Number of predicted frames does NOT match the input.

[u-can-u-up-no-bb]

Thanks for your nice work.

I'm trying to run your code but found the numbers of frames do not match.

I have 100 images in a sequence, and I made it into an mp4 file.
When I run you code, there are 100 frames as input, and at the end I got 92 frames. I print out the array shapes.

video_data.shape = (100, 480, 640, 3)
depth_pred.shape = (92, 480, 640)

Here is my input video:
https://github.com/user-attachments/assets/d4b6a109-4fdd-4028-ad07-72d2892fcce3

This is the command I used (the same as that in README except input):

 python run_infer.py \
    --model_base=jhshao/ChronoDepth \
    --data_dir=assets/100-frame.mp4 \
    --output_dir=./outputs \
    --num_frames=10 \
    --denoise_steps=10 \
    --window_size=9 \
    --half_precision \
    --seed=1234 \

Could you help me get exactly the 100 frames that correspond to my input?

Thank you!