# Instruction following optimization

Starling-dolphin-E26-7B is an experimental language model constructed by fusing multiple high-performance models, designed to handle a wide range of tasks including instruction following and creative writing.

Daring Maid-13B is a smarter and more instruction-following capable version of Noromaid, created by integrating features from multiple excellent models.

Daring Maid-20B is a text generation model based on the fusion of multiple excellent models, aiming to create a smarter and more instruction-following Noromaid model.

H2O.ai's h2ogpt - oasst1 - 512 - 12b is a large instruction - following language model with 12 billion parameters and can be used for commercial purposes. This model is fine - tuned on a specific base model using a specific dataset, and has good performance and a wide range of application scenarios.