Currently, CoVT supports evaluation through VLMEvalKit, which works well for the built-in benchmarks. However, I am trying to evaluate CoVT on other benchmarks and would like to deploy the model using vLLM for inference.Is there an official or recommended way to load CoVT models in vLLM?
I would really appreciate any guidance on how to properly load CoVT into vLLM
Currently, CoVT supports evaluation through VLMEvalKit, which works well for the built-in benchmarks. However, I am trying to evaluate CoVT on other benchmarks and would like to deploy the model using vLLM for inference.Is there an official or recommended way to load CoVT models in vLLM?
I would really appreciate any guidance on how to properly load CoVT into vLLM