The Export OSS Data to MaxCompute Table component reads data files from an Object Storage Service (OSS) directory and writes them to a MaxCompute table. Point the component at an OSS directory and it reads all files in that directory by default.
Input format requirement: JSON Lines (JSONL) only. Each line must be a valid JSON object; the file as a whole is not a valid JSON object. Download a sample data file to see the expected structure.
Supported computing resources
Configure the component
Configure the Export OSS Data to MaxCompute Table component on the pipeline page of Machine Learning Designer in the Platform for AI (PAI) console.
| Tab | Parameter | Required | Default | Description |
|---|---|---|---|---|
| Fields Setting | Output Table Name | Yes | — | Name of the MaxCompute table to write data to. |
| Tuning | Number of CPUs per instance of map task | No | 100 | CPUs allocated to each map task instance. Valid values: 50–800. |
| The memory size per instance of map task | No | 1024 MB | Memory allocated to each map task instance. Valid values: 256–12288 MB. | |
| The maximum size of input data for a map | No | 256 MB | Maximum data each map task instance processes. Valid values: 1 MB to Integer.MAX_VALUE. |
该文章对您有帮助吗？