ai-edge-model-explorer>=0.1.14
lm_eval[api]==0.4.5