I encountered an issue when trying to run the batch_example.py script on Python 3.8.10. The error seems to be related to parsing a protobuf message in the waymo_dataset.py file.
Steps to reproduce:
Setup the environment using the following command:
pip install -e .[waymo]
The traceback is as follows:
Traceback (most recent call last):
File "**/batch_example.py", line 55, in <module>
main()
File "**/batch_example.py", line 14, in main
dataset = UnifiedDataset(
File "**/dataset.py", line 304, in __init__
env.cache_maps(
File "**/waymo_dataset.py", line 373, in cache_maps
self.cache_map(i, cache_path, map_cache_class, map_params)
File "**/waymo_dataset.py", line 342, in cache_map
scenario.ParseFromString(bytearray(data.numpy()))
google.protobuf.message.DecodeError: Error parsing message with type 'waymo.open_dataset.Scenario'
Also, I notice that if we do pip install -e .[nusc,lyft,waymo,interaction], there will be package versions conflict:
ERROR: Cannot install tensorflow==2.11.0, trajdata and trajdata[interaction,lyft,nusc,waymo]==1.3.3 because these package versions have conflicting dependencies.
The conflict is caused by:
trajdata[interaction,lyft,nusc,waymo] 1.3.3 depends on numpy>=1.19
l5kit 1.5.0 depends on numpy~=1.19.0
nuscenes-devkit 1.1.9 depends on numpy
tensorflow 2.11.0 depends on numpy>=1.2
Would appreciate any insights or solutions regarding this. Thank you!
Hello,
I encountered an issue when trying to run the
batch_example.py
script on Python3.8.10
. The error seems to be related to parsing a protobuf message in thewaymo_dataset.py
file.Steps to reproduce:
The traceback is as follows:
Also, I notice that if we do pip install -e .[nusc,lyft,waymo,interaction], there will be package versions conflict:
Would appreciate any insights or solutions regarding this. Thank you!