Kalosm an embeddable framework for pre-trained models in Rust

Kalosm v0.2.0 introduces new features like Tasks and Agents, Task Evaluation, Prompt Auto-Tuning, Regex Validation, Surreal Database Integration, RAG improvements, Performance Improvements, and New Models. Users can now create tasks for identifying keywords in sentences or solving math problems, with examples to help agents learn. The Evaluation Abstraction feature enhances functionality. Prompt Auto-Tuning optimizes prompts for better performance. Surreal DB integration allows storing and retrieving similar documents based on the text’s meaning. The release also includes improvements in RAG, like new chunking strategies such as Hypothetical Questions. The update simplifies regex validation for model output validation.

https://floneum.com/blog/kalosm_0_2/

To top