How to use text-embedding-3-small with different output dimensions?

[anirban-intel]

Could you please provide me with information on how to use text-embedding-3-small model with various output dimensions? Also, is there any estimated timeline for this?

Tasks

Preview Give feedback

No tasks being tracked yet.

[gramhagen]

I can take a crack at adding this functionality. Expected usage would be like this:

from synapse.ml.services.openai import OpenAIEmbedding

embedding = (
    OpenAIEmbedding()
    .setSubscriptionKey(key)
    .setDeploymentName("text-embedding-3-small")
    .setCustomServiceName(service_name)
    .setApiVersion("2024-03-01-preview")
    .setDimensions(256)
    .setTextCol("prompt")
    .setErrorCol("error")
    .setOutputCol("embeddings")
)

embedding.transform(df)