You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: articles/ai-services/openai/concepts/provisioned-throughput.md
+4-3Lines changed: 4 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -41,13 +41,14 @@ An Azure OpenAI Deployment is a unit of management for a specific OpenAI Model.
41
41
## How much throughput per PTU you get for each model
42
42
The amount of throughput (tokens per minute or TPM) a deployment gets per PTU is a function of the input and output tokens in the minute. Generating output tokens requires more processing than input tokens and so the more output tokens generated the lower your overall TPM. The service dynamically balances the input & output costs, so users do not have to set specific input and output limits. This approach means your deployment is resilient to fluctuations in the workload shape.
43
43
44
-
To help with simplifying the sizing effort, the following table outlines the TPM per PTU for the `gpt-4o` and `gpt-4o-mini` models
44
+
To help with simplifying the sizing effort, the following table outlines the TPM per PTU for the `gpt-4o` and `gpt-4o-mini` models. The table also shows Service Level Agreement (SLA) Latency Target Values per model. For more information about the SLA for Azure OpenAI Service, see the [Service Level Agreements (SLA) for Online Services page].(https://www.microsoft.com/licensing/docs/view/Service-Level-Agreements-SLA-for-Online-Services?lang=1)
Copy file name to clipboardExpand all lines: articles/ai-services/openai/overview.md
+20-7Lines changed: 20 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -27,15 +27,28 @@ Azure OpenAI Service provides REST API access to OpenAI's powerful language mode
27
27
| Managed Identity| Yes, via Microsoft Entra ID |
28
28
| UI experience |[Azure portal](https://portal.azure.com) for account & resource management, <br> [Azure AI Studio](https://ai.azure.com) for model exploration and fine-tuning |
29
29
| Model regional availability |[Model availability](./concepts/models.md)|
30
-
| Content filtering | Prompts and completions are evaluated against our content policy with automated systems. High severity content will be filtered. |
30
+
| Content filtering | Prompts and completions are evaluated against our content policy with automated systems. High severity content is filtered. |
31
31
32
32
## Responsible AI
33
33
34
34
At Microsoft, we're committed to the advancement of AI driven by principles that put people first. Generative models such as the ones available in Azure OpenAI have significant potential benefits, but without careful design and thoughtful mitigations, such models have the potential to generate incorrect or even harmful content. Microsoft has made significant investments to help guard against abuse and unintended harm, which includes incorporating Microsoft’s <a href="https://www.microsoft.com/ai/responsible-ai?activetab=pivot1:primaryr6" target="_blank">principles for responsible AI use</a>, adopting a [Code of Conduct](/legal/cognitive-services/openai/code-of-conduct?context=/azure/ai-services/openai/context/context) for use of the service, building [content filters](/azure/ai-services/content-safety/overview) to support customers, and providing responsible AI [information and guidance](/legal/cognitive-services/openai/transparency-note?context=%2Fazure%2Fai-services%2Fopenai%2Fcontext%2Fcontext&tabs=image) that customers should consider when using Azure OpenAI.
35
35
36
-
## How do I get access to Azure OpenAI?
36
+
## Get started with Azure OpenAI Service
37
37
38
-
A Limited Access registration form is not required to access most Azure OpenAI models. Learn more on the [Azure OpenAI Limited Access page](/legal/cognitive-services/openai/limited-access?context=/azure/ai-services/openai/context/context).
38
+
To get started with Azure OpenAI Service, you need to create an Azure OpenAI Service resource in your Azure subscription.
39
+
40
+
Start with the [Create and deploy an Azure OpenAI Service resource](./how-to/create-resource.md) guide.
41
+
1. You can create a resource via Azure portal, Azure CLI, or Azure PowerShell.
42
+
1. When you have an Azure OpenAI Service resource, you can deploy a model such as GPT-4o.
43
+
1. When you have a deployed model, you can:
44
+
45
+
- Try out the Azure AI Studio playgrounds to explore the capabilities of the models.
46
+
- You can also just start making API calls to the service using the REST API or SDKs.
47
+
48
+
For example, you can try [real-time audio](./realtime-audio-quickstart.md) and [assistants](./assistants-quickstart.md) in the playgrounds or via code.
49
+
50
+
> [!NOTE]
51
+
> A Limited Access registration form is required to access some Azure OpenAI Service models or features. Learn more on the [Azure OpenAI Limited Access page](/legal/cognitive-services/openai/limited-access?context=/azure/ai-services/openai/context/context).
39
52
40
53
## Comparing Azure OpenAI and OpenAI
41
54
@@ -47,7 +60,7 @@ With Azure OpenAI, customers get the security capabilities of Microsoft Azure wh
47
60
48
61
### Prompts & completions
49
62
50
-
The completions endpoint is the core component of the API service. This API provides access to the model's text-in, text-out interface. Users simply need to provide an input **prompt** containing the English text command, and the model will generate a text **completion**.
63
+
The completions endpoint is the core component of the API service. This API provides access to the model's text-in, text-out interface. Users simply need to provide an input **prompt** containing the English text command, and the model generates a text **completion**.
51
64
52
65
Here's an example of a simple prompt and completion:
53
66
@@ -70,7 +83,7 @@ Here's an example of a simple prompt and completion:
70
83
71
84
Azure OpenAI processes text by breaking it down into tokens. Tokens can be words or just chunks of characters. For example, the word “hamburger” gets broken up into the tokens “ham”, “bur” and “ger”, while a short and common word like “pear” is a single token. Many tokens start with a whitespace, for example “ hello” and “ bye”.
72
85
73
-
The total number of tokens processed in a given request depends on the length of your input, output and request parameters. The quantity of tokens being processed will also affect your response latency and throughput for the models.
86
+
The total number of tokens processed in a given request depends on the length of your input, output, and request parameters. The quantity of tokens being processed will also affect your response latency and throughput for the models.
74
87
75
88
#### Image tokens
76
89
@@ -106,9 +119,9 @@ Once you create an Azure OpenAI Resource, you must deploy a model before you can
106
119
107
120
### Prompt engineering
108
121
109
-
The GPT-3, GPT-3.5 and GPT-4 models from OpenAI are prompt-based. With prompt-based models, the user interacts with the model by entering a text prompt, to which the model responds with a text completion. This completion is the model’s continuation of the input text.
122
+
The GPT-3, GPT-3.5, and GPT-4 models from OpenAI are prompt-based. With prompt-based models, the user interacts with the model by entering a text prompt, to which the model responds with a text completion. This completion is the model’s continuation of the input text.
110
123
111
-
While these models are extremely powerful, their behavior is also very sensitive to the prompt. This makes [prompt engineering](./concepts/prompt-engineering.md) an important skill to develop.
124
+
While these models are powerful, their behavior is also sensitive to the prompt. This makes [prompt engineering](./concepts/prompt-engineering.md) an important skill to develop.
112
125
113
126
Prompt construction can be difficult. In practice, the prompt acts to configure the model weights to complete the desired task, but it's more of an art than a science, often requiring experience and intuition to craft a successful prompt.
Copy file name to clipboardExpand all lines: articles/ai-studio/includes/create-hub.md
+9-2Lines changed: 9 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -10,17 +10,24 @@ ms.date: 5/21/2024
10
10
ms.custom: include, build-2024
11
11
---
12
12
13
+
> [!NOTE]
14
+
> A hub in Azure AI Studio is a one-stop shop where you manage everything your AI project needs, like security and resources, so you can develop and test faster. To learn more about how hubs can help you, see the [Hubs and projects overview](/azure/ai-studio/concepts/ai-resources) article.
15
+
13
16
To create a hub in [Azure AI Studio](https://ai.azure.com), follow these steps:
14
17
15
18
1. Go to the **Home** page in [Azure AI Studio](https://ai.azure.com) and sign in with your Azure account.
16
-
1. Select **All hubs**from the left pane and then select **+ New hub**.
19
+
1. Select **All resources**on the left pane. If you cannot see this option, in the top bar select **All resources & projects**. Then select **+ New hub**.
17
20
18
21
:::image type="content" source="../media/how-to/hubs/hub-new.png" alt-text="Screenshot of the button to create a new hub." lightbox="../media/how-to/hubs/hub-new.png":::
19
22
20
-
1. In the **Create a new hub** dialog, enter a name for your hub (such as *contoso-hub*) and select **Next**. Leave the default **Connect Azure AI Services** option selected. A new AI services connection is created for the hub.
23
+
1. In the **Create a new hub** dialog, enter a name for your hub (such as *contoso-hub*). If you don't have a resource group, a new **Resource group** will be created linked to the **Subscription** provided. Leave the default **Connect Azure AI Services** option selected.
24
+
1. Select **Next**. If you didn't reuse an existing resource group, a new resource group (*rg-contoso*) is created. Also an Azure AI service (*ai-contoso-hub*) is created for the hub.
21
25
22
26
:::image type="content" source="../media/how-to/hubs/hub-new-connect-services.png" alt-text="Screenshot of the dialog to connect services while creating a new hub." lightbox="../media/how-to/hubs/hub-new-connect-services.png":::
23
27
28
+
> [!NOTE]
29
+
> If you don't see (new) before the **Resource group** and **Connect Azure AI Services** entries then an existing resource is being used. For the purposes of this tutorial, create a seperate entity via **Create new resource group** and **Create new AI Services**. This will allow you to prevent any unexpected charges by deleting the entities after the tutorial.
30
+
24
31
1. Review the information and select **Create**.
25
32
26
33
:::image type="content" source="../media/how-to/hubs/hub-new-review-create.png" alt-text="Screenshot of the dialog to review the settings for the new hub." lightbox="../media/how-to/hubs/hub-new-review-create.png":::
Copy file name to clipboardExpand all lines: articles/machine-learning/concept-compute-target.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -92,7 +92,7 @@ When created, these compute resources are automatically part of your workspace,
92
92
> [!NOTE]
93
93
> To avoid charges when the compute is idle:
94
94
> * For a compute *cluster*, make sure the minimum number of nodes is set to 0, or use [serverless compute](./how-to-use-serverless-compute.md).
95
-
> * For a compute *instance*, [enable idle shutdown](how-to-create-compute-instance.md#configure-idle-shutdown).
95
+
> * For a compute *instance*, [enable idle shutdown](how-to-create-compute-instance.md#configure-idle-shutdown). While stopping the compute instance stops the billing for compute hours, you'll still be billed for disk, public IP, and standard load balancer.
0 commit comments