I had a doubt regarding <https://typesense.org/doc...
# community-help
h
I had a doubt regarding https://typesense.org/docs/0.25.0/api/vector-search.html#using-your-own-models. In this when is the model loaded into memory? Suppose i want to use different models in different collections are all of them loaded into memory?
a
Hey Harsh, We'll be using the models in two places: 1. To generate embeddings for the store fields and index them 2. To generate embeddings for the search queries Since the model will be used constantly, it'll stay loaded in memory at all times. Also, if multiple models are declared, any of them could be needed at any moment, so they all stay in memory concurrently.
h
ok thanks @Alan Martini
🙌 1
f
Also to add, if self-hosting, you can find logs that show when the model was loaded into memory