search

McGill-NLP / AURORA

Code and data for the paper: Learning Action and Reasoning-Centric Image Editing from Videos and Simulation
https://aurora-editing.github.io/
MIT License
10 stars 2 forks source link

Running Evaluation on Aurora-Bench #8

Closed alphacoder01 closed 1 week ago

alphacoder01 commented 4 weeks ago

Hi,

It seems the json files corresponding to different tasks are missing. As mentioned in the (README.MD) Running python disc_edit.py --task "whatsup" leads to the following error

FileNotFoundError: [Errno 2] No such file or directory: 'data/whatsup/itm_test.json'

Can the authors please provide the missing files/correct instructions for running the evaluation with proposed metric.

Thanks!!

alphacoder01 commented 1 week ago

Hi @BennoKrojer I was wondering if there have been any updates or plans to release the files and instructions for evaluation.

Thanks in Advance!!

BennoKrojer commented 1 week ago

Hi Ashish,

sorry for the delay, I somehow missed this! I will look into it later today.

Best, Benno

BennoKrojer commented 1 week ago

Hi,

To unblock you fast I just uploaded the data for now but you will have to change some file paths unfortunately inside the dataloader (edit_dataset.py). I am quite busy the next weeks but I hope this is better than nothing and with changing a few lines you should be good to go!

Here are the jsons: https://github.com/McGill-NLP/AURORA/tree/main/disc_edit_data

Best, Benno