使用 TensorFlow Keras API 进行模型推理Model inference using TensorFlow Keras API
以下笔记本演示了 Azure Databricks 建议的深度学习推理工作流。The following notebook demonstrates the Azure Databricks recommended deep learning inference workflow. 本示例演示了使用 ResNet-50 模型进行的模型推理，该模型将 TensorFlow Keras API 和 Parquet 文件作为输入数据进行训练。This example illustrates model inference using a ResNet-50 model trained with TensorFlow Keras API and Parquet files as input data. 要理解该示例，应熟悉 Spark数据源。To understand the example, you should be familiar with Spark data sources.