Qwen 2.5
LatestAlibaba
•
Open Source•
14B# 115
Released
Sep 19, 2024
# 11
Knowledge Cutoff
Sep 24
# 14
Context Length
128K
Benchmarks
# 131
Code RankedAGI
39.7%
# 88
Agentic RankedAGI
47.4%
# 58
Code LiveBench (old)
20.9%
# 127
Reason RankedAGI
41.6%
# 102
GPQA Diamond
45.5%
# 26
Human Eval
83.5%
# 59
MMLU Pro
63.7%
# 34
MMLU
80.0%
# 25
Halluc. Hughes
4.2%
# 20
IF Evaluation
81.0%
# 131
Math RankedAGI
51.0%
# 121
RAGI RankedAGI
45.2%