new wan trainer

examples/wanvideo/model_training/full/Wan2.1-1.3b-speedcontrol-v1.sh

@@ -0,0 +1,13 @@
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_motion_bucket_id.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "Wan-AI/Wan2.1-T2V-1.3B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-T2V-1.3B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-T2V-1.3B:Wan2.1_VAE.pth,DiffSynth-Studio/Wan2.1-1.3b-speedcontrol-v1:model.safetensors" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.motion_controller." \
+  --output_path "./models/train/Wan2.1-1.3b-speedcontrol-v1_full" \
+  --trainable_models "motion_controller" \
+  --input_contains_motion_bucket_id

examples/wanvideo/model_training/full/Wan2.1-FLF2V-14B-720P.sh

@@ -0,0 +1,14 @@
+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 "Wan-AI/Wan2.1-FLF2V-14B-720P:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-FLF2V-14B-720P:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-FLF2V-14B-720P:Wan2.1_VAE.pth,Wan-AI/Wan2.1-FLF2V-14B-720P:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-FLF2V-14B-720P_full" \
+  --trainable_models "dit" \
+  --input_contains_input_image \
+  --input_contains_end_image

examples/wanvideo/model_training/full/Wan2.1-Fun-1.3B-Control.sh

@@ -0,0 +1,14 @@
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_control.csv \
+  --data_file_keys "video,control_video" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.1-Fun-1.3B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-1.3B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-1.3B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-1.3B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-1.3B-Control_full" \
+  --trainable_models "dit" \
+  --input_contains_control_video

examples/wanvideo/model_training/full/Wan2.1-Fun-1.3B-InP.sh

@@ -0,0 +1,14 @@
+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/Wan2.1-Fun-1.3B-InP:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-1.3B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-1.3B-InP:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-1.3B-InP:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-1.3B-InP_full" \
+  --trainable_models "dit" \
+  --input_contains_input_image \
+  --input_contains_end_image

examples/wanvideo/model_training/full/Wan2.1-Fun-14B-Control.sh

@@ -0,0 +1,14 @@
+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_control.csv \
+  --data_file_keys "video,control_video" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.1-Fun-14B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-14B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-14B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-14B-Control_full" \
+  --trainable_models "dit" \
+  --input_contains_control_video

examples/wanvideo/model_training/full/Wan2.1-Fun-14B-InP.sh

@@ -0,0 +1,14 @@
+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/Wan2.1-Fun-14B-InP:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-14B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-14B-InP:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-14B-InP:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-14B-InP_full" \
+  --trainable_models "dit" \
+  --input_contains_input_image \
+  --input_contains_end_image

examples/wanvideo/model_training/full/Wan2.1-Fun-V1.1-1.3B-Control.sh

@@ -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_reference_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.1-Fun-V1.1-1.3B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-V1.1-1.3B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-V1.1-1.3B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-V1.1-1.3B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-V1.1-1.3B-Control_full" \
+  --trainable_models "dit" \
+  --input_contains_control_video \
+  --input_contains_reference_image

examples/wanvideo/model_training/full/Wan2.1-Fun-V1.1-14B-Control.sh

@@ -0,0 +1,15 @@
+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/Wan2.1-Fun-V1.1-14B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-V1.1-14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-V1.1-14B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-V1.1-14B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-V1.1-14B-Control_full" \
+  --trainable_models "dit" \
+  --input_contains_control_video \
+  --input_contains_reference_image

examples/wanvideo/model_training/full/Wan2.1-I2V-14B-480P.sh

@@ -0,0 +1,13 @@
+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 "Wan-AI/Wan2.1-I2V-14B-480P:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-I2V-14B-480P:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-I2V-14B-480P:Wan2.1_VAE.pth,Wan-AI/Wan2.1-I2V-14B-480P:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-I2V-14B-480P_full" \
+  --trainable_models "dit" \
+  --input_contains_input_image

examples/wanvideo/model_training/full/Wan2.1-I2V-14B-720P.sh

@@ -0,0 +1,13 @@
+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 "Wan-AI/Wan2.1-I2V-14B-720P:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-I2V-14B-720P:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-I2V-14B-720P:Wan2.1_VAE.pth,Wan-AI/Wan2.1-I2V-14B-720P:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-I2V-14B-720P_full" \
+  --trainable_models "dit" \
+  --input_contains_input_image

examples/wanvideo/model_training/full/Wan2.1-T2V-1.3B.sh

@@ -0,0 +1,12 @@
+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 "Wan-AI/Wan2.1-T2V-1.3B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-T2V-1.3B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-T2V-1.3B:Wan2.1_VAE.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-T2V-1.3B_full" \
+  --trainable_models "dit"

examples/wanvideo/model_training/full/Wan2.1-T2V-14B.sh

@@ -0,0 +1,12 @@
+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 "Wan-AI/Wan2.1-T2V-14B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-T2V-14B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-T2V-14B:Wan2.1_VAE.pth" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-T2V-14B_full" \
+  --trainable_models "dit"

examples/wanvideo/model_training/full/accelerate_config_14B.yaml

@@ -0,0 +1,22 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+deepspeed_config:
+  gradient_accumulation_steps: 1
+  offload_optimizer_device: cpu
+  offload_param_device: cpu
+  zero3_init_flag: false
+  zero_stage: 2
+distributed_type: DEEPSPEED
+downcast_bf16: 'no'
+enable_cpu_affinity: false
+machine_rank: 0
+main_training_function: main
+mixed_precision: bf16
+num_machines: 1
+num_processes: 8
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

examples/wanvideo/model_training/full/run_test.py

@@ -0,0 +1,38 @@
+import multiprocessing, os
+
+
+def run_task(scripts, thread_id, thread_num):
+    for script_id, script in enumerate(scripts):
+        if script_id % thread_num == thread_id:
+            log_file_name = script.replace("/", "_") + ".txt"
+            cmd = f"CUDA_VISIBLE_DEVICES={thread_id} bash {script} > data/log/{log_file_name} 2>&1"
+            os.makedirs("data/log", exist_ok=True)
+            print(cmd, flush=True)
+            os.system(cmd)
+    
+
+if __name__ == "__main__":
+    # 1.3B
+    scripts = []
+    for file_name in os.listdir("examples/wanvideo/model_training/full"):
+        if file_name != "run_test.py" and "14B" not in file_name:
+            scripts.append(os.path.join("examples/wanvideo/model_training/full", file_name))
+
+    processes = [multiprocessing.Process(target=run_task, args=(scripts, i, 8)) for i in range(8)]
+    for p in processes:
+        p.start()
+    for p in processes:
+        p.join()
+    
+    # 14B
+    scripts = []
+    for file_name in os.listdir("examples/wanvideo/model_training/full"):
+        if file_name != "run_test.py" and "14B" in file_name:
+            scripts.append(os.path.join("examples/wanvideo/model_training/full", file_name))
+    for script in scripts:
+        log_file_name = script.replace("/", "_") + ".txt"
+        cmd = f"bash {script} > data/log/{log_file_name} 2>&1"
+        print(cmd, flush=True)
+        os.system(cmd)
+    
+    print("Done!")

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

@@ -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_motion_bucket_id.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "Wan-AI/Wan2.1-T2V-1.3B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-T2V-1.3B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-T2V-1.3B:Wan2.1_VAE.pth,DiffSynth-Studio/Wan2.1-1.3b-speedcontrol-v1:model.safetensors" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-1.3b-speedcontrol-v1_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --input_contains_motion_bucket_id

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

@@ -0,0 +1,16 @@
+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 "Wan-AI/Wan2.1-FLF2V-14B-720P:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-FLF2V-14B-720P:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-FLF2V-14B-720P:Wan2.1_VAE.pth,Wan-AI/Wan2.1-FLF2V-14B-720P:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-FLF2V-14B-720P_lora" \
+  --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

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

@@ -0,0 +1,16 @@
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_control.csv \
+  --data_file_keys "video,control_video" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.1-Fun-1.3B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-1.3B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-1.3B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-1.3B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-1.3B-Control_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --input_contains_control_video

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

@@ -0,0 +1,16 @@
+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/Wan2.1-Fun-1.3B-InP:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-1.3B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-1.3B-InP:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-1.3B-InP:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-1.3B-InP_lora" \
+  --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

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

@@ -0,0 +1,16 @@
+accelerate launch examples/wanvideo/model_training/train.py \
+  --dataset_base_path data/example_video_dataset \
+  --dataset_metadata_path data/example_video_dataset/metadata_control.csv \
+  --data_file_keys "video,control_video" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.1-Fun-14B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-14B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-14B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-14B-Control_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --input_contains_control_video

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

@@ -0,0 +1,16 @@
+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/Wan2.1-Fun-14B-InP:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-14B-InP:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-14B-InP:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-14B-InP:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-14B-InP_lora" \
+  --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

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

@@ -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_reference_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.1-Fun-V1.1-1.3B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-V1.1-1.3B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-V1.1-1.3B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-V1.1-1.3B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-V1.1-1.3B-Control_lora" \
+  --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

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

@@ -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_reference_control.csv \
+  --data_file_keys "video,control_video,reference_image" \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "PAI/Wan2.1-Fun-V1.1-14B-Control:diffusion_pytorch_model*.safetensors,PAI/Wan2.1-Fun-V1.1-14B-Control:models_t5_umt5-xxl-enc-bf16.pth,PAI/Wan2.1-Fun-V1.1-14B-Control:Wan2.1_VAE.pth,PAI/Wan2.1-Fun-V1.1-14B-Control:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-Fun-V1.1-14B-Control_lora" \
+  --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

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

@@ -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.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "Wan-AI/Wan2.1-I2V-14B-480P:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-I2V-14B-480P:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-I2V-14B-480P:Wan2.1_VAE.pth,Wan-AI/Wan2.1-I2V-14B-480P:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-I2V-14B-480P_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --input_contains_input_image

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

@@ -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.csv \
+  --height 480 \
+  --width 832 \
+  --dataset_repeat 100 \
+  --model_id_with_origin_paths "Wan-AI/Wan2.1-I2V-14B-720P:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-I2V-14B-720P:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-I2V-14B-720P:Wan2.1_VAE.pth,Wan-AI/Wan2.1-I2V-14B-720P:models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-I2V-14B-720P_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32 \
+  --input_contains_input_image

examples/wanvideo/model_training/lora/Wan2.1-T2V-1.3B.sh

@@ -0,0 +1,14 @@
+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 "Wan-AI/Wan2.1-T2V-1.3B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-T2V-1.3B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-T2V-1.3B:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-T2V-1.3B_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32

examples/wanvideo/model_training/lora/Wan2.1-T2V-14B.sh

@@ -0,0 +1,14 @@
+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 "Wan-AI/Wan2.1-T2V-14B:diffusion_pytorch_model*.safetensors,Wan-AI/Wan2.1-T2V-14B:models_t5_umt5-xxl-enc-bf16.pth,Wan-AI/Wan2.1-T2V-14B:Wan2.1_VAE.pth" \
+  --learning_rate 1e-4 \
+  --num_epochs 5 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Wan2.1-T2V-14B_lora" \
+  --lora_base_model "dit" \
+  --lora_target_modules "q,k,v,o,ffn.0,ffn.2" \
+  --lora_rank 32

examples/wanvideo/model_training/lora/run_test.py

@@ -0,0 +1,25 @@
+import multiprocessing, os
+
+
+def run_task(scripts, thread_id, thread_num):
+    for script_id, script in enumerate(scripts):
+        if script_id % thread_num == thread_id:
+            log_file_name = script.replace("/", "_") + ".txt"
+            cmd = f"CUDA_VISIBLE_DEVICES={thread_id} bash {script} > data/log/{log_file_name} 2>&1"
+            os.makedirs("data/log", exist_ok=True)
+            print(cmd, flush=True)
+            os.system(cmd)
+    
+
+if __name__ == "__main__":
+    scripts = []
+    for file_name in os.listdir("examples/wanvideo/model_training/lora"):
+        if file_name != "run_test.py":
+            scripts.append(os.path.join("examples/wanvideo/model_training/lora", file_name))
+
+    processes = [multiprocessing.Process(target=run_task, args=(scripts, i, 8)) for i in range(8)]
+    for p in processes:
+        p.start()
+    for p in processes:
+        p.join()
+    print("Done!")

examples/wanvideo/model_training/train.py

@@ -0,0 +1,129 @@
+import torch, os, json
+from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
+from diffsynth.trainers.utils import DiffusionTrainingModule, VideoDataset, launch_training_task, wan_parser
+os.environ["TOKENIZERS_PARALLELISM"] = "false"
+
+
+
+class WanTrainingModule(DiffusionTrainingModule):
+    def __init__(
+        self,
+        model_paths=None, model_id_with_origin_paths=None,
+        trainable_models=None,
+        lora_base_model=None, lora_target_modules="q,k,v,o,ffn.0,ffn.2", lora_rank=32,
+        use_gradient_checkpointing=True,
+        use_gradient_checkpointing_offload=False,
+        # Extra inputs
+        input_contains_input_image=False,
+        input_contains_end_image=False,
+        input_contains_control_video=False,
+        input_contains_reference_image=False,
+        input_contains_vace_video=False,
+        input_contains_vace_reference_image=False,
+        input_contains_motion_bucket_id=False,
+    ):
+        super().__init__()
+        # Load models
+        model_configs = []
+        if model_paths is not None:
+            model_paths = json.loads(model_paths)
+            model_configs += [ModelConfig(path=path) for path in model_paths]
+        if model_id_with_origin_paths is not None:
+            model_id_with_origin_paths = model_id_with_origin_paths.split(",")
+            model_configs += [ModelConfig(model_id=i.split(":")[0], origin_file_pattern=i.split(":")[1]) for i in model_id_with_origin_paths]
+        self.pipe = WanVideoPipeline.from_pretrained(torch_dtype=torch.bfloat16, device="cpu", model_configs=model_configs)
+        
+        # Reset training scheduler
+        self.pipe.scheduler.set_timesteps(1000, training=True)
+        
+        # Freeze untrainable models
+        self.pipe.freeze_except([] if trainable_models is None else trainable_models.split(","))
+        
+        # Add LoRA to the base models
+        if lora_base_model is not None:
+            model = self.add_lora_to_model(
+                getattr(self.pipe, lora_base_model),
+                target_modules=lora_target_modules.split(","),
+                lora_rank=lora_rank
+            )
+            setattr(self.pipe, lora_base_model, model)
+            
+        # Store other configs
+        self.use_gradient_checkpointing = use_gradient_checkpointing
+        self.use_gradient_checkpointing_offload = use_gradient_checkpointing_offload
+        self.input_contains_input_image = input_contains_input_image
+        self.input_contains_end_image = input_contains_end_image
+        self.input_contains_control_video = input_contains_control_video
+        self.input_contains_reference_image = input_contains_reference_image
+        self.input_contains_vace_video = input_contains_vace_video
+        self.input_contains_vace_reference_image = input_contains_vace_reference_image
+        self.input_contains_motion_bucket_id = input_contains_motion_bucket_id
+        
+        
+    def forward_preprocess(self, data):
+        # CFG-sensitive parameters
+        inputs_posi = {"prompt": data["prompt"]}
+        inputs_nega = {}
+        
+        # CFG-unsensitive parameters
+        inputs_shared = {
+            # Assume you are using this pipeline for inference,
+            # please fill in the input parameters.
+            "input_video": data["video"],
+            "height": data["video"][0].size[1],
+            "width": data["video"][0].size[0],
+            "num_frames": len(data["video"]),
+            # Please do not modify the following parameters
+            # unless you clearly know what this will cause.
+            "cfg_scale": 1,
+            "tiled": False,
+            "rand_device": self.pipe.device,
+            "use_gradient_checkpointing": self.use_gradient_checkpointing,
+            "use_gradient_checkpointing_offload": self.use_gradient_checkpointing_offload,
+            "cfg_merge": False,
+            "vace_scale": 1,
+        }
+        
+        # Extra inputs
+        if self.input_contains_input_image: inputs_shared["input_image"] = data["video"][0]
+        if self.input_contains_end_image: inputs_shared["end_image"] = data["video"][-1]
+        if self.input_contains_control_video: inputs_shared["control_video"] = data["control_video"]
+        if self.input_contains_reference_image: inputs_shared["reference_image"] = data["reference_image"]
+        if self.input_contains_vace_video: inputs_shared["vace_video"] = data["vace_video"]
+        if self.input_contains_vace_reference_image: inputs_shared["vace_reference_image"] = data["vace_reference_image"]
+        if self.input_contains_motion_bucket_id: inputs_shared["motion_bucket_id"] = data["motion_bucket_id"]
+        
+        # Pipeline units will automatically process the input parameters.
+        for unit in self.pipe.units:
+            inputs_shared, inputs_posi, inputs_nega = self.pipe.unit_runner(unit, self.pipe, inputs_shared, inputs_posi, inputs_nega)
+        return {**inputs_shared, **inputs_posi}
+    
+    
+    def forward(self, data, inputs=None):
+        if inputs is None: inputs = self.forward_preprocess(data)
+        models = {name: getattr(self.pipe, name) for name in self.pipe.in_iteration_models}
+        loss = self.pipe.training_loss(**models, **inputs)
+        return loss
+
+
+if __name__ == "__main__":
+    parser = wan_parser()
+    args = parser.parse_args()
+    dataset = VideoDataset(args=args)
+    model = WanTrainingModule(
+        model_paths=args.model_paths,
+        model_id_with_origin_paths=args.model_id_with_origin_paths,
+        trainable_models=args.trainable_models,
+        lora_base_model=args.lora_base_model,
+        lora_target_modules=args.lora_target_modules,
+        lora_rank=args.lora_rank,
+        use_gradient_checkpointing_offload=args.use_gradient_checkpointing_offload,
+        input_contains_input_image=args.input_contains_input_image,
+        input_contains_end_image=args.input_contains_end_image,
+        input_contains_control_video=args.input_contains_control_video,
+        input_contains_reference_image=args.input_contains_reference_image,
+        input_contains_vace_video=args.input_contains_vace_video,
+        input_contains_vace_reference_image=args.input_contains_vace_reference_image,
+        input_contains_motion_bucket_id=args.input_contains_motion_bucket_id,
+    )
+    launch_training_task(model, dataset, args=args)

examples/wanvideo/model_training/train_i2v.py

@@ -1,54 +0,0 @@
-import torch, os, json
-from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
-from diffsynth.trainers.utils import DiffusionTrainingModule, VideoDataset, launch_training_task, wan_parser
-os.environ["TOKENIZERS_PARALLELISM"] = "false"
-
-
-class WanTrainingModule(DiffusionTrainingModule):
-    def __init__(self, model_paths, task="train_lora", lora_target_modules="q,k,v,o,ffn.0,ffn.2", lora_rank=32):
-        super().__init__()
-        self.pipe = WanVideoPipeline.from_pretrained(
-            torch_dtype=torch.bfloat16,
-            device="cpu",
-            model_configs=[ModelConfig(path=path) for path in model_paths],
-        )
-        self.pipe.scheduler.set_timesteps(1000, training=True)
-        if task == "train_lora":
-            self.pipe.freeze_except([])
-            self.pipe.dit = self.add_lora_to_model(self.pipe.dit, target_modules=lora_target_modules.split(","), lora_rank=lora_rank)
-        else:
-            self.pipe.freeze_except(["dit"])
-        
-    def forward_preprocess(self, data):
-        inputs_posi = {"prompt": data["prompt"]}
-        inputs_nega = {}
-        inputs_shared = {
-            "input_image": data["video"][0],
-            "input_video": data["video"],
-            "height": data["video"][0].size[1],
-            "width": data["video"][0].size[0],
-            "num_frames": len(data["video"]),
-            # Please do not modify the following parameters.
-            "cfg_scale": 1,
-            "tiled": False,
-            "rand_device": self.pipe.device,
-            "use_gradient_checkpointing": True,
-            "cfg_merge": False,
-        }
-        for unit in self.pipe.units:
-            inputs_shared, inputs_posi, inputs_nega = self.pipe.unit_runner(unit, self.pipe, inputs_shared, inputs_posi, inputs_nega)
-        return {**inputs_shared, **inputs_posi}
-    
-    def forward(self, data):
-        inputs = self.forward_preprocess(data)
-        models = {name: getattr(self.pipe, name) for name in self.pipe.in_iteration_models}
-        loss = self.pipe.training_loss(**models, **inputs)
-        return loss
-
-
-if __name__ == "__main__":
-    parser = wan_parser()
-    args = parser.parse_args()
-    dataset = VideoDataset(args=args)
-    model = WanTrainingModule(json.loads(args.model_paths), task=args.task, lora_target_modules=args.lora_target_modules, lora_rank=args.lora_rank)
-    launch_training_task(model, dataset, args=args)

examples/wanvideo/model_training/train_t2v.py

@@ -1,53 +0,0 @@
-import torch, os, json
-from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
-from diffsynth.trainers.utils import DiffusionTrainingModule, VideoDataset, launch_training_task, wan_parser
-os.environ["TOKENIZERS_PARALLELISM"] = "false"
-
-
-class WanTrainingModule(DiffusionTrainingModule):
-    def __init__(self, model_paths, task="train_lora", lora_target_modules="q,k,v,o,ffn.0,ffn.2", lora_rank=32):
-        super().__init__()
-        self.pipe = WanVideoPipeline.from_pretrained(
-            torch_dtype=torch.bfloat16,
-            device="cpu",
-            model_configs=[ModelConfig(path=path) for path in model_paths],
-        )
-        self.pipe.scheduler.set_timesteps(1000, training=True)
-        if task == "train_lora":
-            self.pipe.freeze_except([])
-            self.pipe.dit = self.add_lora_to_model(self.pipe.dit, target_modules=lora_target_modules.split(","), lora_rank=lora_rank)
-        else:
-            self.pipe.freeze_except(["dit"])
-        
-    def forward_preprocess(self, data):
-        inputs_posi = {"prompt": data["prompt"]}
-        inputs_nega = {}
-        inputs_shared = {
-            "input_video": data["video"],
-            "height": data["video"][0].size[1],
-            "width": data["video"][0].size[0],
-            "num_frames": len(data["video"]),
-            # Please do not modify the following parameters.
-            "cfg_scale": 1,
-            "tiled": False,
-            "rand_device": self.pipe.device,
-            "use_gradient_checkpointing": True,
-            "cfg_merge": False,
-        }
-        for unit in self.pipe.units:
-            inputs_shared, inputs_posi, inputs_nega = self.pipe.unit_runner(unit, self.pipe, inputs_shared, inputs_posi, inputs_nega)
-        return {**inputs_shared, **inputs_posi}
-    
-    def forward(self, data):
-        inputs = self.forward_preprocess(data)
-        models = {name: getattr(self.pipe, name) for name in self.pipe.in_iteration_models}
-        loss = self.pipe.training_loss(**models, **inputs)
-        return loss
-
-
-if __name__ == "__main__":
-    parser = wan_parser()
-    args = parser.parse_args()
-    dataset = VideoDataset(args=args)
-    model = WanTrainingModule(json.loads(args.model_paths), task=args.task, lora_target_modules=args.lora_target_modules, lora_rank=args.lora_rank)
-    launch_training_task(model, dataset, args=args)

examples/wanvideo/model_training/validate_full/Wan2.1-1.3b-speedcontrol-v1.py

@@ -0,0 +1,28 @@
+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-T2V-1.3B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="DiffSynth-Studio/Wan2.1-1.3b-speedcontrol-v1", origin_file_pattern="model.safetensors", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-1.3b-speedcontrol-v1_full/epoch-1.safetensors")
+pipe.motion_controller.load_state_dict(state_dict)
+pipe.enable_vram_management()
+
+# Text-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",
+    seed=1, tiled=True,
+    motion_bucket_id=50
+)
+save_video(video, "video_Wan2.1-1.3b-speedcontrol-v1.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-FLF2V-14B-720P.py

@@ -0,0 +1,33 @@
+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="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-FLF2V-14B-720P_full/epoch-1.safetensors")
+pipe.dit.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,
+    sigma_shift=16,
+)
+save_video(video, "video_Wan2.1-FLF2V-14B-720P.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-Fun-1.3B-Control.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/Wan2.1-Fun-1.3B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-Fun-1.3B-Control_full/epoch-1.safetensors")
+pipe.dit.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)]
+
+# 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,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-Fun-1.3B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-Fun-1.3B-InP.py

@@ -0,0 +1,31 @@
+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/Wan2.1-Fun-1.3B-InP", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-InP", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-InP", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-InP", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-Fun-1.3B-InP_full/epoch-1.safetensors")
+pipe.dit.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_Wan2.1-Fun-1.3B-InP.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-Fun-14B-Control.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/Wan2.1-Fun-14B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-Fun-14B-Control_full/epoch-1.safetensors")
+pipe.dit.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)]
+
+# 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,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-Fun-14B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-Fun-14B-InP.py

@@ -0,0 +1,31 @@
+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/Wan2.1-Fun-14B-InP", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-InP", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-InP", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-InP", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-Fun-14B-InP_full/epoch-1.safetensors")
+pipe.dit.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_Wan2.1-Fun-14B-InP.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-Fun-V1.1-1.3B-Control.py

@@ -0,0 +1,33 @@
+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/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-Fun-V1.1-1.3B-Control_full/epoch-1.safetensors")
+pipe.dit.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_Wan2.1-Fun-V1.1-1.3B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-Fun-V1.1-14B-Control.py

@@ -0,0 +1,33 @@
+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/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-Fun-V1.1-14B-Control_full/epoch-1.safetensors")
+pipe.dit.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_Wan2.1-Fun-V1.1-14B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-I2V-14B-480P.py

@@ -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
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-I2V-14B-480P_full/epoch-1.safetensors")
+pipe.dit.load_state_dict(state_dict)
+pipe.enable_vram_management()
+
+input_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",
+    input_image=input_image,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-I2V-14B-480P.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-I2V-14B-720P.py

@@ -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
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-I2V-14B-720P_full/epoch-1.safetensors")
+pipe.dit.load_state_dict(state_dict)
+pipe.enable_vram_management()
+
+input_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",
+    input_image=input_image,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-I2V-14B-720P.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-T2V-1.3B.py

@@ -0,0 +1,25 @@
+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-T2V-1.3B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-T2V-1.3B_full/epoch-1.safetensors")
+pipe.dit.load_state_dict(state_dict)
+pipe.enable_vram_management()
+
+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",
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-T2V-1.3B.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/Wan2.1-T2V-14B.py

@@ -0,0 +1,25 @@
+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-T2V-14B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-14B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-14B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+state_dict = load_state_dict("models/train/Wan2.1-T2V-14B_full/epoch-1.safetensors")
+pipe.dit.load_state_dict(state_dict)
+pipe.enable_vram_management()
+
+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",
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-T2V-14B.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_full/run_test.py

@@ -0,0 +1,25 @@
+import multiprocessing, os
+
+
+def run_task(scripts, thread_id, thread_num):
+    for script_id, script in enumerate(scripts):
+        if script_id % thread_num == thread_id:
+            log_file_name = script.replace("/", "_") + ".txt"
+            cmd = f"CUDA_VISIBLE_DEVICES={thread_id} python -u {script} > data/log/{log_file_name} 2>&1"
+            os.makedirs("data/log", exist_ok=True)
+            print(cmd, flush=True)
+            os.system(cmd)
+    
+
+if __name__ == "__main__":
+    scripts = []
+    for file_name in os.listdir("examples/wanvideo/model_training/validate_full"):
+        if file_name != "run_test.py":
+            scripts.append(os.path.join("examples/wanvideo/model_training/validate_full", file_name))
+
+    processes = [multiprocessing.Process(target=run_task, args=(scripts, i, 8)) for i in range(8)]
+    for p in processes:
+        p.start()
+    for p in processes:
+        p.join()
+    print("Done!")

examples/wanvideo/model_training/validate_lora/Wan2.1-1.3b-speedcontrol-v1.py

@@ -0,0 +1,27 @@
+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-T2V-1.3B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="DiffSynth-Studio/Wan2.1-1.3b-speedcontrol-v1", origin_file_pattern="model.safetensors", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-1.3b-speedcontrol-v1_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+# Text-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",
+    seed=1, tiled=True,
+    motion_bucket_id=50
+)
+save_video(video, "video_Wan2.1-1.3b-speedcontrol-v1.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-FLF2V-14B-720P.py

@@ -0,0 +1,32 @@
+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="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-FLF2V-14B-720P", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-FLF2V-14B-720P_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,
+    sigma_shift=16,
+)
+save_video(video, "video_Wan2.1-FLF2V-14B-720P.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-Fun-1.3B-Control.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/Wan2.1-Fun-1.3B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-Fun-1.3B-Control_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)]
+
+# 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,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-Fun-1.3B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-Fun-1.3B-InP.py

@@ -0,0 +1,30 @@
+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/Wan2.1-Fun-1.3B-InP", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-InP", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-InP", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-1.3B-InP", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-Fun-1.3B-InP_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_Wan2.1-Fun-1.3B-InP.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-Fun-14B-Control.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/Wan2.1-Fun-14B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-Fun-14B-Control_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)]
+
+# 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,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-Fun-14B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-Fun-14B-InP.py

@@ -0,0 +1,30 @@
+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/Wan2.1-Fun-14B-InP", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-InP", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-InP", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-14B-InP", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-Fun-14B-InP_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_Wan2.1-Fun-14B-InP.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-Fun-V1.1-1.3B-Control.py

@@ -0,0 +1,32 @@
+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/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-1.3B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-Fun-V1.1-1.3B-Control_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_Wan2.1-Fun-V1.1-1.3B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-Fun-V1.1-14B-Control.py

@@ -0,0 +1,32 @@
+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/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="PAI/Wan2.1-Fun-V1.1-14B-Control", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-Fun-V1.1-14B-Control_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_Wan2.1-Fun-V1.1-14B-Control.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-I2V-14B-480P.py

@@ -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
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-480P", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-I2V-14B-480P_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+input_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",
+    input_image=input_image,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-I2V-14B-480P.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-I2V-14B-720P.py

@@ -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
+from modelscope import dataset_snapshot_download
+
+
+pipe = WanVideoPipeline.from_pretrained(
+    torch_dtype=torch.bfloat16,
+    device="cuda",
+    model_configs=[
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-I2V-14B-720P", origin_file_pattern="models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-I2V-14B-720P_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+input_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",
+    input_image=input_image,
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-I2V-14B-720P.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-1.3B.py

@@ -0,0 +1,24 @@
+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-T2V-1.3B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-1.3B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-T2V-1.3B_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+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",
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-T2V-1.3B.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/Wan2.1-T2V-14B.py

@@ -0,0 +1,24 @@
+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-T2V-14B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-14B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
+        ModelConfig(model_id="Wan-AI/Wan2.1-T2V-14B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
+    ],
+)
+pipe.load_lora(pipe.dit, "models/train/Wan2.1-T2V-14B_lora/epoch-4.safetensors", alpha=1)
+pipe.enable_vram_management()
+
+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",
+    seed=1, tiled=True
+)
+save_video(video, "video_Wan2.1-T2V-14B.mp4", fps=15, quality=5)

examples/wanvideo/model_training/validate_lora/run_test.py

@@ -0,0 +1,25 @@
+import multiprocessing, os
+
+
+def run_task(scripts, thread_id, thread_num):
+    for script_id, script in enumerate(scripts):
+        if script_id % thread_num == thread_id:
+            log_file_name = script.replace("/", "_") + ".txt"
+            cmd = f"CUDA_VISIBLE_DEVICES={thread_id} python -u {script} > data/log/{log_file_name} 2>&1"
+            os.makedirs("data/log", exist_ok=True)
+            print(cmd, flush=True)
+            os.system(cmd)
+    
+
+if __name__ == "__main__":
+    scripts = []
+    for file_name in os.listdir("examples/wanvideo/model_training/validate_lora"):
+        if file_name != "run_test.py":
+            scripts.append(os.path.join("examples/wanvideo/model_training/validate_lora", file_name))
+
+    processes = [multiprocessing.Process(target=run_task, args=(scripts, i, 8)) for i in range(8)]
+    for p in processes:
+        p.start()
+    for p in processes:
+        p.join()
+    print("Done!")