Trained using 10% of the python portion of the codesearchnet corpus. We used a novel cosine similarity loss function using sentence-transformers/all-MiniLM-L6-v2 to encode the sentences.
Trained using 10% of the python portion of the codesearchnet corpus. We used a novel cosine similarity loss function using sentence-transformers/all-MiniLM-L6-v2 to encode the sentences.