This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(176, 27%, 27%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EAF%3C/text%3E%3C/svg%3E)
I'm trying to access a Llama 3.2 11b Vision model, deployed on a dedicated endpoint, but the sample code does not work:
This is the error on logs:
When using azure.ai lib it was not successful either:
Any thoughts?
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(6.4, 61%, 10%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ES%3C/text%3E%3C/svg%3E)
Hi @Adriano Freitas,
To help troubleshoot the issue could you share the region where you’ve deployed the endpoint? Additionally, if you paste the sample code you’re using, I can assist you more specifically with any modifications that might be needed.
Hi @santoshkc , thanks for your reply.
The region is centralus, and the sample code is the first image. This is exactly the same code provided under the 'Consume' tab, on 'Endpoints'.
Hi @Adriano Freitas,
Thank you for your confirmation. Region "centralus" is not supported. Supported regions are swedencentral, eastus2, eastus, westus3, northcentralus, westus, southcentralus.
I hope you understand. Thank you.
Hi @santoshkc
This is correct for Serverless API, but I'm using the Managed Compute option:
The endpoint is deployed. I have both URL and API key, as shown in previous images. The sample code I've posted in the previous message came from the 'Consume' tab from the created endpoint, same for the log print, which came from 'Logs' tab from the created endpoint.
The problem is not related on endpoint creation, but on consumption.
Hi @Adriano Freitas,
Thank you for your follow-up query.
If you're using Managed Compute for Llama 3.2 Vision and facing issues with consumption, it might be related to how you're making the request. The sample code from the 'Consume' tab should work, double-check that the method and request body are formatted correctly. Pay attention to any error messages returned during the request, as they might provide clues like incorrect headers or payloads.
The "invalid input data" error suggests that the data you're sending to the Llama 3.2 Vision endpoint doesn't match the expected format. Ensure that the input (e.g., image) is correctly encoded, either as a base64 string or through a URL, depending on the endpoint’s requirements. Still if you face any issue, I recommend you raise a support request through the Azure portal.
Thank you.