update Gradio from 4.x -> latest (5.xx) & check compatibility carefully

@imbajin i am working on this

update Gradio from 4.x -> latest (5.xx) & check compatibility carefully

@imbajin i am working on this

@Aryankb THX, some points to note:

1. We need better component support for our streaming output conversations, in 4.x it's not good (Current effect could refer feat(llm): add streaming output feature #172)
2. It's been almost half a year now, what are the major updates/performance improvements in this area?
3. Keep an eye on the official "Breaking Change" & confirm if we have any related issues

update Gradio from 4.x -> latest (5.xx) & check compatibility carefully

@imbajin i am working on this

@Aryankb Excuse me, any progress on this mission?

Because we are currently undergoing a workflow transformation and preparing to introduce a workflow framework, we would like to complete the upgrade of Gradio's major version first to reduce potential conflicts/security issues in the future

@imbajin ok sir. sorry for the delay. Will work on it today. I was busy in my college projects and intern works

@imbajin just wanted to share and know if this gradio update helpful? it happened just 4 days ago and mentions something about streaming

pls refer :- gradio-app/gradio#10688

also can you please guide me how can i test streaming?

@imbajin just wanted to share and know if this gradio update helpful? it happened just 4 days ago and mentions something about streaming

pls refer :- gradio-app/gradio#10688

also can you please guide me how can i test streaming?

Thanks for your info, maybe we do need test to ensure it

After #190 merged? (or u could use that PR's branch)

@imbajin should i use #190 PR's branch? and test it. also how can i test it? can i test directly from ui? if yes which buttons should i click? can you please give a basic overview of this? also i am reading wiki to know all features

@imbajin should i use #190 PR's branch? and test it. also how can i test it? can i test directly from ui? if yes which buttons should i click? can you please give a basic overview of this? also i am reading wiki to know all features

Yes, u're right, clicking the answer button in UI is enough~

Can you please give a basic overview of this?

@MrJs133 maybe add the "Quick Start" operation doc in Wiki first?

@imbajin should i use #190 PR's branch? and test it. also how can i test it? can i test directly from ui? if yes which buttons should i click? can you please give a basic overview of this? also i am reading wiki to know all features

Yes, u're right, clicking the answer button in UI is enough~

> Can you please give a basic overview of this?

@MrJs133 maybe add the "Quick Start" operation doc in Wiki first?

OK I will do it today (Sorry, I accidentally clicked theclose with comment button)

@imbajin should i use #190 PR's branch? and test it. also how can i test it? can i test directly from ui? if yes which buttons should i click? can you please give a basic overview of this? also i am reading wiki to know all features

@Aryankb Already merged into main/master branch (could test it directly)

@imbajin ok sir, thanks. Just wanted to know, what all things i need to configure before that button click?

• fill open ai api key (can i control the token usage?)
• any more steps ?

facing the following error on clicking that button:-

1. fill open ai api key (can i control the token usage?)
2. any more steps ?

2. Yes, u could control it here & could also test it only in raw answer first

BTW, no need to use OpenAI key for testing, you can use any free model completely (By LiteLLM option, see it in README)

2. Nope, but it's best to test the situation where both options are checked at the same time, as this will result in parallel output

@imbajin facing this error on button click

facing the following error on clicking that button:-

i think the error is related to hugegraph server configuration

i think the error is related to hugegraph server configuration

Yes, u need to follow the README's instruction to run a graph server first & config it (apply the configs)

https://github.com/apache/incubator-hugegraph-ai/tree/main/hugegraph-llm#2-environment-requirements

@MrJs133 will add the basic UI instruction soon~

Yes, u need to follow the README's instruction to run a graph server first & config it (apply the configs)

Yes i did everything. Server is running on docker

it worked. actually it's required to load into graph first. then only hugegraph RAG query works

i think the error is related to hugegraph server configuration

Yes, u need to follow the README's instruction to run a graph server first & config it (apply the configs)

main/hugegraph-llm#2-environment-requirements

@MrJs133 will add the basic UI instruction soon~

@Aryankb Already add a basic doc for the UI & logic (U could enhance it later~)

refer https://github.com/apache/incubator-hugegraph-ai/blob/main/hugegraph-llm/quick_start.md