dataset
commavq-trajectory
AdityaNG
or hover any field below to flag it
Overview
Name
commavq-trajectory
Source
AdityaNG
Episodes
0
Robot count
0
Format
other
Description
CommaVQ Trajectory
Based on the commavq dataset.
Contains images of highway driving along with the corresponding control signal (trajectory).
The trajectory has been quantized into one of 256 template trajectories.
This data is formatted to train the LLaVA model:
[
{
"id": "4376800841",
"image": "data_0_to_2500/324a487ead6977b02c746d5117e9825a_16/841.png",
"conversations": [
{
"from": "human",
"value":… See the full description on the dataset page: https://huggingface.co/datasets/AdityaNG/commavq-trajectory.
Robots used
null
Links
HuggingFace dataset