You can use this component to perform offline prediction to EasyRec models.
Prerequisites
You have trained a model by using the Model Training component. For more information, see Model Training.
Configure the component in Machine Learning Designer
Input ports
Port (from left to right) | Recommended upstream component | Parameters of PAI commands | Required |
Input model |
| saved_model_dir | Yes |
Input table |
| input_table | Yes |
Component parameters
Tab | Parameter | Required | Description | Parameters of PAI commands | Default value |
Parameters Setting | Input Feature Columns | No | The feature columns that are selected from the input table for training. You cannot specify this parameter and the Excluded Columns parameter at the same time. | selected_cols | No default value |
Excluded Columns | No | The columns that you want to exclude from the input table. You cannot specify this parameter and the Input Feature Columns parameter at the same time. | excluded_cols | No default value | |
Reserved Columns | No | The names of the generated columns that you want to reserve. | reserved_cols | No default value | |
Output Columns | No | The output columns in the MaxCompute table. | output_cols | probs double | |
miniBatch size | No | The size of the minibatch, which indicates the minimum number of samples contained in a batch. | batch_size | 1024 | |
Specify the algorithm version | Yes | This parameter is available if you select the Advanced Options. Select an algorithm package to run.
| script | No default value | |
Tuning | Worker Count | No | The number of worker nodes. | Passed in as the cluster parameter. | 1 |
Worker CPU | No | The CPU number of each worker. A value of 1 indicates one CPU core. | 6 | ||
Worker Memory | No | The memory size of each worker node. A value of 100 indicates 100 MB. | 30000 | ||
Worker GPU | No | GPUs are not required in most EasyRec trainings. | 0 |
Output ports
Port (from left to right) | Data type | Parameters of PAI commands |
Output table | MaxCompute table | output_table |
PAI command and parameters
PAI -project algo_public -name easy_rec_ext
-Darn="acs:ram::xxx:role/aliyunodpspaidefaultrole"
-Dbatch_size="1024"
-Dbuckets="oss://rec_sln_demo/"
-Dcluster="{\"worker\": {\"count\": 1, \"cpu\": 600, \"gpu\": 0, \"memory\": 30000}}"
-Dcmd="predict"
-Dinput_table="odps://pai_hangzhou/tables/pai_temp_flow_inpwi02on49ooub78p_node_dn3y3lvucm862jr71n_outputTable"
-Dlifecycle="28"
-DossHost="oss-cn-hangzhou-internal.aliyuncs.com"
-Doutput_cols="item_emb string"
-Doutput_table="odps://pai_hangzhou/tables/pai_temp_flow_da1nuzwmbdfyw5kajy_node_5jgko0vlrjiwawp6y8_outputTable"
-Dreserved_cols="item_id"
-Dsaved_model_dir="oss://rec_sln_demo/EasyRec/deploy/rec_sln_demo_dssm_recall_v1/20230425/export/final/item"
-Dscript="oss://rec_sln_demo/easy_rec_ext_0.6.1_res.tar.gz";Parameter | Required | Description |
saved_model_dir | Yes | The directory of the exported model files. |
input_table | Yes | The name of the input table. |
output_table | No | The output table that is automatically created. |
reserved_cols | No | The columns to be copied to the output_table. Separate multiple columns with commas (,). |
output_cols | No | The names and data types of the columns in the output table. Separate multiple columns with commas (,). |
batch_size | No | The size of minibatch. |
arn | Yes | To obtain the value of the arn, perform the following operations: Log on to the PAI console. In the left-side navigation pane, choose Activation & Authorization > Dependent Services. In the Designer section, find OSS and click View Authorization in the Actions column. |
buckets | Yes | The bucket where the model file resides and the bucket where the model is stored. If multiple buckets are used, separate multiple buckets with commas (,). Example: |
ossHost | Yes | The endpoint of OSS. For more information about endpoints, see Regions and endpoints. |
script | No | The path of the OSS bucket where the EasyRec TAR package is stored. For more information about how to configure the EasyRec TAR package, see Release & Upgrade in the EasyRec documentation. |
Example
Create a pipeline as shown in the following figure.
Section
Description
① ② ③
For more information about how to configure the parameters, see Examples in Model Training.
④
Connect the Model Training component to the left input port of Model Prediction, and connect the Read Table-2 component to the right input port of Model Prediction. Set Exclude Column to clk and select user_id and pid for Reserved Columns.
After the pipeline is run, right-click the Model Prediction component and choose View Data > Output Table.
For more information, see 13_rec_sln_demo_dssm_recall_item_embedding_v1 in Vector Recall.