Embedding

The community member is trying to create a VectorStoreIndex from a large list of documents, but the embeddings generation is taking a long time. The community members suggest trying indexing with async, increasing the batch size, and updating the llama_index library to a newer version. The community members also provide specific code examples for creating a VectorStoreIndex with use_async=True. The community members note that increasing the batch size to 2048 helped a little, but the process is still slow for 100k documents. The solution seems to be updating the llama_index library, which resolved the AssertionError and made the embedding process much faster.