fit_pipeline 模块

用于拟合管道的代码。

函数

edge_helper

edge_helper(source_node_id: Optional[str], source_node_name: Optional[str], source_name: Optional[str], target_name: Optional[str], dst_node_id: Optional[str], dst_node_name: Optional[str], graph_json_dict: Dict[str, Any]) -> None

参数

source_node_id
source_node_name
source_name
target_name
dst_node_id
dst_node_name
graph_json_dict

fit_pipeline

运行 AutoML 试验的单次迭代。

在常规 AutoML 试验期间会自动调用此方法。 fit_pipeline 将评估此迭代的管道,使用提供的数据拟合管道,计算与此试验相关的各种指标,并将所有结果记录在指定 AzureML 运行的历史记录中。

fit_pipeline(automl_pipeline: azureml.automl.runtime.automl_pipeline.AutoMLPipeline, control_settings: azureml.automl.runtime._runtime_params.ExperimentControlSettings, resource_settings: azureml.automl.runtime._runtime_params.ExperimentResourceSettings, orchestration_settings: azureml.automl.runtime._runtime_params.ExperimentOrchestrationSettings, pipeline_selection_settings: azureml.automl.runtime._runtime_params.PipelineSelectionSettings, data_settings: azureml.automl.runtime._runtime_params.ExperimentDataSettings, automl_run_context: azureml.automl.runtime.automl_run_context.AutoMLAbstractRunContext, elapsed_time: Optional[int] = None, onnx_cvt: Optional[azureml.automl.runtime.onnx_convert.onnx_converter.OnnxConverter] = None) -> azureml.automl.runtime.fit_output.FitOutput

参数

automl_pipeline
必需

包含管道 ID 和序列化脚本的 AutoMLPipeline 对象。

fit_pipeline_params
必需

用于拟合此管道的用户设置

automl_run_context
必需

子运行上下文对象

elapsed_time
必需

此试验已持续多长时间(以分钟为单位)

onnx_cvt
必需

onnx 转换器。

data_settings
automl_run_context
elapsed_time
默认值: None
onnx_cvt
默认值: None

返回

此子运行的 AzureML 运行属性