When i try the Embeddings on the toy examples, it fails with below error ,
RateLimitError: exceeded quota for this month
is there a limit on the key per user usage ?
From this message I think there is a limit then.
Same here – RateLimitError: exceeded quota for this month - running following code: embedding1 = embedding.embed_query(sentence1)
embedding2 = embedding.embed_query(sentence2)
embedding3 = embedding.embed_query(sentence3)
Hi,
I’ve got the same error, but I have used a workaround, ie to use HF embedding instead and that
got me going. I am not sure if DLAI is happy with this or not because I needed to installed the
sentence_transformer package as well in the env.
If DLAI staffs & mentors are fine I will share my changes here.
Basically the dot products for the 3 comparisons using the HF embedding model all-MiniLM-L6-v2:
np.dot(embedding1, embedding2) == 0.9151646036692593 => similar
np.dot(embedding1, embedding3) == 0.08337090439921915 => dissimilar
np.dot(embedding2, embedding3) == 0.04040370547493643 => dissimilar
Interestingly if you compare the dissimilar dot products vs those in the video, it suggests the HF
embedding actually provides a better “dissimilar” lower score vs openai one.
Rgds,
MCW