跳转到主要内容

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.

发布日期
2024-09-19
参数规模
7.0B
上下文长度
33K
支持模态
text

能力雷达图

20
general
13
coding
29
reasoning
21
science估算
0
agents
0
multimodal

Science 在缺少专门科学评测时使用推理能力代理估算。

排行榜排名

领域#排名分数来源
代码能力榜399
14.0
AA
通用能力榜424
23.0
AA
数学推理250
35.0
AA
推理能力58
63.0
LS
科学能力418
21.0
AA

基准测试分数 (LLM Stats)

Code

HumanEval88.4%自报
Aider55.6%自报
LiveCodeBench18.2%自报

Finance

MMLU-Base68.0%自报
MMLU67.6%自报
TruthfulQA50.6%自报
MMLU-Pro40.1%自报
TheoremQA34.0%自报

General

MBPP0.83 / 100自报
MMLU-Redux66.6%自报
ARC-C60.9%自报
BigCodeBench41.0%自报

Language

Winogrande72.9%自报

Math

GSM8k83.9%自报
MATH46.6%自报
STEM34.0%自报

Reasoning

HellaSwag76.8%自报
CRUXEval-Input-CoT56.5%自报
CRUXEval-Output-CoT56.0%自报

AA 评测指数

Intelligence Index
10.0
Math 500
0.7
Mmlu Pro
0.5
Gpqa
0.3
Scicode
0.1
Livecodebench
0.1
Aime
0.1
Hle
0.0

LLM Stats 分类评分

General
60
Language
60
Math
60
Reasoning
60
Code
50
Finance
50
Healthcare
50
Legal
50
Physics
30

定价

输入价格免费
输出价格免费
混合价格(3:1)免费

速度

Tokens/秒0.0 tokens/s
首Token延迟0.00s
首回答延迟0.00s

可用提供商

(LS 内部计价单位)

暂无提供商数据

外部链接