LLM Tasks
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
3 LLM Tasks (CMMLU, GSM8K, HumanEval) (test)
40.4Average Accuracy
7
3 LLM Tasks Aggregate LLaMa2 (average)
0.405Accuracy
6
Eight benchmark LLM tasks
49,781.23Throughput (Tokens/s)
5
Loading the SOTA2 catalog…