fix

app.py

@@ -82,21 +82,21 @@ def create_args(
     parser.add_argument(
         "--wan_model_dir",
         type=str,
-        default="./models/Wan2.1-I2V-14B-720P",
+        default="/data/models/Wan2.1-I2V-14B-720P",
         required=False,
         help="The dir of the Wan I2V 14B model.",
     )
     parser.add_argument(
         "--fantasytalking_model_path",
         type=str,
-        default="./models/fantasytalking_model.ckpt",
+        default="/data/models/fantasytalking_model.ckpt",
         required=False,
         help="The .ckpt path of fantasytalking model.",
     )
     parser.add_argument(
         "--wav2vec_model_dir",
         type=str,
-        default="./models/wav2vec2-base-960h",
+        default="/data/models/wav2vec2-base-960h",
         required=False,
         help="The dir of wav2vec model.",
     )
@@ -291,8 +291,8 @@ with gr.Blocks(title="FantasyTalking Video Generation") as demo:
             gr.Examples(
                 examples=[
                     [
-                        "assets/images/woman.png",
-                        "assets/audios/woman.wav",
+                        "./assets/images/woman.png",
+                        "./assets/audios/woman.wav",
                     ],
                 ],
                 inputs=[image_input, audio_input],

infer.py

@@ -11,8 +11,8 @@ from model import FantasyTalkingAudioConditionModel
 from utils import save_video, get_audio_features, resize_image_by_longest_edge
 from pathlib import Path
 from datetime import datetime
-from modelscope import snapshot_download
-
+# from modelscope import snapshot_download
+from huggingface_hub import snapshot_download
 
 def parse_args():
     parser = argparse.ArgumentParser(description="Simple example of a training script.")
@@ -125,9 +125,9 @@ def parse_args():
 def load_models(args):
     # Load Wan I2V models
 
-    snapshot_download("Wan-AI/Wan2.1-I2V-14B-720P", local_dir="models/Wan2.1-I2V-14B-720P")
-    snapshot_download("facebook/wav2vec2-base-960h", local_dir="models/wav2vec2-base-960h")
-    snapshot_download("acvlab/FantasyTalking", local_dir="models")
+    snapshot_download("Wan-AI/Wan2.1-I2V-14B-720P", local_dir="/data/models/Wan2.1-I2V-14B-720P")
+    snapshot_download("facebook/wav2vec2-base-960h", local_dir="/data/models/wav2vec2-base-960h")
+    snapshot_download("acvlab/FantasyTalking", local_dir="/data/models")
 
 
     model_manager = ModelManager(device="cpu")