TACO: Topics in Algorithmic COde generation dataset
Open paper
Benchmark
Model
Rank
Results
code-generation-on-taco-code
GPT-4
#1
easy pass@1: 31.50%
code-generation-on-taco-code
Starcoder-15.5B
#2
easy pass@1: 11.6%
code-generation-on-taco-code
CodeLlama-7B-Python
#3
easy pass@1: 9.32%
Rank counts only results with a code link.