The community member @Logan M tried using the SemanticSimilarityEvaluator, but encountered an SSLCertVerificationError - RetryError when running the code. They asked if anyone else has faced this issue and if there is a synchronous way to achieve this in llama-index.
In the comments, other community members suggested that the SemanticSimilarityEvaluator works by finding semantic similarity between texts and does not require an API or SSL. They speculated that the issue might be related to the OpenAI embeddings used by the evaluator, and suggested trying the code again or using a different embedding model like BAAI/bge-base-en-v1.5 instead of the mpnet V2 model.
@Logan M tried the SemanticSimilarityEvaluator. The code result = await evaluator.aevaluate(
response=student_answer,
reference=correct_answer,
) is throwing SSLCertVerificationError - RetryError: RetryError[<Future at 0x2852955d0 state=finished raised APIConnectionError>] every time. What to do!? Anybody else here faced this before? Also is there no synchronous way to achieve this in llama-index?
