Merge pull request #217 from ruanzhengxin-zhuxing/master

Add pai plugin tmpl
aliyun · Apr 26, 2024 · e164599 · e164599
2 parents 81d7d84 + 768af3c
commit e164599
Show file tree

Hide file tree

Showing 3 changed files with 96 additions and 2 deletions.
diff --git a/README-CN.md b/README-CN.md
@@ -452,7 +452,8 @@ ROS 模板的示例和最佳实践。模板分类如下：
 | [build-a-dialogue-model-based-on-ChatGLM-and-LangChain.yml](documents/solution/ai/build-a-dialogue-model-based-on-ChatGLM-and-LangChain.yml)                   | 基于ChatGLM和LangChain搭建对话模型。｜ [解决方案](https://www.aliyun.com/solution/tech-solution/calbadm)         |
 | [build-a-dialogue-model-based-on-Qwen-and-LangChain.yml](documents/solution/ai/build-a-dialogue-model-based-on-Qwen-and-LangChain.yml)                   | 基于通义千问和LangChain搭建对话模型。                                                                           |
 | [ack-ai-fine-tuning.yml](documents/solution/ai/ack-ai-fine-tuning.yml)      | 使用云原生AI套件提交模型微调训练任务与部署GPU共享推理服务。  ｜ [解决方案](https://www.aliyun.com/solution/tech-solution/one_gpu) |
-| [exclusive-qa-service.yml](exclusive-qa-service.yml)     | 向量检索与通义千问搭建专属问答服务。  |
+| [exclusive-qa-service.yml](documents/solution/ai/exclusive-qa-service.yml)     | 向量检索与通义千问搭建专属问答服务。  |
+| [discover-and-reach-high-value-users.yml](documents/solution/ai/discover-and-reach-high-value-users.yml)     | 高价值用户挖掘及触达  |
 
 - data-analysis
 

diff --git a/README.md b/README.md
@@ -454,7 +454,8 @@ Examples and best practices of ROS templates. The templates are categorized as f
 | [build-a-dialogue-model-based-on-ChatGLM-and-LangChain.yml](documents/solution/ai/build-a-dialogue-model-based-on-ChatGLM-and-LangChain.yml)                   | Build a dialogue model based on ChatGLM and LangChain.                                       |
 | [build-a-dialogue-model-based-on-Qwen-and-LangChain.yml](documents/solution/ai/build-a-dialogue-model-based-on-Qwen-and-LangChain.yml)                   | Build a dialogue model based on Qwen and LangChain.                                       |
 | [ack-ai-fine-tuning.yml](documents/solution/ai/ack-ai-fine-tuning.yml)  | Use the cloud-native AI suite to submit model fine-tuning training tasks and deploy GPU shared inference services.     |
-| [exclusive-qa-service.yml](exclusive-qa-service.yml)     | DashVector and Tongyi QianWen build exclusive Q&A service.     |
+| [exclusive-qa-service.yml](documents/solution/ai/exclusive-qa-service.yml)     | DashVector and Tongyi QianWen build exclusive Q&A service.     |
+| [discover-and-reach-high-value-users.yml](documents/solution/ai/discover-and-reach-high-value-users.yml)     | Discover and reach high value users.  |
 
 
 

diff --git a/documents/solution/ai/discover-and-reach-high-value-users.yml b/documents/solution/ai/discover-and-reach-high-value-users.yml
@@ -0,0 +1,92 @@
+ROSTemplateFormatVersion: '2015-09-01'
+Parameters:
+  TrainingDataPath:
+    Type: String
+    Label:
+      zh-cn: 训练输入数据OSS路径
+      en: Training Job Data OSS Path
+    Description:
+      zh-cn: 选择之前步骤中上传至OSS的训练数据数据目录。
+      en: Select the data directory uploaded to OSS in the previous step.
+  InferenceDataPath:
+    Type: String
+    Label:
+      zh-cn: 推理输入数据OSS路径
+      en: Inference Job Data OSS Path
+    Description:
+      zh-cn: 选择之前步骤中上传至OSS的人群圈选任务的输入数据路径
+      en: Select the input data path of the crowd selection task uploaded to OSS in the previous step.
+  TargetOutputPath:
+    Type: String
+    Label:
+      zh-cn: 人群圈选任务输出数据路径
+      en: Outputs Data Path
+    Description:
+      zh-cn: 输出数据所在的地域和OSS目录。输出数据路径需要为空目录，否则可能会导致输出目录中存在多个任务的人群文件。
+      en: OSS directory where the output data is located. The output data path needs to be an empty directory, otherwise crowd files for multiple tasks may exist in the output directory.
+Resources:
+  Campaign:
+    Type: ALIYUN::PAIPlugin::Campaign
+    Properties:
+      Name: 
+        Fn::Sub: tech_solu_campaign_name-${ALIYUN::StackId}
+  TrainingJob:
+    Type: ALIYUN::PAIPlugin::TrainingJob
+    Properties:
+      Algorithm: sms_recall
+      CampaignId:
+        Ref: Campaign
+      DataPath: 
+        Ref: TrainingDataPath
+      UserConfig:
+        end_date: "20220208"
+        op_target: sms
+        start_date: "20220101"
+      WaitForTrainingFinish: true
+      Name:
+        Fn::Sub: tech_solu_training_job_name-${ALIYUN::StackId}
+  InferenceJob:
+    Type: ALIYUN::PAIPlugin::InferenceJob
+    Properties:
+      Algorithm: sms_recall
+      CampaignId:
+        Ref: Campaign
+      DataPath:
+        Ref: InferenceDataPath
+      UserConfig:
+        percentage_threshold: 0.5
+        exe_date: "20220201"
+      WaitForInferenceFinish: true
+      TrainingJobId:
+        Ref: TrainingJob
+      TargetPath:
+        Ref: TargetOutputPath
+      Name:
+        Fn::Sub: tech_solu_inference_job_name-${ALIYUN::StackId}
+Outputs:
+  GroupId:
+    Description:
+      zh-cn: 圈选出的人群ID
+      en: Circled crowd ID
+    Value:
+      Fn::GetAtt:
+      - InferenceJob
+      - GroupId
+Metadata:
+  ALIYUN::ROS::Interface:
+    ParameterGroups:
+      - Parameters:
+        - TrainingDataPath
+        Label:
+          default:
+            en: Training Job
+            zh-cn: 策略训练任务
+      - Parameters:
+        - InferenceDataPath
+        - TargetOutputPath
+        Label:
+          default:
+            en: Inference Job
+            zh-cn: 人群圈选任务
+    TemplateTags:
+      - acs:technical-solution:ai:高价值用户挖掘及触达-tech_solu_17