Closed mashwishi closed 1 year ago
Is this a question about importing a chat?
About understanding the usage of the responseID
, choiceID
, and _reqID
..
Yes, you can assume they are necessary.
I'm not too sure what they do, but I'm safely assuming:
responseID
is the actual response you're on in the conversation
choiceID
is which response you chose out of the possible responses
and _reqID
is probably some form of a index to keep track of responses
As for useJSON
, it's mostly for the advanced image information.
Any other questions? Closing otherwise.
Do i have to manually put it ?
Do i have to manually put it ?
No, It's optional. It is used when you want to begin new chat that is linked to previous conversations (So Bard able to understand the context, or you can link to a custom prompts)
Yes, please read the documentation. Closing.
I am not asking for "optional" or not.
What i am pointing here is is it "OKAY" to apply "conversationID" or it is required to use responseID, choiceID, and _reqID.
That's the question here, As can you see the issue title is Use cases, I know it is optional but it isn't declare if all of this are required to be in the post request.
Now, Do i have to manually specify the "value" of the responseID, choiceID, and _reqID.
You know I am from google bard ai module and it only require us to put "conversationID" the reason i have this kind of question, I don't think this question is included in the docs. Well export is in different section while this topic points the "creating a new conversation"
Oh... So if I understand correctly:
You came from GoogleBard, which only gave you a conversationID... now, you've migrated to bard-ai
, you're asking how responseID, choiceID, and _reqID should be filled in.
From the code I have right now, that's not possible... however in Version 2, we may address this issue... if only 1 ID is required, the better!
Please stay tuned, especially to #30.
Thank you, This will help us lessen our code process!
Okay, after a little bit of digging through GoogleBard code, here's how it works: The so-called "conversationID" in GoogleBard is a abstracted ID (that technically doesn't mean anything to Bard)... GoogleBard takes that ID, and associates it to a JSON with all of the real IDs that Bard uses. These real IDs are split into four:
conversationID: string
responseID: string
choiceID: string
_reqID: string
Because of this, to find these real IDs, you're going to have to look through the JSON files that GoogleBard generates.
These are the IDs that bard-ai
gives, and are the IDs that are able to be used universally across all "unofficial" Google Bard APIs. You may need to migrate some things, but honestly I believe it's for the better if you do it in this way.
@ThatXliner @Aldhanekaa @RFS-ADRENO What do y'all think?
I personally would like to keep it raw.
I am quite confuse about:
I can understand the
conversationID
however why do we need theresponseID
,choiceID
, and_reqID
we need brief explanation such us what does it do, how important it is.conversationID
based on what I understand this is the one that keeps or holds the entire conversation. but I am lost atresponseID
,choiceID
, and_reqID
. Can you enlighten me about this? or is it possible to useconversationID
id only? because i have conversationID via supabase database that group them together and maybe i can use that to specify which conversation they are talking right?Also how do you guys use the
useJSON
if ever i use the advanced way? do i have to include it on themyConversationContinued
thing to i still have to do it next to the message?