Closed evelinehong closed 3 months ago
Hi there, most of the dataset is sourced from the dataset used to train VPT, which can be found at their repo. Unfortunately, the rest of the dataset (~15M frames) is not publicly available (it is difficult to host such a large amount of data). See the STEVE-1 paper for more details about the dataset.
Hi, do you provide the 54M dataset?