Isaac Lee
The biggest barrier to trying a new setting (even for users with experience) is the time it takes to figure out the correct lane IDs. Since information from `get_lane_vehicle_count()` or...
it is difficult to figure out roadIds which are required for most information-querying APIs. One has to read the roadnet file in order to figure out which road is which,...
Type error when running: `python run_atari.py --gamma_ext 0.999` and have applied below changes in `mpi_util.py`: `available_gpus = guess_available_gpus()`-> `available_gpus = 1` and `os.environ['CUDA_VISIBLE_DEVICES'] = str(1)` -> `os.environ['CUDA_VISIBLE_DEVICES'] = str(available_gpus[local_rank])` ```...
The Rainbow DQN paper uses human-expert normalized scores, so I am not sure how to evaluate the training results against the original paper. Do you know what values were used...
replay
1. **Can't find `replay.txt` and `roadnetlog.json`** I am running training & inference on a remote server and want to watch the replays on my local desktop. I see that the...
`kitti.pkl` is 0 bytes. It would be great if you could re-upload the weights. Thanks!
What's the train / val / test splits used for training the pretrained weights? I want to run validation on CO3D dataset using the released pretrained weights.
### What problem is this feature trying to solve? Claude code now supports [remote mcp server connection](https://docs.anthropic.com/en/release-notes/claude-code). ### Describe the solution you'd like Zen should also be hosted online as...