You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In most algorithms, computing the dot-product for embeddings to check for text similarity will be done against 2 float arrays so it would make more sense to read the data as float rather than double.
There could always be typed CreateEmbeddingResponseFloat and CreateEmbeddingResponseDouble classes.
Great library btw, love using it!
Thanks!
The text was updated successfully, but these errors were encountered:
The response from OpenAI has 18 digits. Choosing a float type would result in significant data loss. However, I can create a generic type response for those who prioritize performance over precision. Let me think about implementation.
In most algorithms, computing the dot-product for embeddings to check for text similarity will be done against 2 float arrays so it would make more sense to read the data as float rather than double.
There could always be typed CreateEmbeddingResponseFloat and CreateEmbeddingResponseDouble classes.
Great library btw, love using it!
Thanks!
The text was updated successfully, but these errors were encountered: