-
Notifications
You must be signed in to change notification settings - Fork 14
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Rerank 3.5 #267
base: main
Are you sure you want to change the base?
Rerank 3.5 #267
Conversation
🌿 Preview your docs: https://cohere-preview-828f66fa-7c92-43d4-bc90-2e255b769c65.docs.buildwithfern.com |
🌿 Preview your docs: https://cohere-preview-52642da7-2957-4a1f-be0b-067de33b4920.docs.buildwithfern.com |
🌿 Preview your docs: https://cohere-preview-cafd35e9-eb46-4fc7-a026-b6d6c1efaefd.docs.buildwithfern.com |
🌿 Preview your docs: https://cohere-preview-8ba6dc11-59b8-4b4f-b4e6-2b6d869e06fa.docs.buildwithfern.com |
🌿 Preview your docs: https://cohere-preview-3dff02d3-1788-4a22-9996-2cc799711cd0.docs.buildwithfern.com |
🌿 Preview your docs: https://cohere-preview-986f855f-2277-4c51-ae7d-719de0345e92.docs.buildwithfern.com |
This PR introduces a new model,
rerank-english-v3.5
, which offers state-of-the-art performance in English and non-English languages. The model supports the same languages asembed-multilingual-v3.0
and has a context length of 4096 tokens.Summary of Changes
rerank-english-v3.5
model is added to the list of latest models, with a description highlighting its capabilities and context length.rerank-v3.5
section is introduced, providing details about the model's constraints, including the number of documents, maximum number of chunks, and tokens per document and query.rerank-v3.5
, specifying that bothrerank-v3.0
andrerank-v3.5
models break documents into 4096 token chunks.rerank-v3.0
models is clarified, stating that the endpoint will throw an error if the user attempts to pass more than 1000 documents at a time.rerank-v3.0
andrerank-v3.5
models are specified, ensuring that both models have a context length of 4096 tokens and that queries larger than 2048 tokens are truncated.