AI orchestration framework to build customizable, production-ready LLM applications. Connect components (models, vector DBs, file converters) to pipelines or agents that can interact with your data. With advanced retrieval methods, it's best suited for building RAG, question answering, semantic search or conversational agent chatbots.
This brings a change we had in v1 that updates the Sentence Transformer specific variable max_seq_length that allows us to change the max sequence length. Previously we though using model_max_length through tokenizer_kwargs would fix this, but a recent inveistgation from @bglearning shows it does not. So we make an update here such that model_max_length updates the value of max_seq_length.
coveralls
commented
2 months ago
Related Issues
Proposed Changes:
This brings a change we had in v1 that updates the Sentence Transformer specific variable
max_seq_lengththat allows us to change the max sequence length. Previously we though usingmodel_max_lengththroughtokenizer_kwargswould fix this, but a recent inveistgation from @bglearning shows it does not. So we make an update here such thatmodel_max_lengthupdates the value ofmax_seq_length.How did you test it?
Notes for the reviewer
Checklist
fix:,feat:,build:,chore:,ci:,docs:,style:,refactor:,perf:,test:.