Mono_Magic_Drive

Code for monocular view conditioned scene generation using Diffusion Models

Training and Validation

How to train :

accelerate launch --mixed_precision fp16 --gpu_ids all --num_processes 8 train.py \
  +exp=224x400 runner=8gpus

How to test:

python test.py

Make changes in the config file accordingly here:

out.mp4

Name		Name	Last commit message	Last commit date
Latest commit History 8 Commits
assets		assets
bev		bev
configs		configs
dataset		dataset
misc		misc
model		model
pipeline		pipeline
runner		runner
README.md		README.md
env.txt		env.txt
test.py		test.py
test_view.py		test_view.py
train.py		train.py