向远方。阳光洒在水面上，闪烁着金色的光芒，为这壮丽的场景增添了一抹温暖。镜头拉近，可以看到船上的旗帜迎风飘扬，象征着不屈的精神与冒险的勇气。这段画面充满力量，激励人心，展现了面对挑战时的无畏与执着。",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    input_image=image,
+    seed=0, tiled=True,
+    # You can input `end_image=xxx` to control the last frame of the video.
+    # The model will automatically generate the dynamic content between `input_image` and `end_image`.
+)
+save_video(video, "video.mp4", fps=15, quality=5)
\ No newline at end of file

From 0d81626fe78a311bf44d8dcafe80ed56fb6c6aa7 Mon Sep 17 00:00:00 2001
From: "lzw478614@alibaba-inc.com" <lzw478614@alibaba-inc.com>
Date: Thu, 21 Aug 2025 20:08:49 +0800
Subject: [PATCH 05/10] update wan2.2-fun

---
 diffsynth/models/wan_video_dit.py    |  1 +
 diffsynth/pipelines/wan_video_new.py | 30 ++++++++++++++++------------
 2 files changed, 18 insertions(+), 13 deletions(-)

diff --git a/diffsynth/models/wan_video_dit.py b/diffsynth/models/wan_video_dit.py
index ba28dcb..9420736 100644
--- a/diffsynth/models/wan_video_dit.py
+++ b/diffsynth/models/wan_video_dit.py
@@ -294,6 +294,7 @@ class WanModel(torch.nn.Module):
     ):
         super().__init__()
         self.dim = dim
+        self.in_dim = dim
         self.freq_dim = freq_dim
         self.has_image_input = has_image_input
         self.patch_size = patch_size
diff --git a/diffsynth/pipelines/wan_video_new.py b/diffsynth/pipelines/wan_video_new.py
index 9e83282..53df7d9 100644
--- a/diffsynth/pipelines/wan_video_new.py
+++ b/diffsynth/pipelines/wan_video_new.py
@@ -663,25 +663,23 @@ class WanVideoUnit_ImageEmbedderFused(PipelineUnit):
 class WanVideoUnit_FunControl(PipelineUnit):
     def __init__(self):
         super().__init__(
-            input_params=("control_video", "num_frames", "height", "width", "tiled", "tile_size", "tile_stride", "clip_feature", "y"),
+            input_params=("control_video", "num_frames", "height", "width", "tiled", "tile_size", "tile_stride", "clip_feature", "y", "latents"),
             onload_model_names=("vae",)
         )
 
-    def process(self, pipe: WanVideoPipeline, control_video, num_frames, height, width, tiled, tile_size, tile_stride, clip_feature, y):
+    def process(self, pipe: WanVideoPipeline, control_video, num_frames, height, width, tiled, tile_size, tile_stride, clip_feature, y, latents):
         if control_video is None:
             return {}
         pipe.load_models_to_device(self.onload_model_names)
         control_video = pipe.preprocess_video(control_video)
         control_latents = pipe.vae.encode(control_video, device=pipe.device, tiled=tiled, tile_size=tile_size, tile_stride=tile_stride).to(dtype=pipe.torch_dtype, device=pipe.device)
         control_latents = control_latents.to(dtype=pipe.torch_dtype, device=pipe.device)
+        y_dim = pipe.dit.in_dim-control_latents.shape[1]-latents.shape[1]
         if clip_feature is None or y is None:
             clip_feature = torch.zeros((1, 257, 1280), dtype=pipe.torch_dtype, device=pipe.device)
-            y = torch.zeros((1, 16, (num_frames - 1) // 4 + 1, height//8, width//8), dtype=pipe.torch_dtype, device=pipe.device)
-            if pipe.dit2 is not None:
-                y = torch.zeros((1, 20, (num_frames - 1) // 4 + 1, height//8, width//8), dtype=pipe.torch_dtype, device=pipe.device)
+            y = torch.zeros((1, y_dim, (num_frames - 1) // 4 + 1, height//8, width//8), dtype=pipe.torch_dtype, device=pipe.device)
         else:
-            if pipe.dit2 is None:
-                y = y[:, -16:]
+            y = y[:, -y_dim:]
         y = torch.concat([control_latents, y], dim=1)
         return {"clip_feature": clip_feature, "y": y}
     
@@ -734,13 +732,19 @@ class WanVideoUnit_FunCameraControl(PipelineUnit):
         control_camera_latents = control_camera_latents.contiguous().view(b, f // 4, 4, c, h, w).transpose(2, 3)
         control_camera_latents = control_camera_latents.contiguous().view(b, f // 4, c * 4, h, w).transpose(1, 2)
         control_camera_latents_input = control_camera_latents.to(device=pipe.device, dtype=pipe.torch_dtype)
-
-        image = pipe.preprocess_image(input_image.resize((width, height))).to(pipe.device)
-        vae_input = torch.concat([image.transpose(0, 1), torch.zeros(3, num_frames-1, height, width).to(image.device)], dim=1)
-        y = pipe.vae.encode([vae_input.to(dtype=pipe.torch_dtype, device=pipe.device)], device=pipe.device, tiled=tiled, tile_size=tile_size, tile_stride=tile_stride)[0]
+        
+        input_image = input_image.resize((width, height))
+        input_latents = pipe.preprocess_video([input_image])
+        input_latents = pipe.vae.encode(input_latents, device=pipe.device)
+        y = torch.zeros_like(latents).to(pipe.device)
+        y[:, :, :1] = input_latents
         y = y.to(dtype=pipe.torch_dtype, device=pipe.device)
 
-        if pipe.dit2 is not None:
+        if y.shape[1] != pipe.dit.in_dim - latents.shape[1]:
+            image = pipe.preprocess_image(input_image.resize((width, height))).to(pipe.device)
+            vae_input = torch.concat([image.transpose(0, 1), torch.zeros(3, num_frames-1, height, width).to(image.device)], dim=1)
+            y = pipe.vae.encode([vae_input.to(dtype=pipe.torch_dtype, device=pipe.device)], device=pipe.device, tiled=tiled, tile_size=tile_size, tile_stride=tile_stride)[0]
+            y = y.to(dtype=pipe.torch_dtype, device=pipe.device)
             msk = torch.ones(1, num_frames, height//8, width//8, device=pipe.device)
             msk[:, 1:] = 0
             msk = torch.concat([torch.repeat_interleave(msk[:, 0:1], repeats=4, dim=1), msk[:, 1:]], dim=1)
@@ -1061,7 +1065,7 @@ def model_fn_wan_video(
     if clip_feature is not None and dit.require_clip_embedding:
         clip_embdding = dit.img_emb(clip_feature)
         context = torch.cat([clip_embdding, context], dim=1)
-    
+
     # Add camera control
     x, (f, h, w) = dit.patchify(x, control_camera_latents_input)
     

From c4b97cd591f0f3b1c3b9682b0f7d9530d52b7e9e Mon Sep 17 00:00:00 2001
From: "lzw478614@alibaba-inc.com" <lzw478614@alibaba-inc.com>
Date: Fri, 22 Aug 2025 09:38:19 +0800
Subject: [PATCH 06/10] update wan2.2-fun

---
 diffsynth/models/wan_video_dit.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/diffsynth/models/wan_video_dit.py b/diffsynth/models/wan_video_dit.py
index 9420736..1a54728 100644
--- a/diffsynth/models/wan_video_dit.py
+++ b/diffsynth/models/wan_video_dit.py
@@ -294,7 +294,7 @@ class WanModel(torch.nn.Module):
     ):
         super().__init__()
         self.dim = dim
-        self.in_dim = dim
+        self.in_dim = in_dim
         self.freq_dim = freq_dim
         self.has_image_input = has_image_input
         self.patch_size = patch_size

From cb49f0283f77b89d8ba22f708c2766496c12dbe4 Mon Sep 17 00:00:00 2001
From: lzws <63908509+lzws@users.noreply.github.com>
Date: Fri, 22 Aug 2025 14:18:16 +0800
Subject: [PATCH 07/10] Rename Wan2.2-Fun-A14B-Inp.py to Wan2.2-Fun-A14B-InP.py

---
 .../{Wan2.2-Fun-A14B-Inp.py => Wan2.2-Fun-A14B-InP.py}          | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
 rename examples/wanvideo/model_inference/{Wan2.2-Fun-A14B-Inp.py => Wan2.2-Fun-A14B-InP.py} (97%)

diff --git a/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Inp.py b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py
similarity index 97%
rename from examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Inp.py
rename to examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py
index 87c8488..c63e522 100644
--- a/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Inp.py
+++ b/examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py
@@ -32,4 +32,4 @@ video = pipe(
     # You can input `end_image=xxx` to control the last frame of the video.
     # The model will automatically generate the dynamic content between `input_image` and `end_image`.
 )
-save_video(video, "video.mp4", fps=15, quality=5)
\ No newline at end of file
+save_video(video, "video.mp4", fps=15, quality=5)

From f0916e6baecfa70393b01a4ec1f9f50e68142c99 Mon Sep 17 00:00:00 2001
From: lzws <lzw478614@alibaba-inc.com>
Date: Tue, 26 Aug 2025 16:37:47 +0800
Subject: [PATCH 08/10] update wan2.2-fun training scripts

---
 README.md                                     |  3 ++
 README_zh.md                                  |  3 ++
 examples/wanvideo/README.md                   |  3 ++
 examples/wanvideo/README_zh.md                |  3 ++
 .../full/Wan2.2-Fun-A14B-Control-Camera.sh    | 35 +++++++++++++++++
 .../full/Wan2.2-Fun-A14B-Control.sh           | 35 +++++++++++++++++
 .../full/Wan2.2-Fun-A14B-InP.sh               | 33 ++++++++++++++++
 .../lora/Wan2.2-Fun-A14B-Control-Camera.sh    | 39 +++++++++++++++++++
 .../lora/Wan2.2-Fun-A14B-Control.sh           | 39 +++++++++++++++++++
 .../lora/Wan2.2-Fun-A14B-InP.sh               | 37 ++++++++++++++++++
 .../Wan2.2-Fun-A14B-Control-Camera.py         | 34 ++++++++++++++++
 .../validate_full/Wan2.2-Fun-A14B-Control.py  | 35 +++++++++++++++++
 .../validate_full/Wan2.2-Fun-A14B-InP.py      | 32 +++++++++++++++
 .../Wan2.2-Fun-A14B-Control-Camera.py         | 32 +++++++++++++++
 .../validate_lora/Wan2.2-Fun-A14B-Control.py  | 33 ++++++++++++++++
 .../validate_lora/Wan2.2-Fun-A14B-InP.py      | 31 +++++++++++++++
 16 files changed, 427 insertions(+)
 create mode 100644 examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh
 create mode 100644 examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh
 create mode 100644 examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh
 create mode 100644 examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh
 create mode 100644 examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh
 create mode 100644 examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh
 create mode 100644 examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py
 create mode 100644 examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py
 create mode 100644 examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py
 create mode 100644 examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py
 create mode 100644 examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py
 create mode 100644 examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py

diff --git a/README.md b/README.md
index b1c8716..78f2114 100644
--- a/README.md
+++ b/README.md
@@ -204,6 +204,9 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./examples/wanvideo/model_inference/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-TI2V-5B.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
+|[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./examples/wanvideo/model_inference/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-1.3B.py)|
 |[Wan-AI/Wan2.1-T2V-14B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-14B)||[code](./examples/wanvideo/model_inference/Wan2.1-T2V-14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-T2V-14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-T2V-14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-T2V-14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-14B.py)|
 |[Wan-AI/Wan2.1-I2V-14B-480P](https://modelscope.cn/models/Wan-AI/Wan2.1-I2V-14B-480P)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.1-I2V-14B-480P.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-I2V-14B-480P.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-I2V-14B-480P.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-I2V-14B-480P.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-I2V-14B-480P.py)|
diff --git a/README_zh.md b/README_zh.md
index a8ff395..6f8aa1f 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -204,6 +204,9 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./examples/wanvideo/model_inference/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-TI2V-5B.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
+|[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./examples/wanvideo/model_inference/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-1.3B.py)|
 |[Wan-AI/Wan2.1-T2V-14B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-14B)||[code](./examples/wanvideo/model_inference/Wan2.1-T2V-14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-T2V-14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-T2V-14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-T2V-14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-14B.py)|
 |[Wan-AI/Wan2.1-I2V-14B-480P](https://modelscope.cn/models/Wan-AI/Wan2.1-I2V-14B-480P)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.1-I2V-14B-480P.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-I2V-14B-480P.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-I2V-14B-480P.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-I2V-14B-480P.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-I2V-14B-480P.py)|
diff --git a/examples/wanvideo/README.md b/examples/wanvideo/README.md
index 4e5195a..2587928 100644
--- a/examples/wanvideo/README.md
+++ b/examples/wanvideo/README.md
@@ -51,6 +51,9 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./model_inference/Wan2.2-I2V-A14B.py)|[code](./model_training/full/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./model_inference/Wan2.2-T2V-A14B.py)|[code](./model_training/full/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./model_inference/Wan2.2-TI2V-5B.py)|[code](./model_training/full/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_lora/Wan2.2-TI2V-5B.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
+|[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./model_inference/Wan2.1-T2V-1.3B.py)|[code](./model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_lora/Wan2.1-T2V-1.3B.py)|
 |[Wan-AI/Wan2.1-T2V-14B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-14B)||[code](./model_inference/Wan2.1-T2V-14B.py)|[code](./model_training/full/Wan2.1-T2V-14B.sh)|[code](./model_training/validate_full/Wan2.1-T2V-14B.py)|[code](./model_training/lora/Wan2.1-T2V-14B.sh)|[code](./model_training/validate_lora/Wan2.1-T2V-14B.py)|
 |[Wan-AI/Wan2.1-I2V-14B-480P](https://modelscope.cn/models/Wan-AI/Wan2.1-I2V-14B-480P)|`input_image`|[code](./model_inference/Wan2.1-I2V-14B-480P.py)|[code](./model_training/full/Wan2.1-I2V-14B-480P.sh)|[code](./model_training/validate_full/Wan2.1-I2V-14B-480P.py)|[code](./model_training/lora/Wan2.1-I2V-14B-480P.sh)|[code](./model_training/validate_lora/Wan2.1-I2V-14B-480P.py)|
diff --git a/examples/wanvideo/README_zh.md b/examples/wanvideo/README_zh.md
index bcc076f..b591758 100644
--- a/examples/wanvideo/README_zh.md
+++ b/examples/wanvideo/README_zh.md
@@ -51,6 +51,9 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./model_inference/Wan2.2-I2V-A14B.py)|[code](./model_training/full/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./model_inference/Wan2.2-T2V-A14B.py)|[code](./model_training/full/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./model_inference/Wan2.2-TI2V-5B.py)|[code](./model_training/full/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_lora/Wan2.2-TI2V-5B.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
+|[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./model_inference/Wan2.1-T2V-1.3B.py)|[code](./model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_lora/Wan2.1-T2V-1.3B.py)|
 |[Wan-AI/Wan2.1-T2V-14B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-14B)||[code](./model_inference/Wan2.1-T2V-14B.py)|[code](./model_training/full/Wan2.1-T2V-14B.sh)|[code](./model_training/validate_full/Wan2.1-T2V-14B.py)|[code](./model_training/lora/Wan2.1-T2V-14B.sh)|[code](./model_training/validate_lora/Wan2.1-T2V-14B.py)|
 |[Wan-AI/Wan2.1-I2V-14B-480P](https://modelscope.cn/models/Wan-AI/Wan2.1-I2V-14B-480P)|`input_image`|[code](./model_inference/Wan2.1-I2V-14B-480P.py)|[code](./model_training/full/Wan2.1-I2V-14B-480P.sh)|[code](./model_training/validate_full/Wan2.1-I2V-14B-480P.py)|[code](./model_training/lora/Wan2.1-I2V-14B-480P.sh)|[code](./model_training/validate_lora/Wan2.1-I2V-14B-480P.py)|
diff --git a/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh b/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh
new file mode 100644
index 0000000..4352a05
--- /dev/null
+++ b/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh
@@ -0,0 +1,35 @@
+accelerate launch --config_file examples/wanvideo/model_training/full/accelerate_config_14B.yaml examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_camera_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control-Camera:high_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control-Camera:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control-Camera:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control-Camera_high_niose_full" \
+  --trainable_models "dit" \
+  --extra_inputs "input_image,camera_control_direction,camera_control_speed" \
+  --max_timestep_boundary 0.358 \
+  --min_timestep_boundary 0
+# boundary corresponds to timesteps [900, 1000]
+
+accelerate launch --config_file examples/wanvideo/model_training/full/accelerate_config_14B.yaml examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_camera_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control-Camera:low_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control-Camera:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control-Camera:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control-Camera_low_noise_full" \
+  --trainable_models "dit" \
+  --extra_inputs "input_image,camera_control_direction,camera_control_speed" \
+  --max_timestep_boundary 1 \
+  --min_timestep_boundary 0.358
+# boundary corresponds to timesteps [0, 900]
\ No newline at end of file
diff --git a/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh b/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh
new file mode 100644
index 0000000..2cc7da0
--- /dev/null
+++ b/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh
@@ -0,0 +1,35 @@
+accelerate launch --config_file examples/wanvideo/model_training/full/accelerate_config_14B.yaml examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_reference_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control:high_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control_high_niose_full" \
+  --trainable_models "dit" \
+  --extra_inputs "control_video,reference_image" \
+  --max_timestep_boundary 0.358 \
+  --min_timestep_boundary 0
+# boundary corresponds to timesteps [900, 1000]
+
+accelerate launch --config_file examples/wanvideo/model_training/full/accelerate_config_14B.yaml examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_reference_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control:low_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control_low_noise_full" \
+  --trainable_models "dit" \
+  --extra_inputs "control_video,reference_image" \
+  --max_timestep_boundary 1 \
+  --min_timestep_boundary 0.358
+# boundary corresponds to timesteps [0, 900]
\ No newline at end of file
diff --git a/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh b/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh
new file mode 100644
index 0000000..225b888
--- /dev/null
+++ b/examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh
@@ -0,0 +1,33 @@
+accelerate launch --config_file examples/wanvideo/model_training/full/accelerate_config_14B.yaml examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-InP:high_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-InP:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-InP_high_niose_full" \
+  --trainable_models "dit" \
+  --extra_inputs "input_image,end_image" \
+  --max_timestep_boundary 0.358 \
+  --min_timestep_boundary 0
+# boundary corresponds to timesteps [900, 1000]
+
+accelerate launch --config_file examples/wanvideo/model_training/full/accelerate_config_14B.yaml examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-InP:low_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-InP:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-InP_low_noise_full" \
+  --trainable_models "dit" \
+  --extra_inputs "input_image,end_image" \
+  --max_timestep_boundary 1 \
+  --min_timestep_boundary 0.358
+# boundary corresponds to timesteps [0, 900]
\ No newline at end of file
diff --git a/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh b/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh
new file mode 100644
index 0000000..b97e800
--- /dev/null
+++ b/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh
@@ -0,0 +1,39 @@
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_camera_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control-Camera:high_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control-Camera:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control-Camera:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control-Camera_high_niose_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --extra_inputs "input_image,camera_control_direction,camera_control_speed" \
+  --max_timestep_boundary 0.358 \
+  --min_timestep_boundary 0
+# boundary corresponds to timesteps [900, 1000]
+
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_camera_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control-Camera:low_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control-Camera:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control-Camera:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control-Camera_low_noise_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --extra_inputs "input_image,camera_control_direction,camera_control_speed" \
+  --max_timestep_boundary 1 \
+  --min_timestep_boundary 0.358
+# boundary corresponds to timesteps [0, 900]
\ No newline at end of file
diff --git a/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh b/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh
new file mode 100644
index 0000000..f0af530
--- /dev/null
+++ b/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh
@@ -0,0 +1,39 @@
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_reference_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control:high_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control_high_niose_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --extra_inputs "control_video,reference_image" \
+  --max_timestep_boundary 0.358 \
+  --min_timestep_boundary 0
+# boundary corresponds to timesteps [900, 1000]
+
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_reference_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-Control:low_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-Control:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-Control_low_noise_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --extra_inputs "control_video,reference_image" \
+  --max_timestep_boundary 1 \
+  --min_timestep_boundary 0.358
+# boundary corresponds to timesteps [0, 900]
\ No newline at end of file
diff --git a/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh b/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh
new file mode 100644
index 0000000..94cf196
--- /dev/null
+++ b/examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh
@@ -0,0 +1,37 @@
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-InP:high_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-InP:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-InP_high_niose_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --extra_inputs "input_image,end_image" \
+  --max_timestep_boundary 0.358 \
+  --min_timestep_boundary 0
+# boundary corresponds to timesteps [900, 1000]
+
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.2-Fun-A14B-InP:low_noise_model/diffusion_pytorch_model*.safetensors,PAI/Wan2.2-Fun-A14B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.2-Fun-A14B-InP:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.2-Fun-A14B-InP_low_noise_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --extra_inputs "input_image,end_image" \
+  --max_timestep_boundary 1 \
+  --min_timestep_boundary 0.358
+# boundary corresponds to timesteps [0, 900]
\ No newline at end of file
diff --git a/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py
new file mode 100644
index 0000000..5c3a3e4
--- /dev/null
+++ b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py
@@ -0,0 +1,34 @@
+import torch
+from PIL import Image
+from diffsynth import save_video, VideoData, load_state_dict
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.2-Fun-A14B-Control-Camera_high_noise_full/epoch-1.safetensors")
+pipe.dit.load_state_dict(state_dict)
+state_dict = load_state_dict("models/train/Wan2.2-Fun-A14B-Control-Camera_low_noise_full/epoch-1.safetensors")
+pipe.dit2.load_state_dict(state_dict)
+pipe.enable_vram_management()
+
+video = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)
+
+# First and last frame to video
+video = pipe(
+    prompt="from sunset to night, a small town, light, house, river",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    input_image=video[0],
+    camera_control_direction="Left", camera_control_speed=0.0,
+    seed=0, tiled=True
+)
+save_video(video, "video_Wan2.2-Fun-A14B-Comtrol-Camera.mp4", fps=15, quality=5)
diff --git a/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py
new file mode 100644
index 0000000..685b48c
--- /dev/null
+++ b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py
@@ -0,0 +1,35 @@
+import torch
+from PIL import Image
+from diffsynth import save_video, VideoData, load_state_dict
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.2-Fun-A14B-Control_high_noise_full/epoch-1.safetensors")
+pipe.dit.load_state_dict(state_dict)
+state_dict = load_state_dict("models/train/Wan2.2-Fun-A14B-Control_low_noise_full/epoch-1.safetensors")
+pipe.dit2.load_state_dict(state_dict)
+pipe.enable_vram_management()
+
+video = VideoData("data/example_video_dataset/video1_softedge.mp4", height=480, width=832)
+video = [video[i] for i in range(81)]
+reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
+
+# Control video
+video = pipe(
+    prompt="from sunset to night, a small town, light, house, river",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    control_video=video, reference_image=reference_image,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.2-Fun-A14B-Control.mp4", fps=15, quality=5)
diff --git a/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py
new file mode 100644
index 0000000..154c4e4
--- /dev/null
+++ b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py
@@ -0,0 +1,32 @@
+import torch
+from PIL import Image
+from diffsynth import save_video, VideoData, load_state_dict
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.2-Fun-A14B-InP_high_noise_full/epoch-1.safetensors")
+pipe.dit.load_state_dict(state_dict)
+state_dict = load_state_dict("models/train/Wan2.2-Fun-A14B-InP_low_noise_full/epoch-1.safetensors")
+pipe.dit2.load_state_dict(state_dict)
+pipe.enable_vram_management()
+video = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)
+
+# First and last frame to video
+video = pipe(
+    prompt="from sunset to night, a small town, light, house, river",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    input_image=video[0], end_image=video[80],
+    seed=0, tiled=True
+)
+save_video(video, "video_Wan2.2-Fun-A14B-InP.mp4", fps=15, quality=5)
diff --git a/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py
new file mode 100644
index 0000000..85dd973
--- /dev/null
+++ b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py
@@ -0,0 +1,32 @@
+import torch
+from PIL import Image
+from diffsynth import save_video, VideoData, load_state_dict
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.2-Fun-A14B-Control-Camera_high_noise_lora/epoch-4.safetensors", alpha=1)
+pipe.load_lora(pipe.dit2, "models/train/Wan2.2-Fun-A14B-Control-Camera_low_noise_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+video = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)
+
+# First and last frame to video
+video = pipe(
+    prompt="from sunset to night, a small town, light, house, river",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    input_image=video[0],
+    camera_control_direction="Left", camera_control_speed=0.0,
+    seed=0, tiled=True
+)
+save_video(video, "video_Wan2.2-Fun-A14B-Comtrol-Camera.mp4", fps=15, quality=5)
diff --git a/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py
new file mode 100644
index 0000000..11dda5b
--- /dev/null
+++ b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py
@@ -0,0 +1,33 @@
+import torch
+from PIL import Image
+from diffsynth import save_video, VideoData
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.2-Fun-A14B-Control_high_noise_lora/epoch-4.safetensors", alpha=1)
+pipe.load_lora(pipe.dit2, "models/train/Wan2.2-Fun-A14B-Control_low_noise_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+video = VideoData("data/example_video_dataset/video1_softedge.mp4", height=480, width=832)
+video = [video[i] for i in range(81)]
+reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
+
+# Control video
+video = pipe(
+    prompt="from sunset to night, a small town, light, house, river",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    control_video=video, reference_image=reference_image,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.2-Fun-A14B-Control.mp4", fps=15, quality=5)
diff --git a/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py
new file mode 100644
index 0000000..90afa46
--- /dev/null
+++ b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py
@@ -0,0 +1,31 @@
+import torch
+from PIL import Image
+from diffsynth import save_video, VideoData
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-InP", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.2-Fun-A14B-InP_high_noise_lora/epoch-4.safetensors", alpha=1)
+pipe.load_lora(pipe.dit2, "models/train/Wan2.2-Fun-A14B-InP_low_noise_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+video = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)
+
+# First and last frame to video
+video = pipe(
+    prompt="from sunset to night, a small town, light, house, river",
+    negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
+    input_image=video[0], end_image=video[80],
+    seed=0, tiled=True
+)
+save_video(video, "video_Wan2.2-Fun-A14B-InP.mp4", fps=15, quality=5)

From 2b7a2548b49fa9b114b90f13a39489c83ed38401 Mon Sep 17 00:00:00 2001
From: lzws <lzw478614@alibaba-inc.com>
Date: Tue, 26 Aug 2025 17:11:48 +0800
Subject: [PATCH 09/10] update wan2.2-fun model overview in readme

---
 README.md                      | 2 +-
 README_zh.md                   | 2 +-
 examples/wanvideo/README.md    | 2 +-
 examples/wanvideo/README_zh.md | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/README.md b/README.md
index 78f2114..3f26d96 100644
--- a/README.md
+++ b/README.md
@@ -204,7 +204,7 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./examples/wanvideo/model_inference/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-TI2V-5B.py)|
-|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-InP)|`input_image`, `end_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
 |[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
 |[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./examples/wanvideo/model_inference/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-1.3B.py)|
diff --git a/README_zh.md b/README_zh.md
index 6f8aa1f..dd532b8 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -204,7 +204,7 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./examples/wanvideo/model_inference/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-TI2V-5B.py)|
-|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-InP)|`input_image`, `end_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
 |[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
 |[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./examples/wanvideo/model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./examples/wanvideo/model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./examples/wanvideo/model_inference/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./examples/wanvideo/model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-1.3B.py)|
diff --git a/examples/wanvideo/README.md b/examples/wanvideo/README.md
index 2587928..e402a8a 100644
--- a/examples/wanvideo/README.md
+++ b/examples/wanvideo/README.md
@@ -51,7 +51,7 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./model_inference/Wan2.2-I2V-A14B.py)|[code](./model_training/full/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./model_inference/Wan2.2-T2V-A14B.py)|[code](./model_training/full/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./model_inference/Wan2.2-TI2V-5B.py)|[code](./model_training/full/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_lora/Wan2.2-TI2V-5B.py)|
-|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-InP)|`input_image`, `end_image`|[code](./model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
 |[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
 |[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./model_inference/Wan2.1-T2V-1.3B.py)|[code](./model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_lora/Wan2.1-T2V-1.3B.py)|
diff --git a/examples/wanvideo/README_zh.md b/examples/wanvideo/README_zh.md
index b591758..e758098 100644
--- a/examples/wanvideo/README_zh.md
+++ b/examples/wanvideo/README_zh.md
@@ -51,7 +51,7 @@ save_video(video, "video1.mp4", fps=15, quality=5)
 |[Wan-AI/Wan2.2-I2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-I2V-A14B)|`input_image`|[code](./model_inference/Wan2.2-I2V-A14B.py)|[code](./model_training/full/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-I2V-A14B.py)|[code](./model_training/lora/Wan2.2-I2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-I2V-A14B.py)|
 |[Wan-AI/Wan2.2-T2V-A14B](https://modelscope.cn/models/Wan-AI/Wan2.2-T2V-A14B)||[code](./model_inference/Wan2.2-T2V-A14B.py)|[code](./model_training/full/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_full/Wan2.2-T2V-A14B.py)|[code](./model_training/lora/Wan2.2-T2V-A14B.sh)|[code](./model_training/validate_lora/Wan2.2-T2V-A14B.py)|
 |[Wan-AI/Wan2.2-TI2V-5B](https://modelscope.cn/models/Wan-AI/Wan2.2-TI2V-5B)|`input_image`|[code](./model_inference/Wan2.2-TI2V-5B.py)|[code](./model_training/full/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_full/Wan2.2-TI2V-5B.py)|[code](./model_training/lora/Wan2.2-TI2V-5B.sh)|[code](./model_training/validate_lora/Wan2.2-TI2V-5B.py)|
-|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.1-Fun-A14B-InP)|`input_image`, `end_image`|[code](./model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
+|[PAI/Wan2.2-Fun-A14B-InP](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-InP)|`input_image`, `end_image`|[code](./model_inference/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/full/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-InP.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-InP.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-InP.py)|
 |[PAI/Wan2.2-Fun-A14B-Control](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control)|`control_video`, `reference_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control.py)|
 |[PAI/Wan2.2-Fun-A14B-Control-Camera](https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-Control-Camera)|`control_camera_video`, `input_image`|[code](./model_inference/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/full/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py)|[code](./model_training/lora/Wan2.2-Fun-A14B-Control-Camera.sh)|[code](./model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py)|
 |[Wan-AI/Wan2.1-T2V-1.3B](https://modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B)||[code](./model_inference/Wan2.1-T2V-1.3B.py)|[code](./model_training/full/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_full/Wan2.1-T2V-1.3B.py)|[code](./model_training/lora/Wan2.1-T2V-1.3B.sh)|[code](./model_training/validate_lora/Wan2.1-T2V-1.3B.py)|

From 86e509ad31c05b5ba6c3e2dffbb0678579f82221 Mon Sep 17 00:00:00 2001
From: lzws <lzw478614@alibaba-inc.com>
Date: Tue, 26 Aug 2025 17:22:41 +0800
Subject: [PATCH 10/10] update wan2.2-fun training scripts

---
 .../validate_full/Wan2.2-Fun-A14B-Control-Camera.py    | 10 +++++-----
 .../validate_lora/Wan2.2-Fun-A14B-Control-Camera.py    | 10 +++++-----
 2 files changed, 10 insertions(+), 10 deletions(-)

diff --git a/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py
index 5c3a3e4..700d503 100644
--- a/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py
+++ b/examples/wanvideo/model_training/validate_full/Wan2.2-Fun-A14B-Control-Camera.py
@@ -9,10 +9,10 @@ pipe = WanVideoPipeline.from_pretrained(
     torch_dtype=torch.bfloat16,
     device="cuda",
     model_configs=[
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
     ],
 )
 state_dict = load_state_dict("models/train/Wan2.2-Fun-A14B-Control-Camera_high_noise_full/epoch-1.safetensors")
@@ -31,4 +31,4 @@ video = pipe(
     camera_control_direction="Left", camera_control_speed=0.0,
     seed=0, tiled=True
 )
-save_video(video, "video_Wan2.2-Fun-A14B-Comtrol-Camera.mp4", fps=15, quality=5)
+save_video(video, "video_Wan2.2-Fun-A14B-Control-Camera.mp4", fps=15, quality=5)
diff --git a/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py
index 85dd973..f1e1840 100644
--- a/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py
+++ b/examples/wanvideo/model_training/validate_lora/Wan2.2-Fun-A14B-Control-Camera.py
@@ -9,10 +9,10 @@ pipe = WanVideoPipeline.from_pretrained(
     torch_dtype=torch.bfloat16,
     device="cuda",
     model_configs=[
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
-        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Comtrol-Camera", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.2-Fun-A14B-Control-Camera", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
     ],
 )
 pipe.load_lora(pipe.dit, "models/train/Wan2.2-Fun-A14B-Control-Camera_high_noise_lora/epoch-4.safetensors", alpha=1)
@@ -29,4 +29,4 @@ video = pipe(
     camera_control_direction="Left", camera_control_speed=0.0,
     seed=0, tiled=True
 )
-save_video(video, "video_Wan2.2-Fun-A14B-Comtrol-Camera.mp4", fps=15, quality=5)
+save_video(video, "video_Wan2.2-Fun-A14B-Control-Camera.mp4", fps=15, quality=5)