llmware-ai
Keeping up with OpenAI
The core focus of the LLMWare project is building pipelines with small open source models, but we believe strongly that the project should be open and supportive of all model providers (small, large, open, proprietary) to build the best solutions.
We appreciate that many members of the LLMWare community use OpenAI, especially for use in "hybrid" strategies that combine both open source and OpenAI in different use cases, or use OpenAI as a benchmark for comparison purposes.
We are rarely at the leading edge of integrating OpenAI models - and generally will not prioritize that among our core development team. Since this is one of our "blind spots," we always welcome contributions from members of the community looking to improve OpenAI support and keep llmware up-to-date on the latest OpenAI models.
We also welcome contributions to support and integrate different authentication and configuration patterns with Azure OpenAI.
The key classes are in the llmware.models module: OpenAIGenModel, along with updating the models supported in the default catalog in llmware.model_configs.
