You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
|
|
|
|
|
"""
|
|
|
|
|
|
第十二章示例2：BFCL快速开始
|
|
|
|
|
|
|
|
|
|
|
|
对应文档：12.2.5 在HelloAgents中实现BFCL评估 - 方式1
|
|
|
|
|
|
|
|
|
|
|
|
这是最简单的BFCL评估方式，一行代码完成评估、报告生成和官方评估。
|
|
|
|
|
|
"""
|
|
|
|
|
|
|
|
|
|
|
|
from hello_agents import SimpleAgent, HelloAgentsLLM
|
|
|
|
|
|
from hello_agents.tools import BFCLEvaluationTool
|
|
|
|
|
|
|
|
|
|
|
|
# 1. 创建要评估的智能体
|
|
|
|
|
|
llm = HelloAgentsLLM()
|
|
|
|
|
|
agent = SimpleAgent(name="TestAgent", llm=llm)
|
|
|
|
|
|
|
|
|
|
|
|
# 2. 创建BFCL评估工具
|
|
|
|
|
|
bfcl_tool = BFCLEvaluationTool()
|
|
|
|
|
|
|
|
|
|
|
|
# 3. 运行评估（自动完成所有步骤）
|
|
|
|
|
|
results = bfcl_tool.run(
|
|
|
|
|
|
agent=agent,
|
|
|
|
|
|
category="simple_python", # 评估类别
|
|
|
|
|
|
max_samples=5 # 评估样本数（0表示全部）
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
# 4. 查看结果
|
|
|
|
|
|
print(f"准确率: {results['overall_accuracy']:.2%}")
|
|
|
|
|
|
print(f"正确数: {results['correct_samples']}/{results['total_samples']}")
|
|
|
|
|
|
|
|
|
|
|
|
# 运行输出示例：
|
|
|
|
|
|
# ============================================================
|
|
|
|
|
|
# BFCL一键评估
|
|
|
|
|
|
# ============================================================
|
|
|
|
|
|
#
|
|
|
|
|
|
# 配置:
|
|
|
|
|
|
# 智能体: TestAgent
|
|
|
|
|
|
# 类别: simple_python
|
|
|
|
|
|
# 样本数: 5
|
|
|
|
|
|
#
|
|
|
|
|
|
# 评估进度: 100%|██████████| 5/5 [00:15<00:00, 3.12s/样本]
|
|
|
|
|
|
#
|
|
|
|
|
|
# ✅ 评估完成
|
|
|
|
|
|
# 总样本数: 5
|
|
|
|
|
|
# 正确样本数: 5
|
|
|
|
|
|
# 准确率: 100.00%
|
|
|
|
|
|
#
|
|
|
|
|
|
# 准确率: 100.00%
|
|
|
|
|
|
# 正确数: 5/5
|
|
|
|
|
|
|
