Qwen 2.5
LatestAlibaba
•
Open Source•
32B# 116
Released
Sep 19, 2024
# 11
Knowledge Cutoff
Sep 24
# 14
Context Length
128K
Benchmarks
# 126
Code RankedAGI
42.9%
# 82
Agentic RankedAGI
51.3%
# 54
Code LiveBench (old)
27.0%
# 115
Reason RankedAGI
44.5%
# 93
GPQA Diamond
49.5%
# 14
Human Eval
88.4%
# 44
MMLU Pro
69.2%
# 28
MMLU
83.9%
# 18
Halluc. Hughes
3.0%
# 21
IF Evaluation
79.5%
# 129
Math RankedAGI
51.7%
# 107
RAGI RankedAGI
47.5%