how can use the vlogger project to get the teddy_travel vlog in your github like the example https://youtu.be/ZRD1-jHbEGk?

[pzwstudy]

No description provided.

[zhuangshaobin]

Firstly, run the following command to get script, actors and protagonist:
python sample_scripts/vlog_write_script.py
The generated scripts will be saved in results/vlog/$your_story_dir/script.
The generated reference images will be saved in results/vlog/$your_story_dir/img.
⚠️ Enter your openai key in the 7th line of the file vlogger/planning_utils/gpt4_utils.py

Secondly, run the following command to get the vlog:
python sample_scripts/vlog_read_script_sample.py
The generated scripts will be saved in results/vlog/$your_story_dir/video.

[pzwstudy]

thanks!
pil_image = Image.open(args.reference_image_path[reference_lists[i][0] - 1])results an error "list index out of range"

`File "C:\Users\HIT\Desktop\vlogger\Vlogger\sample_scripts\vlog_read_script_sample.py", line 190, in main
pil_image = Image.open(args.reference_image_path[reference_lists[i][0] - 1])
File "C:\Users\HIT\Desktop\vlogger\Vlogger\sample_scripts\vlog_read_script_sample.py", line 307, in
main(omega_conf)
omegaconf.errors.ConfigIndexError: list index out of range
full_key: reference_image_path[1]
object_type=list

this shows the default reference_image_path: ["results/vlog/teddy_travel/ref_img/teddy.jpg"] only has an element,
how can l adjust my code?

[zhuangshaobin]

This is because the output of LLM may sometimes make mistakes. You can add a line of code to slightly correct it.
You can modify the error code in the picture as follows：
if reference_lists[i][0] > len(args.reference_image_path):
reference_lists[i][0] = 0 # you can set the 0 or len(args.reference_image_path)
pil_image = Image.open(args.reference_image_path[reference_lists[i][0] - 1])
Hope the above information can help you！

[pzwstudy]

Thanks,it's normal now!

[pzwstudy]

File "D:\anaconda\envs\vlog\lib\site-packages\moviepy\video\io\ffmpeg_writer.py", line 88, in __init__ '-r', '%.02f' % fps, TypeError: must be real number, not NoneType lt likes the fps has no init

[pzwstudy]

1.How to eval the FVD score ?
Can you offer a script?

[zhuangshaobin]

As we mentioned in the paper, the code we used to test fvd in the paper is based on the codebase of stylegan-v, follow this link below:
https://github.com/universome/stylegan-v

[pzwstudy]

OK，thanks