Gorilla Benchmark API Bench

coding

APIBench, a comprehensive dataset of over 11,000 instruction-API pairs from HuggingFace, TorchHub, and TensorHub APIs for evaluating language models' ability to generate accurate API calls.

Leaderboard

Showing 3 of 3 results

Llama 3.1 405B Instruct

35.3%

i
Llama 3.1 70B Instruct

29.7%

i
Llama 3.1 8B Instruct

8.2%

i