search

facebookresearch / AVT

Code release for ICCV 2021 paper "Anticipative Video Transformer"
Apache License 2.0
151 stars 28 forks source link

Env Create Conflicts #16

Closed GOZGOXGOYGO closed 2 years ago

GOZGOXGOYGO commented 2 years ago

Hi,

Your work is so impressive. Actually, I am a beginner in this field. I try to use the env.yaml file that you provide to create a conda env through conda env create -f env.yaml python=3.7.7. But it reports that found conflicts. Then what should I do?

Thank you in advance!

rohitgirdhar commented 2 years ago

Hi, that is strange, I exported that YAML file from my environment. You can perhaps try to create the environment from scratch: the main dependencies are pytorch, hydra, huggingface transformers. And you can install other dependencies if it complains when you run.

GOZGOXGOYGO commented 2 years ago

Thanks a lot. Create the environment from scratch seems helpful.

May I ask which part of the code is Backbone model, and which part of the code is Head model? I'm sorry I'm a little confused about the five files in the models folder.

rohitgirdhar commented 2 years ago

The backbones are in video_classification.py and heads in future_prediction.py. The combined model is constructed in the base_model.py.

GOZGOXGOYGO commented 2 years ago

Thanks so much for your answer, but I don't see the class AVTh in future_prediction.py being called in base_model.py or any other code files. https://github.com/facebookresearch/AVT/blob/b372773a2fd75e295da3ab737111363b4c860546/models/future_prediction.py#L51

rohitgirdhar commented 2 years ago

It's because I use hydra instantiate to create class objects. The AVT-h is being defined in the config file here, which initializes this object in base_model.py