Hi there, 

we are trying to enroll our US business location for CSP indirect reseller (for our DE location we are successfully registered and enrolled). 

I created an Entra tenant and used the enrollment form, but I fail when completing the form to kick everything off. I receive the below error message: 

We have one central website, but it won’t accept the entry. What can i provide to make this work? 

I can not even open a support request, because I end up in a closed form when i follow the red link : 

Any recommendations and ideas are really welcome. 

Thanks 

Ann

​Hi there, we are trying to enroll our US business location for CSP indirect reseller (for our DE location we are successfully registered and enrolled). I created an Entra tenant and used the enrollment form, but I fail when completing the form to kick everything off. I receive the below error message:  We have one central website, but it won’t accept the entry. What can i provide to make this work? I can not even open a support request, because I end up in a closed form when i follow the red link :  Any recommendations and ideas are really welcome. Thanks Ann  Read More

​

I have to format a Surface 10 Pro Business without using a recovery image but using the Windows 11 key. The SSD disk is not recognized because the Sata controller driver is missing. Can anyone tell me the model or the driver download link? Thank you

​I have to format a Surface 10 Pro Business without using a recovery image but using the Windows 11 key. The SSD disk is not recognized because the Sata controller driver is missing. Can anyone tell me the model or the driver download link? Thank you  Read More

​

Bonjour,

je suis en windows 10 22h2 build 19045.2787

quelles sont les manips pour l’upgrader, car il est figé .

Merci

​Bonjour,je suis en windows 10 22h2 build 19045.2787quelles sont les manips pour l’upgrader, car il est figé .Merci    Read More

​

The Future of AI: Distillation Just Got Easier

Part 3 – Deploying your LoRA Fine-tuned Llama 3.1 8B model, why it’s a breeze!

Learn how Azure AI makes it effortless to deploy your LoRA fine-tuned models using Azure AI.  (🚀🔥 Github recipe repo).

By Cedric Vidal, Principal AI Advocate, Microsoft

Part of the Future of AI 🚀 series initiated by Marco Casalaina with his Exploring Multi-Agent AI Systems blog post.

A Llama on a rocket launched in space, generated using Azure OpenAI DALL-E 3

Welcome back to our series on leveraging Azure AI Studio to accelerate your AI development journey. In our previous posts, we’ve explored synthetic dataset generation and the process of fine-tuning models. Today, we’re diving into the crucial step that turns your hard work into actionable insights: deploying your fine-tuned model. In this installment, we’ll guide you through deploying your model using Azure AI Studio and the Python SDK, ensuring a seamless transition from development to production.

Why Deploying GPU Accelerated Inference Workloads is Hard

Deploying GPU-accelerated inference workloads comes with a unique set of challenges that make the process significantly more complex compared to standard CPU workloads. Below are some of the primary difficulties encountered:

GPU Resource Allocation: GPUs are specialized and limited resources, requiring precise allocation to avoid wastage and ensure efficiency. Unlike CPUs that can be easily provisioned in larger numbers, the specialized nature of GPUs means that effective allocation strategies are crucial to optimize performance.
GPU Scaling: Scaling GPU workloads is inherently more challenging due to the high cost and limited availability of GPU resources. it requires careful planning to balance cost efficiency with workload demands, unlike more straightforward CPU resource scaling.
Load Balancing for GPU Instances: Implementing load balancing for GPU-based tasks is complex due to the necessity of evenly distributing tasks across available GPU instances. This step is vital to prevent bottlenecks, avoid overload in certain instances, and ensure optimal performance of each GPU unit.
Model Partitioning and Sharding: Large models that cannot fit into a single GPU memory require partitioning and sharding. This process involves splitting the model across multiple GPUs, which introduces additional layers of complexity in terms of load distribution and resource management.
Containerization and Orchestration: While containerization simplifies the deployment process by packaging models and dependencies, managing GPU resources within containers and orchestrating them across nodes adds another layer of complexity. Effective orchestration setups need to be fine-tuned to handle the subtle dynamics of GPU resource utilization and management.
LoRA Adapter Integration: LoRA, which stands for Low-order Rank Adaptation, is a powerful optimization technique that reduces the number of trainable parameters by decomposing the original weight matrices into lower-rank matrices. This makes it efficient for fine-tuning large models with fewer resources. However, integrating LoRA adapters into deployment pipelines involves additional steps to efficiently store, load and merge the lightweight adapters with the base model and serve the final model, which increases the complexity of the deployment process.
Monitoring GPU Inference Endpoints: Monitoring GPU inference endpoints is complex due to the need for specialized metrics to capture GPU utilization, memory bandwidth, and thermal limits, not to mention model specific metrics such as token counts or request counts. These metrics are vital for understanding performance bottlenecks and ensuring efficient operation but require intricate tools and expertise to collect and analyze accurately.
Model Specific Considerations: It’s important to acknowledge that the deployment process is often specific to the base model architecture you are working with. Each new version of a model or a different model vendor will require a fair amount of adaptations in your deployment pipeline. This could include changes in preprocessing steps, modifications in environment configurations, or adjustments in the integration or versions of third-party libraries. Therefore, it’s crucial to stay updated with the model documentation and vendor-specific deployment guidelines to ensure a smooth and efficient deployment process.
Model Versioning Complexity: Keeping track of multiple versions of a model can be intricate. Each version may exhibit distinct behaviors and performance metrics, necessitating thorough evaluation to manage updates, rollbacks, and compatibility with other systems. We’ll cover the subject of model evaluation more thoroughly in the next blog post. Another difficulty with versioning is storing the weights of the different LoRA adapters and keeping track of the versions of the base models they must be adapted onto.
Cost Planning: Planning the costs for GPU inference workloads is challenging due to the variable nature of GPU usage and the higher costs associated with GPU resources. Predicting the precise amount of GPU time required for inference under different workloads can be difficult, leading to unexpected expenses.

Understanding and addressing these difficulties is crucial for successfully deploying GPU-accelerated inference workloads, ensuring that the full potential of GPU capabilities is harnessed.

Azure AI Serverless: A Game Changer

Azure AI Serverless is a game changer because it effectively addresses a lot of challenges with deploying GPU-accelerated inference workloads. By leveraging the serverless architecture, it abstracts away the complexities associated with GPU resource allocation, model specific deployment considerations, and API management. This means you can deploy your models without worrying about the underlying infrastructure management, allowing you to focus on your application’s needs. Additionally, Azure AI Serverless supports a diverse collection of models and abstracts away the choice and provisioning of GPU hardware accelerators, ensuring efficient and fast inference times. The platform’s integration with managed services enables robust container orchestration, simplifying the deployment process even further and enhancing overall operational efficiency.

Attractive pay as you go cost model

One of the standout features of Azure AI Serverless is its token-based cost model, which greatly simplifies cost planning. With token-based billing, you are charged based on the number of tokens processed by your model, making it easy to predict costs based on expected usage patterns. This model is particularly beneficial for applications with variable loads, as you only pay for what you use.

Because the managed infrastructure needs to maintain LoRA adapters in memory and swap them on demand, there is an additional per hour cost associated with fine tuned serverless endpoints but it is billed by the hour only while the endpoint is being used. This makes it super easy to plan ahead future bills depending on your expected usage profile.

Also, the hourly cost is meant to go down, it already went down dramatically from $3.09/hour for a Llama 2 7B based model to $0.74/hour for a Llama 3.1 8B based model.

By paying attention to these critical factors, you can ensure that your model deployment is robust, secure, and capable of meeting the demands of your application.

Region Availability

When deploying your Llama 3.1 fine-tuned model, it’s important to consider the geographical regions where the model can be deployed. As of now, Azure AI Studio supports the deployment of Llama 3.1 fine-tuned models in the following regions: East US, East US 2, North Central US, South Central US, West US, and West US 3. Choosing a region that’s closer to your end-users can help reduce latency and improve performance. Ensure you select the appropriate region based on your target audience for optimal results.

For the most up-to-date information on region availability for other models, please refer to this guide on deploying models serverlessly.

Let’s get coding with Azure AI Studio and the Python SDK

Before proceeding to deployment, you’ll need a model that you have previously fine-tuned. One way is to use the process described in the two preceding installments of this fine-tuning blog post series: the first one covers synthetic dataset generation using RAFT and the second one covers fine-tuning. This ensures that you can fully benefit from the deployment steps using Azure AI Studio.

Note: All code samples that follow have been extracted from the 3_deploy.ipynb notebook of the raft-recipe GitHub repository. The snippets have been simplified and some intermediary steps left aside for ease of reading. You can either head over there, clone the repo and start experimenting right away or stick with me here for an overview.

Step 1: Set Up Your Environment

First, ensure you have the necessary libraries installed. You’ll need the Azure Machine Learning SDK for Python. You can install it using pip:

pip install azure-ai-ml

Next, you’ll need to import the required modules and authenticate your Azure ML workspace. This is standard, the MLClient is the gateway to the ML Workspace which gives you access to everything AI and ML on Azure.

from azure.ai.ml import MLClient
from azure.identity import (
DefaultAzureCredential,
InteractiveBrowserCredential,
)
from azure.ai.ml.entities import MarketplaceSubscription, ServerlessEndpoint

try:
credential = DefaultAzureCredential()
credential.get_token(“https://management.azure.com/.default”)
except Exception as ex:
credential = InteractiveBrowserCredential()

try:
client = MLClient.from_config(credential=credential)
except:
print(“Please create a workspace configuration file in the current directory.”)

# Get AzureML workspace object.
workspace = client._workspaces.get(client.workspace_name)
workspace_id = workspace._workspace_id

Step 2: Resolving the previously registered fine-tuned model

Before deploying, you need to resolve your fine-tuned model in the Azure ML workspace.

Since the fine-tuning job might still be running, you may want to wait for the model to be registered, here’s a simple helper function you can use.

def wait_for_model(client, model_name):
“””Wait for the model to be available, typically waiting for a finetuning job to complete.”””
import time

attempts = 0
while True:
try:
model = client.models.get(model_name, label=”latest”)
return model
except:
print(f”Model not found yet #{attempts}”)
attempts += 1
time.sleep(30)

The above function is basic but will make sure your deployment can proceed as soon as your model becomes available.

print(f”Waiting for fine tuned model {FINETUNED_MODEL_NAME} to complete training…”)
model = wait_for_model(client, FINETUNED_MODEL_NAME)
print(f”Model {FINETUNED_MODEL_NAME} is ready”)

Step 3: Subscribe to the model provider

Before deploying a model fine-tuned using a base model from a third-party non-Microsoft source, you need to subscribe to the model provider’s marketplace offering. This subscription allows you to access and use the model within Azure ML.

print(f”Deploying model asset id {model_asset_id}”)

from azure.core.exceptions import ResourceExistsError
marketplace_subscription = MarketplaceSubscription(
model_id=base_model_id,
name=subscription_name,
)

try:
marketplace_subscription = client.marketplace_subscriptions.begin_create_or_update(marketplace_subscription).result()
except ResourceExistsError as ex:
print(f”Marketplace subscription {subscription_name} already exists for model {base_model_id}”)

Details on how to construct the base_model_id and subscription_name are available in the 3_deploy.ipynb notebook.

Step 4: Deploy the model as a serverless endpoint

This section manages the deployment of a serverless endpoint for your fine-tuned model using the Azure ML client. It checks for an existing endpoint and creates one if it doesn’t exist, then proceeds with the deployment.

from azure.core.exceptions import ResourceNotFoundError
try:
serverless_endpoint = client.serverless_endpoints.get(endpoint_name)
print(f”Found existing endpoint {endpoint_name}”)
except ResourceNotFoundError as ex:
serverless_endpoint = ServerlessEndpoint(name=endpoint_name, model_id=model_asset_id)
serverless_endpoint = client.serverless_endpoints.begin_create_or_update(serverless_endpoint).result()

print(“Waiting for deployment to complete…”)
serverless_endpoint = ServerlessEndpoint(name=endpoint_name, model_id=model_id)

created_endpoint = client.serverless_endpoints.begin_create_or_update(serverless_endpoint).result()
print(“Deployment complete”)

Step 5: Check that the endpoint is correctly deployed

As part of a deployment pipeline, it is a good practice to include integration tests that check that the model is correctly deployed and fails fast instead of waiting for steps down the line to fail without context.

import requests

url = f”{endpoint.scoring_uri}/v1/chat/completions”

prompt = “What do you know?”
payload = {
“messages”:[ { “role”:”user”,”content”: prompt } ],
“max_tokens”:1024
}
headers = {“Content-Type”: “application/json”, “Authorization”: endpoint_keys.primary_key}

response = requests.post(url, json=payload, headers=headers)

response.json()

This code assumes that the deployed model is a chat model for simplicity. The code available in the 3_deploy.ipynb notebook is more generic and will cover both completion and chat models.

Conclusion

Deploying your fine-tuned model with Azure AI Studio and the Python SDK not only simplifies the process but also empowers you with unparalleled control, ensuring you have a robust and reliable platform for your deployment needs.

Stay tuned for our next blog post, in two weeks we will delve into assessing the performance of your deployed model through rigorous evaluation methodologies. Until then, head out to the Github repo and happy coding!

​Microsoft Tech Community – Latest Blogs –Read More 

Salam

I have a question regarding official training for Exchange Server 2019. I’m aware that Microsoft offers various training materials for its products, but I couldn’t find any official course specifically designed for Exchange 2019 in the catalog.

Could someone confirm if there was ever an official training course or certification for Exchange Server 2019? I’ve seen training for previous versions like Exchange 2010, but it seems like there wasn’t anything equivalent for 2019. Any clarification would be appreciated.

​SalamI have a question regarding official training for Exchange Server 2019. I’m aware that Microsoft offers various training materials for its products, but I couldn’t find any official course specifically designed for Exchange 2019 in the catalog.Could someone confirm if there was ever an official training course or certification for Exchange Server 2019? I’ve seen training for previous versions like Exchange 2010, but it seems like there wasn’t anything equivalent for 2019. Any clarification would be appreciated.  Read More

​

On a beautiful day in Korea, we brought together a diverse group of Microsoft MVPs (Most Valuable Professional), MLSAs (Microsoft Learn Student Ambassadors), RDs (Regional Directors), Microsoft employees, and guests from Japan to create a truly inclusive and inspiring event: Korea Influencer Day. The gathering aimed to build cross-border connections and foster collaboration while empowering communities with shared knowledge and tech trends. With a carefully crafted agenda, we succeeded in sparking meaningful conversations among university students, community leaders, and professionals.

In this post, we’ll walk through the event highlights and share best practices on how to organize inclusive in-person community events. We will also reflect on the valuable feedback received to inspire others to create impactful community gatherings.

Memorable Moments and Reflections

1. Inspiring Cross-Cultural Exchange

A defining feature of the event was the meaningful collaboration between Korean and Japanese MVPs. Kazuyuki Miyake, Japanese Microsoft Azure MVP and RD, and Ryota Nakamura, Japanese Business Applications MVP, introduced their local community trends to Korean community leaders.

Kazuyuki shared his experiences and said, “Participating in Influencer Day in Korea was a milestone. Sharing insights from Japan’s AOAI Dev Day that I successfully organized and proposing the next edition in Seoul marked great progress. I believe collaboration between Microsoft MVPs and RDs can spark a powerful movement. I was especially impressed by the proactive Korean Microsoft Learn Student Ambassadors, whose enthusiasm and curiosity promise a bright future.”

2. Networking through Speed Dating: A Surprising Success

Initially met with hesitation, the speed dating session turned out to be a highlight. It encouraged conversations between individuals from different backgrounds, leading to insights and connections that may not have otherwise emerged. MLSAs engaged with MVPs, attendees shared cultural perspectives between Korea and Japan, and discussions sparked about future collaborations.

JinSeok Kim, a Korean Developer Technologies MVP, who also played a key role as a translator between Korean and Japanese attendees, offered valuable feedback for future events: “While the format encouraged organic interaction, some feedback suggested adding conversation starters or a topic-drawing activity to make it easier for shy participants to dive into meaningful discussions.” 

Atsushi Yokohama, an AI Platform MVP from Japan, visited Seoul for the first time to connect with community leaders in Korea. He shared his experience of the event, saying, “It was my first time interacting with Microsoft MVPs from Korea, but I’m grateful to have been able to engage in friendly technical discussions with all of them. This experience has definitely boosted my motivation. I now feel inspired to help strengthen community interactions across Asia.”

3. Empowering the Next Generation of Leaders

The event provided invaluable exposure for Korean MLSA students, whose energy and curiosity left a lasting impression. Many expressed their ambition to grow within the community, including one MLSA student, Minseok Song’s newly formed goal to achieve GOLD MLSA status this year after attending the event.

He continued his reflections and said, “At the event, I asked several questions while talking with the MVPs, and everyone was kind enough to explain things, making it a productive and rewarding experience for me. These conversations inspired me to become someone who can help others, just like you and the MVPs.” This reflection shows how inclusive events can inspire future leaders by connecting them with role models and mentors.

4. Female Tech Influencers and Expanding Community Impact

One of the most impactful sessions was the speech by female tech influencers, highlighting the importance of diversity and gender inclusiveness in the tech space. Representation matters, and hearing from these leaders not only inspired attendees but also promoted the idea that diverse voices are key to creating a thriving tech ecosystem.

The panel discussion on increasing community impact through collaboration also underscored the potential of generative AI to transform communities across Korea and Japan, opening doors for future joint initiatives.

SungHo You, Microsoft Technical Trainer and Justin Yoo, Microsoft Cloud Advocate who participated in the event, shared their thoughts: “The Korea Influencer Day was a pivotal event for the Korean developer community. It brought together diverse community leaders, fostering meaningful interactions, empathy, and moments of joy, especially with Japanese MVPs. I want to particularly commend the efforts to promote gender diversity within the Microsoft tech community, which was positively influenced by the collaboration between Microsoft and the SA team.”

Best Practices for Organizing Inclusive In-Person Events

Drawing on the success of Korea Influencer Day, here are some key practices to consider when planning inclusive events:

Curate a Diverse Agenda

Ensure that the schedule reflects a range of topics and speakers from various backgrounds, including professionals, students, and community leaders.
Highlight underrepresented voices, such as female tech leaders or community members from different regions or fields.

Design for Interactivity and Connection

Incorporate speed networking sessions or icebreaker activities to foster interaction among attendees from different backgrounds.
Use creative formats like Show & Tell or small-group discussions to encourage knowledge sharing.

Provide Conversation Starters or Prompts

Offer topic cards or a discussion board to spark conversations, helping participants break the ice during networking sessions.
Create personalized introductions to connect individuals based on shared interests.

Make Cross-Cultural Exchange a Priority

If attendees come from diverse regions or countries, include sessions that promote cultural understanding, such as cultural exchange talks or panels discussing shared challenges and solutions.

Support Newcomers and Aspiring Leaders

Engage with students and newcomers, offering mentorship opportunities to help them grow within the community.
Recognize and celebrate their achievements to encourage continued participation.

Balance Structure with Flexibility

While structured agendas are important, allow time for unstructured networking to enable organic connections and deeper conversations.

Gather and Act on Feedback

Ask attendees for feedback to understand what worked well and where improvements can be made.
Implement these learnings in future events to enhance inclusiveness and engagement.

Korea Influencer Day sparked creativity through stories of personal tech projects to inspiring students to become future leaders, the event demonstrated the value of bringing people together across cultures, backgrounds, and interests.

By designing events that celebrate diversity, foster interaction, and empower individuals, we can create meaningful experiences that have a lasting impact on communities. Whether you’re organizing a small community meetup or a large-scale event, the lessons from Korea Influencer Day can guide you in creating an environment where everyone feels welcome, heard, and inspired to contribute.

What’s next? As one participant from Japan suggested, we can look forward to taking place in Seoul. Until then, let’s continue building bridges and sharing knowledge to shape the future together.

​Microsoft Tech Community – Latest Blogs –Read More 

Hello,

I need help with a few things.

1. When someone fills out the form (on the right), it then populates into the list (on the left). When I comment on the populated row within the list, the user I am tagging is not getting notifications. How can I fix this? I do not want anyone else having access to the full list, just my comment.

2. Is there a way to create a workflow for approvals? I.e. if it’s a primary buyer I want these 3 people to approve in order, but if it’s an investor then I want these 4 people to approve in order?

​Hello, I need help with a few things. 1. When someone fills out the form (on the right), it then populates into the list (on the left). When I comment on the populated row within the list, the user I am tagging is not getting notifications. How can I fix this? I do not want anyone else having access to the full list, just my comment. 2. Is there a way to create a workflow for approvals? I.e. if it’s a primary buyer I want these 3 people to approve in order, but if it’s an investor then I want these 4 people to approve in order?    Read More

​

I am adding data sources to our collections in Purview so I can start doing scans for our Data Catalog.  I have added 5 data sources already and installed the Azure Arc extension and registered the data source as a SQL Server (Not arc enable just a regular SQL Server) and I have run some scans on those data sources.

Recently, I added another data source and just registered it as a regular SQL Server and skipped the Azure arc extension.

Do I have to install the extension for the scans to work?  Is the extension just for if I wanted to manage the SQL Servers, which are on Prem, in the Azure environment…which I don’t at this time.

Thank you

​I am adding data sources to our collections in Purview so I can start doing scans for our Data Catalog.  I have added 5 data sources already and installed the Azure Arc extension and registered the data source as a SQL Server (Not arc enable just a regular SQL Server) and I have run some scans on those data sources. Recently, I added another data source and just registered it as a regular SQL Server and skipped the Azure arc extension. Do I have to install the extension for the scans to work?  Is the extension just for if I wanted to manage the SQL Servers, which are on Prem, in the Azure environment…which I don’t at this time. Thank you  Read More

​

Hello Community, 

I am having issues integrating Jira with Sentinel. I have connected the logic app to Jira API. Now the logic app is not flowing. I receive a 406 error code during a step in the logic. I made sure the project was the same as listed in Jira. I even tried using the project key. 

​Hello Community,  I am having issues integrating Jira with Sentinel. I have connected the logic app to Jira API. Now the logic app is not flowing. I receive a 406 error code during a step in the logic. I made sure the project was the same as listed in Jira. I even tried using the project key.        Read More

​

Microsoft partners like Open Lake and Zafin deliver transact-capable offers, which allow you to purchase directly from Azure Marketplace. Learn about these offers below:

Compliance Process Automation for Microsoft Teams: Designed for businesses of all sizes, Open Lake Technology’s Compliance Process Automation is an end-to-end supervision and analytics tool designed to ensure your company is compliant with financial regulations including MiFID II and Dodd-Frank. Built on Microsoft Teams, the tool includes automatic monitoring, real-time alerts, auditability, and a clear overview of your compliance status.

Compliance Recording for Microsoft Teams: This turnkey and managed solution deploys Open Lake Technology’s Compliance Recording to your business environment. Compliance Recording integrates with Microsoft Teams, Open Lake’s Compliance Automated Process, on-premises telephon systems, trading telephony systems, and legacy systems. With this solution, you can record chat, voice, video, and screen sharing to ensure full compliance.

A logo reading Zafin

Zafin Cloud: Zafin enables financial institutions to design and manage pricing, products, and packages while simplifying and modernizing core banking systems. Zafin Cloud, built on Microsoft Azure, enables financial institutions to modernize, augment, and extend their core banking technologies. The platform includes a catalog of customer-centric products and services, a personalized reward system, unified data, and much more.

​Microsoft Tech Community – Latest Blogs –Read More