Search papers, labs, and topics across Lattice.
This paper introduces DynamiCS, a dynamic cluster-based sampling method for vision-language pre-training that downsamples large data clusters and upsamples small ones to address the underrepresentation of long-tail concepts. By dynamically adjusting the sampling strategy at each epoch and preserving the relative order of semantic clusters, DynamiCS improves the representation of rare concepts during training. Experiments demonstrate that DynamiCS reduces computational cost while improving performance on long-tail concepts compared to existing methods that focus solely on flattening the semantic distribution.
VLMs can get a boost in long-tail performance and train more efficiently by dynamically upsampling underrepresented data clusters each epoch.
The computational cost of training a vision-language model (VLM) can be reduced by sampling the training data. Previous work on efficient VLM pre-training has pointed to the importance of semantic data balance, adjusting the distribution of topics in the data to improve VLM accuracy. However, existing efficient pre-training approaches may disproportionately remove rare concepts from the training corpus. As a result, \emph{long-tail concepts} remain insufficiently represented in the training data and are not effectively captured during training. In this work, we introduce a \emph{dynamic cluster-based sampling approach (DynamiCS)} that downsamples large clusters of data and upsamples small ones. The approach is dynamic in that it applies sampling at each epoch. We first show the importance of dynamic sampling for VLM training. Then, we demonstrate the advantage of our cluster-scaling approach, which maintains the relative order of semantic clusters in the data and emphasizes the long-tail. This approach contrasts with current work, which focuses only on flattening the semantic distribution of the data. Our experiments show that DynamiCS reduces the computational cost of VLM training and provides a performance advantage for long-tail concepts.
