How do i pass in np.ndarray with ndim > 1

[pzerone]

I'm working on a project based of face recognition where i have to store the face-embedding into postgres DB. the face-recog part returns a numpy nd-array with 120 dimensions. which is the cropped out portion of the face. When i try to store this, pgvector complains that ndim needs to be 1. I cannot find any relevent documentation regarding this. How do i change it? wny help would be appreciated

[ankane]

Hi @pzerone, it sounds like you're trying to pass a matrix to a vector column. If you have multiple embeddings, you'll typically want to store them as multiple rows.

[pzerone]

Hey @ankane. Thanks for the heads up. Now it makes sense. I was infact trying to pass in a matrix. Thank you for your help.