164 lines
3.6 KiB
Markdown
164 lines
3.6 KiB
Markdown
# 智能算法展示平台Python SDK
|
||
|
||
本SDK提供了与智能算法展示平台API交互的Python客户端，支持算法调用、仿真数据生成等功能。
|
||
|
||
## 安装
|
||
|
||
### 从源码安装
|
||
|
||
```bash
|
||
cd sdk/python
|
||
pip install -e .
|
||
```
|
||
|
||
### 从PyPI安装
|
||
|
||
```bash
|
||
pip install algorithm-showcase-sdk
|
||
```
|
||
|
||
## 使用示例
|
||
|
||
### 1. 初始化客户端
|
||
|
||
```python
|
||
from algorithm_showcase import AlgorithmShowcaseClient
|
||
|
||
# 初始化客户端
|
||
client = AlgorithmShowcaseClient(
|
||
base_url="http://localhost:8001/api/v1", # API基础URL
|
||
api_key="your-api-key" # 可选，API密钥
|
||
)
|
||
```
|
||
|
||
### 2. 获取算法列表
|
||
|
||
```python
|
||
# 获取所有算法
|
||
algorithms = client.get_algorithms()
|
||
|
||
# 获取特定类型的算法
|
||
algorithms = client.get_algorithms(algorithm_type="computer_vision")
|
||
|
||
# 打印算法信息
|
||
for algo in algorithms:
|
||
print(f"算法ID: {algo.id}")
|
||
print(f"算法名称: {algo.name}")
|
||
print(f"算法描述: {algo.description}")
|
||
print(f"算法类型: {algo.type}")
|
||
print("---")
|
||
```
|
||
|
||
### 3. 获取算法详情
|
||
|
||
```python
|
||
# 获取算法详情
|
||
algorithm = client.get_algorithm(algorithm_id="algorithm-123456")
|
||
|
||
print(f"算法ID: {algorithm.id}")
|
||
print(f"算法名称: {algorithm.name}")
|
||
print(f"算法描述: {algorithm.description}")
|
||
print(f"算法类型: {algorithm.type}")
|
||
print("版本信息:")
|
||
for version in algorithm.versions:
|
||
print(f" - 版本: {version.version} (默认: {version.is_default})")
|
||
print(f" URL: {version.url}")
|
||
print(f" 参数: {version.params}")
|
||
```
|
||
|
||
### 4. 调用算法
|
||
|
||
```python
|
||
# 准备输入数据
|
||
input_data = {
|
||
"text": "这是一段测试文本"
|
||
}
|
||
|
||
# 准备算法参数
|
||
params = {
|
||
"confidence_threshold": 0.5,
|
||
"model_name": "resnet50"
|
||
}
|
||
|
||
# 调用算法
|
||
result = client.call_algorithm(
|
||
algorithm_id="algorithm-123456",
|
||
version_id="version-123456",
|
||
input_data=input_data,
|
||
params=params
|
||
)
|
||
|
||
print(f"调用ID: {result.id}")
|
||
print(f"状态: {result.status}")
|
||
print(f"响应时间: {result.response_time} 秒")
|
||
print(f"输出数据: {result.output_data}")
|
||
```
|
||
|
||
### 5. 获取调用结果
|
||
|
||
```python
|
||
# 获取调用结果
|
||
result = client.get_call_result(call_id="call-123456")
|
||
|
||
print(f"调用ID: {result.id}")
|
||
print(f"状态: {result.status}")
|
||
print(f"响应时间: {result.response_time} 秒")
|
||
print(f"输出数据: {result.output_data}")
|
||
```
|
||
|
||
### 6. 生成仿真输入数据
|
||
|
||
```python
|
||
# 生成文本数据
|
||
text_data = client.generate_simulation_data(
|
||
prompt="生成一段关于人工智能的新闻文本",
|
||
data_type="text"
|
||
)
|
||
print(f"生成的文本: {text_data['data']}")
|
||
|
||
# 生成结构化数据
|
||
structured_data = client.generate_simulation_data(
|
||
prompt="生成一个包含姓名、年龄、职业的用户信息",
|
||
data_type="structured"
|
||
)
|
||
print(f"生成的结构化数据: {structured_data['data']}")
|
||
```
|
||
|
||
## API文档
|
||
|
||
### 核心方法
|
||
|
||
- `get_algorithms(algorithm_type=None)`: 获取算法列表
|
||
- `get_algorithm(algorithm_id)`: 获取算法详情
|
||
- `get_algorithm_versions(algorithm_id)`: 获取算法版本列表
|
||
- `call_algorithm(algorithm_id, version_id, input_data, params=None)`: 调用算法
|
||
- `get_call_result(call_id)`: 获取调用结果
|
||
- `generate_simulation_data(prompt, data_type="text")`: 生成仿真输入数据
|
||
|
||
### 数据模型
|
||
|
||
- `Algorithm`: 算法信息
|
||
- `AlgorithmVersion`: 算法版本信息
|
||
- `AlgorithmCallRequest`: 算法调用请求
|
||
- `AlgorithmCallResult`: 算法调用结果
|
||
|
||
## 错误处理
|
||
|
||
SDK会捕获API请求中的错误，并抛出异常。建议在使用时添加异常处理：
|
||
|
||
```python
|
||
try:
|
||
result = client.call_algorithm(...)
|
||
except Exception as e:
|
||
print(f"调用失败: {e}")
|
||
```
|
||
|
||
## 版本兼容性
|
||
|
||
- Python 3.7+
|
||
- requests 2.31.0+
|
||
|
||
## 许可证
|
||
|
||
MIT License
|