Get the latest tech news

voyage-3.5 and voyage-3.5-lite: improved quality for a new retrieval frontier


TL;DR – We’re excited to introduce voyage-3.5 and voyage-3.5-lite, the latest generation of our embedding models. These models offer improved retrieval quality over voyage-3 and voyage-3-lite at th…

Both models support embeddings in 2048, 1024, 512, and 256 dimensions, with multiple quantization options enabled by Matryoshka learning and quantization-aware training. Each dataset consists of a corpus (e.g., technical documentation, court opinions) and queries (e.g., questions, summaries). CategoryDescriptionsDatasets TECHTechnical documentationCohere, 5G, OneSignal, LangChain, PyTorchCODECode snippets, docstringsLeetCodeCpp-rtl, LeetCodeJava-rtl, LeetCodePython-rtl, HumanEval-rtl, MBPP-rtl, DS1000-referenceonly-rtl, DS1000-rtl, APPS-rtlLAWCases, court opinions, statutes, patents LeCaRDv2, LegalQuAD, LegalSummarization, AILA casedocs, AILA statutes FINANCESEC filings, finance QARAG benchmark (Apple-10K-2022), FinanceBench, TAT-QA-rtl, Finance Alpaca, FiQA-Personal-Finance-rtl, Stock News Sentiment, ConvFinQA-rtl, FinQA-rtl, HC3 FinanceWEBReviews, forum posts, policy pagesHuffpostsports, Huffpostscience, Doordash, Health4CALONG-CONTEXTLong documents on assorted topics: government reports, academic papers, and dialogues NarrativeQA, Needle, Passkey, QMSum, SummScreenFD, WikimQA CONVERSATIONMeeting transcripts, dialoguesDialog Sum, QA Conv, HQA Models.

Get the Android app

Or read this on Hacker News

Read more on:

Photo of improved quality

improved quality

Photo of voyage-3.5

voyage-3.5