Ty for the Video. How can i use my database as source? I already have the database in Cognitive search and i also created indexes. But where can i tell the app that it should use it instead of the pdf's.
May be this sample will work for you. Make sure to select cognitive search and select the column that contains your data. ruclips.net/video/-sr44ZldZoI/видео.html
I am uploading new file directly to storage, but its not coming in search. Only was was to add to the data folder in github and run azd up. Is there a way to run processing for just one file.
Current solution do not support. But you can do that using azure functions that gets triggered when you upload a new file in storage account to extract data from form recognizer and send to cognitive search.
For uploading all files at once it does work, but if I want to add file after deployment simply uploading the file into container doesn't work i.e ai is not able to search through the doc. So, how one can add/update documents?
What was the cost of this? I see you only used one document so the Azure AI Document Intelligence service to process so it would be like 10 dollars for the month correct? Thanks in advance.
@@Dewiride Error: () Unknown field 'embedding' in vector field list. Parameter name: vectorFields Code: Message: Unknown field 'embedding' in vector field list. Parameter name: vectorFields
has anyone encountered the error on line 35 and 39 of prepdocs.ps1: This command cannot be run due to error: The system cannot find the file specified? we have tried to make the updates described in this video to change python to python3 but still getting the same error message.
Python to python3 must only be changed only for mac users and only if alias of python is not created. Go to your terminal and just type python and see if it works. If it works, then you need not change prepdocs.
@@Dewiride sir , the chatbot which you have created in the video("Azure Open AI Search Demo: Custom Enterprise Data for ChatGPT Setup | Train your Own Data") i want to save chathistory from this chat bot
The idea is same and even implementation will be same. Ok, let me create a separate video on that. But Azure Storage is not recommended. I will do on Azure Cosmos DB.
I can't deploy azure openai fo my subscription as it needs form to fill up and it reject when I am using my personal subscription. How can we request this to try the scenario out?
Hi, great video, my company has provided me with a resource group, how can i use that resource group and resources to run this repo locally, i have seen the explaination in github repo but it is still very confusing as im new to this, thanks
Do you know how i can integerate this bot with my main website? Like a pop up bot? It’s already deployed on the azure backend but i want to integerate it to my main website also it keeps crashing when i hit it with many too many requests. Any way to optimise it?
We have deployed the backend using azure and have a frontend team. When they tried to run frontend app separately, it was running but was getting internal error on auth_setup call.
It requires azure subscription plus access to Azure open AI service. Cost is based on individual services that get created. Following services are used - cognitive search, form recognizer, web app, azure open ai. You can have a look at their pricing and calculate based on your requirement. The cost will depend on the pricing tier you choose.
Love your content❤. Sir please make a video on creating a chatbot with our own data set.and intregate that chatbot in MS teams. Sir please make a video on it.🙏
I uploaded some documents then i want to deleted few documents from them but still answering the questions from the deleted documents. how to remove them from the vector database? And i don’t some services from it how to delete services? If I delete it through azure does it cause any error in gpt?
Ty for the Video. How can i use my database as source? I already have the database in Cognitive search and i also created indexes. But where can i tell the app that it should use it instead of the pdf's.
May be this sample will work for you. Make sure to select cognitive search and select the column that contains your data. ruclips.net/video/-sr44ZldZoI/видео.html
This uses Azure SQL DB connected to Cognitive Search - ruclips.net/video/mtXxtSZsIpI/видео.html
Followed everything - when i try to run it online, I get an openai error, when i run it locally, I get a Failed to start backend error - any ideas?
can you change the gpt3.5 turbo to gpt4? I tried doing that but its failing everytime I try to deploy using azd up
Make sure gpt-4 is deployed with the name you are using during the deployment.
how to include prompts? thanks
I am uploading new file directly to storage, but its not coming in search. Only was was to add to the data folder in github and run azd up. Is there a way to run processing for just one file.
Current solution do not support. But you can do that using azure functions that gets triggered when you upload a new file in storage account to extract data from form recognizer and send to cognitive search.
For uploading all files at once it does work, but if I want to add file after deployment simply uploading the file into container doesn't work i.e ai is not able to search through the doc. So, how one can add/update documents?
Hoping this sample might work for you - ruclips.net/video/bO0oGtu_gwo/видео.html
What was the cost of this? I see you only used one document so the Azure AI Document Intelligence service to process so it would be like 10 dollars for the month correct? Thanks in advance.
Correct. Cost depends on the pages you read.
Thanks for the video
but i have a question
How can i connect this demo with mysql database?
i have 2 question
Getting an error regarding the index...can anyone help me,?
What error?
@@Dewiride Error: () Unknown field 'embedding' in vector field list.
Parameter name: vectorFields
Code:
Message: Unknown field 'embedding' in vector field list.
Parameter name: vectorFields
do you get the error while deploying the solution?
The model is up and running but it is taking information from online sources as well. What to do?
Modify the system prompt to only look for information from sources.
has anyone encountered the error on line 35 and 39 of prepdocs.ps1: This command cannot be run due to error: The system cannot find the file specified? we have tried to make the updates described in this video to change python to python3 but still getting the same error message.
Python to python3 must only be changed only for mac users and only if alias of python is not created. Go to your terminal and just type python and see if it works. If it works, then you need not change prepdocs.
Hello sir,
Do you know how ro save chatbot history in the azure storage
We already created a video on this - ruclips.net/video/0joaIgHjbEk/видео.html.
@@Dewiride sir , the chatbot which you have created in the video("Azure Open AI Search Demo: Custom Enterprise Data for ChatGPT Setup | Train your Own Data") i want to save chathistory from this chat bot
The idea is same and even implementation will be same. Ok, let me create a separate video on that. But Azure Storage is not recommended. I will do on Azure Cosmos DB.
@@Dewiride ok sir , thankyou
@@Dewiride Hello sir, can i know when the video will be released.
Note: Azure Open AI Region South Central US has issues during deployment. Please select East US.
I can't deploy azure openai fo my subscription as it needs form to fill up and it reject when I am using my personal subscription. How can we request this to try the scenario out?
Apply here using business email - go.microsoft.com/fwlink/?linkid=2218724
can i make same app in node js?
Yes, you can.
Hi, great video, my company has provided me with a resource group, how can i use that resource group and resources to run this repo locally, i have seen the explaination in github repo but it is still very confusing as im new to this, thanks
The best way is to follow the instructions mentioned in the README of the repo to run the solution locally.
how can we create an api call for this bot which can be integrated with our website?
There are 2 api calls - ask and chat endpoints. You can use them in your custom website
Do you know how i can integerate this bot with my main website?
Like a pop up bot?
It’s already deployed on the azure backend but i want to integerate it to my main website also it keeps crashing when i hit it with many too many requests. Any way to optimise it?
@@Dewiridejust saw your other video. Will try to implement that and keep you updated. Love your videos bdw.
How can I separate backend and frontend? I want to use only frontend app and backend service is deployed already.
The description has the link to the repo that contains both front end and backend. You can separate them.
We have deployed the backend using azure and have a frontend team. When they tried to run frontend app separately, it was running but was getting internal error on auth_setup call.
Does it take to have a subscription to do this? If so, which ones and how much does it cost?
It requires azure subscription plus access to Azure open AI service. Cost is based on individual services that get created. Following services are used - cognitive search, form recognizer, web app, azure open ai. You can have a look at their pricing and calculate based on your requirement.
The cost will depend on the pricing tier you choose.
@@Dewiride Thx
Can I adjust the UI of the web app
Yes, you can. The UI source code is under app folder.
Repository link in the description.
Love your content❤. Sir please make a video on creating a chatbot with our own data set.and intregate that chatbot in MS teams. Sir please make a video on it.🙏
Sure I will
Integrated with MS Teams - ruclips.net/video/mtXxtSZsIpI/видео.html
Thanks for the video. Could you please also setup a Whatsapp Chat bot powered by Azure Open AI ?
Ok sure
I uploaded some documents then i want to deleted few documents from them but still answering the questions from the deleted documents. how to remove them from the vector database?
And i don’t some services from it how to delete services? If I delete it through azure does it cause any error in gpt?
If you used Indexer, then you define a property that tells the index to not consider and delete on next indexer run.