Amir Yazdanbakhsh

New methods improve training efficiency and accuracy of large language models.

Research on optimizing deep learning models with sparsity and quantization techniques.

Research on enhancing language models' efficiency using linear attention and speculative decoding.