theluyuan/DiffSynth-Studio
1
0
Fork 0
mirror of https://github.com/modelscope/DiffSynth-Studio.git synced 2026-03-24 10:18:12 +00:00
Code Issues Packages Projects Releases Wiki Activity

wan-refactor

Browse Source
This commit is contained in:
Artiprocher
2025-06-13 13:04:35 +08:00
parent 7e6a3c7897
commit 8dd24169cc
47 changed files with 810 additions and 310 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/wanvideo/model_training/lora/Wan2.1-1.3b-speedcontrol-v1.sh
View File

@@ -12,4 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_motion_bucket_id
--extra_inputs "motion_bucket_id"

3
examples/wanvideo/model_training/lora/Wan2.1-FLF2V-14B-720P.sh
View File

@@ -12,5 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_input_image \
--input_contains_end_image
--extra_inputs "input_image,end_image"

2
examples/wanvideo/model_training/lora/Wan2.1-Fun-1.3B-Control.sh
View File

@@ -13,4 +13,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_control_video
--extra_inputs "control_video"

3
examples/wanvideo/model_training/lora/Wan2.1-Fun-1.3B-InP.sh
View File

@@ -12,5 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_input_image \
--input_contains_end_image
--extra_inputs "input_image,end_image"

2
examples/wanvideo/model_training/lora/Wan2.1-Fun-14B-Control.sh
View File

@@ -13,4 +13,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_control_video
--extra_inputs "control_video"

3
examples/wanvideo/model_training/lora/Wan2.1-Fun-14B-InP.sh
View File

@@ -12,5 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_input_image \
--input_contains_end_image
--extra_inputs "input_image,end_image"

15
examples/wanvideo/model_training/lora/Wan2.1-Fun-V1.1-1.3B-Control-Camera.sh Normal file
View File

@@ -0,0 +1,15 @@
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 \
--height 480 \
--width 832 \
--dataset_repeat 100 \
--model_id_with_origin_paths "PAI/Wan2.1-Fun-V1.1-1.3B-Control-Camera:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-V1.1-1.3B-Control-Camera:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-V1.1-1.3B-Control-Camera:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-V1.1-1.3B-Control-Camera:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
--learning_rate 1e-5 \
--num_epochs 5 \
--remove_prefix_in_ckpt "pipe.dit." \
--output_path "./models/train/Wan2.1-Fun-V1.1-1.3B-Control-Camera_full" \
--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"

3
examples/wanvideo/model_training/lora/Wan2.1-Fun-V1.1-1.3B-Control.sh
View File

@@ -13,5 +13,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_control_video \
--input_contains_reference_image
--extra_inputs "control_video,reference_image"

3
examples/wanvideo/model_training/lora/Wan2.1-Fun-V1.1-1.3B-InP.sh
View File

@@ -12,5 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_input_image \
--input_contains_end_image
--extra_inputs "input_image,end_image"

15
examples/wanvideo/model_training/lora/Wan2.1-Fun-V1.1-14B-Control-Camera.sh Normal file
View File

@@ -0,0 +1,15 @@
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 \
--height 480 \
--width 832 \
--dataset_repeat 100 \
--model_id_with_origin_paths "PAI/Wan2.1-Fun-V1.1-14B-Control-Camera:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-V1.1-14B-Control-Camera:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-V1.1-14B-Control-Camera:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-V1.1-14B-Control-Camera:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
--learning_rate 1e-5 \
--num_epochs 5 \
--remove_prefix_in_ckpt "pipe.dit." \
--output_path "./models/train/Wan2.1-Fun-V1.1-14B-Control-Camera_full" \
--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"

3
examples/wanvideo/model_training/lora/Wan2.1-Fun-V1.1-14B-Control.sh
View File

@@ -13,5 +13,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_control_video \
--input_contains_reference_image
--extra_inputs "control_video,reference_image"

3
examples/wanvideo/model_training/lora/Wan2.1-Fun-V1.1-14B-InP.sh
View File

@@ -12,5 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_input_image \
--input_contains_end_image
--extra_inputs "input_image,end_image"

2
examples/wanvideo/model_training/lora/Wan2.1-I2V-14B-480P.sh
View File

@@ -12,4 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_input_image
--extra_inputs "input_image"

2
examples/wanvideo/model_training/lora/Wan2.1-I2V-14B-720P.sh
View File

@@ -12,4 +12,4 @@ accelerate launch examples/wanvideo/model_training/train.py \
--lora_base_model "dit" \
--lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
--lora_rank 32 \
--input_contains_input_image
--extra_inputs "input_image"

3
examples/wanvideo/model_training/lora/Wan2.1-VACE-1.3B-Preview.sh
View File

@@ -13,6 +13,5 @@ accelerate launch examples/wanvideo/model_training/train.py \
--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 \
--extra_inputs "vace_video,vace_reference_image" \
--use_gradient_checkpointing_offload

3
examples/wanvideo/model_training/lora/Wan2.1-VACE-1.3B.sh
View File

@@ -13,6 +13,5 @@ accelerate launch examples/wanvideo/model_training/train.py \
--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 \
--extra_inputs "vace_video,vace_reference_image" \
--use_gradient_checkpointing_offload

3
examples/wanvideo/model_training/lora/Wan2.1-VACE-14B.sh
View File

@@ -14,6 +14,5 @@ accelerate launch examples/wanvideo/model_training/train.py \
--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 \
--extra_inputs "vace_video,vace_reference_image" \
--use_gradient_checkpointing_offload