Machine reading comprehension predict

Limits

Only Deep Learning Containers (DLC) resources are supported.

Configure the component in Machine Learning Designer

• Input ports

  Input port	Data type	Recommended upstream component	Required
  input saved model	OSS	machine reading comprehension training	Yes
  Data for Prediction	OSS	Read OSS data	Yes

• Component parameters

  Configure the component on the pipeline page of Machine Learning Designer.

  Tab	Parameter	Description
  Fields Setting	Language	The language of the input file. Default value: zh. zh en
  	Input Schema	The data schema of each column in the input file. Separate multiple columns with commas (,). Default value: qas_id:str:1,context_text:str:1,question_text:str:1,answer_text:str:1,start_position_character:str:1,title:str:1.
  	Question Column	The column that contains questions in the input file. Default value: question_text.
  	Context Column	The column that contains text passages in the input file. Default value: context_text.
  	Answer Column	The column that contains answers in the input file. Default value: answer_text.
  	Id Column	The ID column in the input file. Default value: qas_id.
  	Start Position Column	The column that contains the starting positions of answer spans in the input file. If the answer to a question can be found in the text passage, the starting position of the answer span is recorded in this column. Default value: start_position_character.
  	Output data file	The OSS path for the output answer file.
  	Use User-defined Model	Whether to use a custom model. Default value: no. no yes
  	OSS Directory for Alink Model	Required only if Use User-defined Model is set to yes. The OSS path of the custom model.
  Parameters Setting	batchSize	The number of samples processed at a time. In multi-server, multi-GPU training, this is the per-GPU batch size. The value must be of the INT type. Default value: 256.
  	Sequence Length	The maximum text passage length. The value must be of the INT type. Default value: 384.
  	Max Query Length	The maximum question length. The value must be of the INT type. Default value: 64.
  	Max Answer Length	The maximum answer length. The value must be of the INT type. Default value: 30.
  	Doc Stride	The sliding window stride for sliced text passages. The value must be of the INT type. Default value: 128.
  	pretrainModelNameOrPath	The name or path of a system-provided pre-trained model. Default value: hfl/macbert-base-zh. User Defined hfl/macbert-base-zh hfl/macbert-large-zh bert-base-uncased bert-large-uncased
  	Additional Parameters	Custom parameters for fine-tuning the model. Format: {A: xxx, B: xxx}. Separate keys and values with colons (:). Separate multiple parameters with commas (,).
  Tuning	GPU Machine type	The GPU-accelerated instance type. The default value is gn5-c8g1.2xlarge, which provides 8 vCPUs, 80 GB of memory, and a single P100 GPU.
  	num_GPU_worker	The number of GPUs per worker. Default value: 1.

Example

The following figure shows a sample pipeline that uses this component.

To configure the component, perform the following steps:

1. Prepare a prediction dataset and upload the dataset to an OSS bucket. For more information, see the "Upload an object" section in the Get started by using the OSS console topic.

   A dataset can be in the TSV or TEXT format and contains the following columns: ID column, text column, question column, answer column (optional), start position column (optional), and title column (optional).

   In this example, a TSV file is used to show how to train a model.

2. Use the Read File Data -3 component to read the prediction dataset. Set the OSS Data Path parameter of the Read File Data component to the OSS path in which the prediction dataset is stored.

3. Connect the Read File Data-3 component to the machine reading comprehension predict component as an upstream node and configure the machine reading comprehension predict component. For more information, see the "Component parameters" section of this topic.

References

• You can use this component together with the machine reading comprehension training component. For more information, see machine reading comprehension training.

• For more information about Machine Learning Designer components, see Designer overview.

• Machine Learning Designer provides various preset algorithm components that you can use based on your business scenario. For more information, see Designer component reference.