InternLM2 Math 7B

About

InternLM2 Math 7B is a bilingual large language model specialized in advanced mathematical reasoning and problem-solving across Chinese and English languages. It functions as a solver, prover, verifier, and augmentor of math problems, showcasing capabilities in theorem proving, code interpretation, and reward modeling. Trained on approximately 100 billion high-quality math-related tokens and 2 million bilingual supervised data points, the model is designed for both research and commercial applications, including educational tools and automated theorem proving systems. Despite its strengths, it has limitations in specific problem types and potential biases from its training data.

Capabilities

Rankings