pushing

gradio_app.py

@@ -232,7 +232,7 @@ def run_epic_inference(video_path, fps, num_frames, target_pose, mode):
         "--video_length", str(num_frames),
         "--save_name", "output",
         "--mode", mode,
-        "--fps", fps
+        "--fps", str(fps)
     ]
     try:
         result = subprocess.run(command, capture_output=True, text=True, check=True)

inference/cli_demo_camera_i2v_pcd.py

@@ -160,6 +160,7 @@ def generate_video(
     infer_with_mask: bool = False,
     pool_style: str = 'avg',
     pipe_cpu_offload: bool = False,
+    fps: int = 8,
 ):
     """
     Generates a video based on the given prompt and saves it to the specified path.
@@ -216,6 +217,7 @@ def generate_video(
         downscale_coef=downscale_coef,
         in_channels=controlnet_input_channels,
         use_zero_conv=use_zero_conv,
+        sample_frames = num_frames,  # 49 frames
         **controlnet_kwargs,   
     )
     if controlnet_model_path:
@@ -348,8 +350,8 @@ def generate_video(
         video_generate = video_generate_all[0]
 
         # 6. Export the generated frames to a video file. fps must be 8 for original video.
-        export_to_video(video_generate, output_path_file, fps=8)
-        export_to_video(reference_frames, output_path_file_reference, fps=8)
+        export_to_video(video_generate, output_path_file, fps=fps)
+        export_to_video(reference_frames, output_path_file_reference, fps=fps)
         out_reference_frames = [
             stack_images_horizontally(frame_reference, frame_out)
             for frame_out, frame_reference in zip(video_generate, reference_frames)
@@ -360,7 +362,7 @@ def generate_video(
             stack_images_horizontally(frame_out, frame_reference)
             for frame_out, frame_reference in zip(out_reference_frames, anchor_video)
             ]
-        export_to_video(out_reference_frames, output_path_file_out_reference, fps=8)
+        export_to_video(out_reference_frames, output_path_file_out_reference, fps=fps)
 
 
 if __name__ == "__main__":
@@ -415,6 +417,7 @@ if __name__ == "__main__":
     parser.add_argument("--controlnet_input_channels", type=int, default=6)
     parser.add_argument("--controlnet_transformer_num_layers", type=int, default=8)
     parser.add_argument("--enable_model_cpu_offload", action="store_true", default=False, help="Enable model CPU offload")
+    parser.add_argument("--fps", type=int, default=8, help="Frames per second for the output video")
 
     args = parser.parse_args()
     dtype = torch.float16 if args.dtype == "float16" else torch.bfloat16
@@ -451,4 +454,5 @@ if __name__ == "__main__":
         infer_with_mask=args.infer_with_mask,
         pool_style=args.pool_style,
         pipe_cpu_offload=args.enable_model_cpu_offload,
+        fps=args.fps,
     )