theluyuan/DiffSynth-Studio
1
0
Fork 0
mirror of https://github.com/modelscope/DiffSynth-Studio.git synced 2026-03-24 01:48:13 +00:00
Code Issues Packages Projects Releases Wiki Activity

VACE

Browse Source
This commit is contained in:
Artiprocher
2025-06-11 18:48:44 +08:00
parent 7d29ee1fbb
commit 6a833c7134
15 changed files with 332 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
examples/wanvideo/model_training/full/Wan2.1-VACE-1.3B-Preview.sh Normal file
View File

@@ -0,0 +1,17 @@
accelerate launch examples/wanvideo/model_training/train.py \
--dataset_base_path data/example_video_dataset \
--dataset_metadata_path data/example_video_dataset/metadata_vace.csv \
--data_file_keys "video,vace_video,vace_reference_image" \
--height 480 \
--width 832 \
--num_frames 49 \
--dataset_repeat 100 \
--model_id_with_origin_paths "iic/VACE-Wan2.1-1.3B-Preview:diffusion_pytorch_model*.safetensors,iic/VACE-Wan2.1-1.3B-Preview:models_t5_umt5-xxl-enc-bf16.pth,iic/VACE-Wan2.1-1.3B-Preview:Wan2.1_VAE.pth" \
--learning_rate 1e-4 \
--num_epochs 2 \
--remove_prefix_in_ckpt "pipe.vace." \
--output_path "./models/train/Wan2.1-VACE-1.3B-Preview_full" \
--trainable_models "vace" \
--input_contains_vace_video \
--input_contains_vace_reference_image \
--use_gradient_checkpointing_offload

17
examples/wanvideo/model_training/full/Wan2.1-VACE-1.3B.sh Normal file
View File

@@ -0,0 +1,17 @@
accelerate launch examples/wanvideo/model_training/train.py \
--dataset_base_path data/example_video_dataset \
--dataset_metadata_path data/example_video_dataset/metadata_vace.csv \
--data_file_keys "video,vace_video,vace_reference_image" \
--height 480 \
--width 832 \
--num_frames 49 \
--dataset_repeat 100 \
--model_id_with_origin_paths "Wan-AI/Wan2.1-VACE-1.3B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-VACE-1.3B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-VACE-1.3B:Wan2.1_VAE.pth" \
--learning_rate 1e-4 \
--num_epochs 2 \
--remove_prefix_in_ckpt "pipe.vace." \
--output_path "./models/train/Wan2.1-VACE-1.3B_full" \
--trainable_models "vace" \
--input_contains_vace_video \
--input_contains_vace_reference_image \
--use_gradient_checkpointing_offload

17
examples/wanvideo/model_training/full/Wan2.1-VACE-14B.sh Normal file
View File

@@ -0,0 +1,17 @@
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_vace.csv \
--data_file_keys "video,vace_video,vace_reference_image" \
--height 480 \
--width 832 \
--num_frames 17 \
--dataset_repeat 100 \
--model_id_with_origin_paths "Wan-AI/Wan2.1-VACE-14B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-VACE-14B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-VACE-14B:Wan2.1_VAE.pth" \
--learning_rate 1e-4 \
--num_epochs 2 \
--remove_prefix_in_ckpt "pipe.vace." \
--output_path "./models/train/Wan2.1-VACE-14B_full" \
--trainable_models "vace" \
--input_contains_vace_video \
--input_contains_vace_reference_image \
--use_gradient_checkpointing_offload

18
examples/wanvideo/model_training/lora/Wan2.1-VACE-1.3B-Preview.sh Normal file
View File

@@ -0,0 +1,18 @@
accelerate launch examples/wanvideo/model_training/train.py \
--dataset_base_path data/example_video_dataset \
--dataset_metadata_path data/example_video_dataset/metadata_vace.csv \
--data_file_keys "video,vace_video,vace_reference_image" \
--height 480 \
--width 832 \
--dataset_repeat 100 \
--model_id_with_origin_paths "iic/VACE-Wan2.1-1.3B-Preview:diffusion_pytorch_model*.safetensors,iic/VACE-Wan2.1-1.3B-Preview:models_t5_umt5-xxl-enc-bf16.pth,iic/VACE-Wan2.1-1.3B-Preview:Wan2.1_VAE.pth" \
--learning_rate 1e-4 \
--num_epochs 5 \
--remove_prefix_in_ckpt "pipe.vace." \
--output_path "./models/train/Wan2.1-VACE-1.3B-Preview_lora" \
--lora_base_model "vace" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_vace_video \
--input_contains_vace_reference_image \
--use_gradient_checkpointing_offload

18
examples/wanvideo/model_training/lora/Wan2.1-VACE-1.3B.sh Normal file
View File

@@ -0,0 +1,18 @@
accelerate launch examples/wanvideo/model_training/train.py \
--dataset_base_path data/example_video_dataset \
--dataset_metadata_path data/example_video_dataset/metadata_vace.csv \
--data_file_keys "video,vace_video,vace_reference_image" \
--height 480 \
--width 832 \
--dataset_repeat 100 \
--model_id_with_origin_paths "Wan-AI/Wan2.1-VACE-1.3B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-VACE-1.3B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-VACE-1.3B:Wan2.1_VAE.pth" \
--learning_rate 1e-4 \
--num_epochs 5 \
--remove_prefix_in_ckpt "pipe.vace." \
--output_path "./models/train/Wan2.1-VACE-1.3B_lora" \
--lora_base_model "vace" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_vace_video \
--input_contains_vace_reference_image \
--use_gradient_checkpointing_offload

19
examples/wanvideo/model_training/lora/Wan2.1-VACE-14B.sh Normal file
View File

@@ -0,0 +1,19 @@
accelerate launch examples/wanvideo/model_training/train.py \
--dataset_base_path data/example_video_dataset \
--dataset_metadata_path data/example_video_dataset/metadata_vace.csv \
--data_file_keys "video,vace_video,vace_reference_image" \
--height 480 \
--width 832 \
--num_frames 17 \
--dataset_repeat 100 \
--model_id_with_origin_paths "Wan-AI/Wan2.1-VACE-14B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-VACE-14B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-VACE-14B:Wan2.1_VAE.pth" \
--learning_rate 1e-4 \
--num_epochs 5 \
--remove_prefix_in_ckpt "pipe.vace." \
--output_path "./models/train/Wan2.1-VACE-14B_lora" \
--lora_base_model "vace" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_vace_video \
--input_contains_vace_reference_image \
--use_gradient_checkpointing_offload

30
examples/wanvideo/model_training/validate_full/Wan2.1-VACE-1.3B-Preview.py Normal file
View File

@@ -0,0 +1,30 @@
import torch
from PIL import Image
from diffsynth import save_video, VideoData, load_state_dict
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="iic/VACE-Wan2.1-1.3B-Preview", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="iic/VACE-Wan2.1-1.3B-Preview", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="iic/VACE-Wan2.1-1.3B-Preview", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
state_dict = load_state_dict("models/train/VACE-Wan2.1-1.3B-Preview_full/epoch-1.safetensors")
pipe.vace.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(49)]
reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
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",
vace_video=video, vace_reference_image=reference_image, num_frames=49,
seed=1, tiled=True
)
save_video(video, "video_Wan2.1-VACE-1.3B-Preview.mp4", fps=15, quality=5)

30
examples/wanvideo/model_training/validate_full/Wan2.1-VACE-1.3B.py Normal file
View File

@@ -0,0 +1,30 @@
import torch
from PIL import Image
from diffsynth import save_video, VideoData, load_state_dict
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-1.3B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-1.3B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-1.3B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
state_dict = load_state_dict("models/train/Wan2.1-VACE-1.3B_full/epoch-1.safetensors")
pipe.vace.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(49)]
reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
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",
vace_video=video, vace_reference_image=reference_image, num_frames=49,
seed=1, tiled=True
)
save_video(video, "video_Wan2.1-VACE-1.3B.mp4", fps=15, quality=5)

30
examples/wanvideo/model_training/validate_full/Wan2.1-VACE-14B.py Normal file
View File

@@ -0,0 +1,30 @@
import torch
from PIL import Image
from diffsynth import save_video, VideoData, load_state_dict
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-14B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-14B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-14B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
state_dict = load_state_dict("models/train/Wan2.1-VACE-14B_full/epoch-1.safetensors")
pipe.vace.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(17)]
reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
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",
vace_video=video, vace_reference_image=reference_image, num_frames=17,
seed=1, tiled=True
)
save_video(video, "video_Wan2.1-VACE-14B.mp4", fps=15, quality=5)

29
examples/wanvideo/model_training/validate_lora/Wan2.1-VACE-1.3B-Preview.py Normal file
View File

@@ -0,0 +1,29 @@
import torch
from PIL import Image
from diffsynth import save_video, VideoData
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="iic/VACE-Wan2.1-1.3B-Preview", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="iic/VACE-Wan2.1-1.3B-Preview", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="iic/VACE-Wan2.1-1.3B-Preview", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
pipe.load_lora(pipe.vace, "models/train/Wan2.1-VACE-1.3B-Preview_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(49)]
reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
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",
vace_video=video, vace_reference_image=reference_image, num_frames=49,
seed=1, tiled=True
)
save_video(video, "video_Wan2.1-VACE-1.3B-Preview.mp4", fps=15, quality=5)

29
examples/wanvideo/model_training/validate_lora/Wan2.1-VACE-1.3B.py Normal file
View File

@@ -0,0 +1,29 @@
import torch
from PIL import Image
from diffsynth import save_video, VideoData
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-1.3B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-1.3B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-1.3B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
pipe.load_lora(pipe.vace, "models/train/Wan2.1-VACE-1.3B_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(49)]
reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
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",
vace_video=video, vace_reference_image=reference_image, num_frames=49,
seed=1, tiled=True
)
save_video(video, "video_Wan2.1-VACE-1.3B.mp4", fps=15, quality=5)

29
examples/wanvideo/model_training/validate_lora/Wan2.1-VACE-14B.py Normal file
View File

@@ -0,0 +1,29 @@
import torch
from PIL import Image
from diffsynth import save_video, VideoData
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-14B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-14B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.1-VACE-14B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
pipe.load_lora(pipe.vace, "models/train/Wan2.1-VACE-14B_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(17)]
reference_image = VideoData("data/example_video_dataset/video1.mp4", height=480, width=832)[0]
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",
vace_video=video, vace_reference_image=reference_image, num_frames=17,
seed=1, tiled=True
)
save_video(video, "video_Wan2.1-VACE-14B.mp4", fps=15, quality=5)