Qwen2.5 Coder Instruct 7B
AlibabaQwenOpen WeightApache 2.0 · Commercial OK
描述
Qwen2.5-Coder is a specialized coding model trained on 5.5 trillion tokens of code data, supporting 92 programming languages with a 128K context window. It excels in code generation, completion, and repair while maintaining strong performance in math and general tasks. The model demonstrates exceptional capabilities in multi-programming language tasks and code reasoning.
發布日期
2024-09-19
參數規模
7.0B
上下文長度
33K
支援模態
text
能力雷達圖
20
general
13
coding
29
reasoning
21
science估算
0
agents
0
multimodal
Science 在缺少專門科學評測時使用推理能力代理估算。
排行榜排名
基準測試分數 (LLM Stats)
Code
HumanEval
88.4%自報
Aider
55.6%自報
LiveCodeBench
18.2%自報
Finance
MMLU-Base
68.0%自報
MMLU
67.6%自報
TruthfulQA
50.6%自報
MMLU-Pro
40.1%自報
TheoremQA
34.0%自報
General
MBPP
0.83 / 100自報
MMLU-Redux
66.6%自報
ARC-C
60.9%自報
BigCodeBench
41.0%自報
Language
Winogrande
72.9%自報
Math
GSM8k
83.9%自報
MATH
46.6%自報
STEM
34.0%自報
Reasoning
HellaSwag
76.8%自報
CRUXEval-Input-CoT
56.5%自報
CRUXEval-Output-CoT
56.0%自報
AA 評測指數
Intelligence Index10.0
Math 5000.7
Mmlu Pro0.5
Gpqa0.3
Scicode0.1
Livecodebench0.1
Aime0.1
Hle0.0
LLM Stats 分類評分
General60
Language60
Math60
Reasoning60
Code50
Finance50
Healthcare50
Legal50
Physics30
定價
輸入價格免費
輸出價格免費
混合價格(3:1)免費
速度
Tokens/秒0.0 tokens/s
首Token延遲0.00s
首回答延遲0.00s
可用提供商
(LS 內部計價單位)暫無提供商資料