Alibaba Cloud / Qwen Team

Qwen2 72B Instruct

Zero-eval
#1CMMLU
#1TheoremQA
#2EvalPlus
+1 more

by Alibaba Cloud / Qwen Team

+
+
+
+
About

Qwen2 72B was developed as the flagship model in the Qwen2 generation, designed to provide advanced language understanding with 72 billion parameters. Built to deliver strong performance across diverse tasks, it represented a significant advancement in Qwen's model capabilities when introduced.

+
+
+
+
Timeline
AnnouncedJul 23, 2024
ReleasedJul 23, 2024
+
+
+
+
License & Family
License
tongyi-qianwen
Performance Overview
Performance metrics and category breakdown

Overall Performance

17 benchmarks
Average Score
73.6%
Best Score
91.1%
High Performers (80%+)
9
+
+
+
+
All Benchmark Results for Qwen2 72B Instruct
Complete list of benchmark scores with detailed information
GSM8k
text
0.91
91.1%
Self-reported
CMMLU
text
0.90
90.1%
Self-reported
HellaSwag
text
0.88
87.6%
Self-reported
HumanEval
text
0.86
86.0%
Self-reported
Winogrande
text
0.85
85.1%
Self-reported
C-Eval
text
0.84
83.8%
Self-reported
BBH
text
0.82
82.4%
Self-reported
MMLU
text
0.82
82.3%
Self-reported
MBPP
text
0.80
80.2%
Self-reported
EvalPlus
text
0.79
79.0%
Self-reported
Showing 1 to 10 of 17 benchmarks
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+