Hey all, I am just starting with semantic search. I have enabled auto embeddings using the

multilingual-e5-base

model and setting the field option to

optional

. As we have a lot of data I assume this will take quite some time to index (~300k items, each having around 300-500 words, resulting in around 150M tokens). I would expect that the "regular" search by text match should still work until indexing is done, but our requests are failing. is this due to lack of resources? the error response is unfortunately empty as it gets cancelled after exactly 5seconds. we have 2vcpus and 4GB ram allocated.