Sparse LLM

A large language model that uses techniques like pruning or quantization to reduce the number of non-zero parameters in the model making it more parameter-efficient and easier to deploy on resource-constrained devices.