Getting started with Azure Machine Learning Prompt Flow
HTML-код
- Опубликовано: 7 июн 2024
- Getting started with Azure Machine Learning Prompt Flow, this will get you running with AML Prompt Flow, stay tuned for a future video where i will be discussing Prompt Flow in Azure AI Studio :)
troubleshooting link mentioned in video;
learn.microsoft.com/en-us/azu...
This video is incredible, exactly what I needed and so clearly explained - thank you!
Exactly what i needed. thanks!
Excellent Demo, thanks.
I have never added a comment on any video before, but this is really amazing and exactly what i was looking for. Thanks for sharing. well done.
Awesome, thank you for the feedback!
This is really fantastic recipe. Thank you for recording this!
This is extraordinary content. Extremely well done.
This is Awesome! Thanks.
Thanks a lot. Bumped across Prompt flow in Azure AI studio today without having any idea. Loved your video. Can you please tell me why we'd need a text embedding model?
Awesome
how do I update my vector index with additional data?
I used a similar prompt as yours however the model is still answering out of context questions. Is there a way to ensure that the model strictly answers the topics related to the docs in Azure AI?
Very nice session
but one question Can't we use azure alone instead of azure open AI ?
is there any alternative than using Azure Open AI as its asks for registering etc.,
if you are using a free azure account, then most probably you are not able to create an azure open ai service.
Will the system automatically change the API versions when necessary?
learn.microsoft.com/en-us/azure/ai-services/openai/how-to/working-with-models?tabs=powershell
Is it possible to control the chunking strategy in azure promt
Not if you are using the Vector Index Wizard
Can we creat a interface for end user to upload a doc and do RAG in Azure ?
Yes you can, but you will need to write all the code yourself, security, read the docs in, chunking, run through forms recognizer if needed, process through embedding model, store in vector db etc...
@@shepMLgreat video, do you plan to do any other videos to create a UI and show this flow? It would be really useful
Getting this error can you help
Prompt flow deployment "ml-workspace-service-xyz" of endpoint "ml-workspace-service-xyz" failed
Creating deployment
Hide details
ResourceNotReady: User container has crashed or terminated: Liveness probe failed: