[Feature] One-Click Deployment for GenAI Examples

[zhiweizhangintc]

trafficstars

Priority

P2-High

OS type

Ubuntu

Hardware type

Gaudi2

Running nodes

Single Node

Description

Based on feedback from enterprise developers, OPEA will introduce "One-Click Deployment" scripts in the v1.4 release to simplify setup for the following 8 GenAI examples. These scripts should support both Docker and Kubernetes environments to minimize developer effort.

Please avoid duplication between test scripts(deployment part) and "One-Click Deployment" scripts.

[joshuayao]

Hi @lvliang-intel It would be great if we have a RFC first, thanks.

[poussa]

We already have two one-click scripts, one in Enterprise-RAG repo and another in Enterprise-Inference repo. Can we align, please?

[zhiweizhangintc]

@joshuayao @lvliang-intel The 8 GenAI examples newly created as sub-feature are not consistent to the 8 GenAI Examples from issue ticket i created on March 27th (e.g. AgentQnA is missing). please double check and correct. Thanks.

[joshuayao]

@joshuayao @lvliang-intel The 8 GenAI examples newly created as sub-feature are not consistent to the 8 GenAI Examples from issue ticket i created on March 27th (e.g. AgentQnA is missing). please double check and correct. Thanks.

Hi @zhiweizhangintc Thanks for your information. This is the ticket you filed. Which component should be removed from the scope after adding AgentQnA?

[joshuayao]

We already have two one-click scripts, one in Enterprise-RAG repo and another in Enterprise-Inference repo. Can we align, please?

@lvliang-intel

[zhiweizhangintc]

@joshuayao @lvliang-intel The 8 GenAI examples newly created as sub-feature are not consistent to the 8 GenAI Examples from issue ticket i created on March 27th (e.g. AgentQnA is missing). please double check and correct. Thanks.

Hi @zhiweizhangintc Thanks for your information. This is the ticket you filed. Which component should be removed from the scope after adding AgentQnA?

@joshuayao This is the 8 GenAI Examples i filed on March 27th which we should focus on.

"8 GenAI Examples are: ChatQnA, DocuSum, AgentQnA, Code Gen, Code Tran, FAQGen, Visual QnA, Audio QnA"

[joshuayao]

@joshuayao @lvliang-intel The 8 GenAI examples newly created as sub-feature are not consistent to the 8 GenAI Examples from issue ticket i created on March 27th (e.g. AgentQnA is missing). please double check and correct. Thanks.

Hi @zhiweizhangintc Thanks for your information. This is the ticket you filed. Which component should be removed from the scope after adding AgentQnA?

@joshuayao This is the 8 GenAI Examples i filed on March 27th which we should focus on.

"8 GenAI Examples are: ChatQnA, DocuSum, AgentQnA, Code Gen, Code Tran, FAQGen, Visual QnA, Audio QnA"

Thanks. Updated.

[lvliang-intel]

We already have two one-click scripts, one in Enterprise-RAG repo and another in Enterprise-Inference repo. Can we align, please?

@lvliang-intel Yes, we will stay aligned with E-RAG. The difference is that we will use a unified set of one-click scripts to apply to all examples.

[ZePan110]

DocSum,FaqGen https://github.com/opea-project/GenAIExamples/actions/runs/16064991750/job/45337717474 CodeTrans https://github.com/opea-project/GenAIExamples/actions/runs/16063947380/job/45334922590

[letonghan]

AgentQnA test by @ZePan110 : https://github.com/opea-project/GenAIExamples/actions/runs/16067420884/job/45344464090

[joshuayao]

For AgentQnA, support Gaudi deployment. For the other 7 examples, support both Xeon and Gaudi deployment.