plaguss/mistal-7b-prm-openrlhf
TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Dec 9, 2024Architecture:Transformer Cold

The plaguss/mistal-7b-prm-openrlhf is a 7 billion parameter causal language model, likely based on the Mistral architecture, that has been fine-tuned using a Preference Ranking Model (PRM) approach. This model is designed for tasks requiring nuanced evaluation of outputs, as suggested by its PRM training, and is demonstrated with an example involving scoring different outputs for a given question. Its primary strength lies in its ability to assign scores to generated text, indicating its suitability for tasks like response ranking or quality assessment.

Loading preview...