Merge pull request #6816 from voutilad/global-training-ga

prmerger-automator[bot] · web-flow · commit 26ec4ecdd9e1 · 2025-09-22T15:50:50.000Z
Release Global Training for Fine Tuning as Generally Available
diff --git a/articles/ai-foundry/openai/includes/fine-tune-models.md b/articles/ai-foundry/openai/includes/fine-tune-models.md
@@ -18,7 +18,7 @@ ms.custom:
 > The supported regions for fine-tuning might vary if you use Azure OpenAI models in an Azure AI Foundry project versus outside a project.
 >
 
-|  Model ID  | Standard training regions | Global training (preview) | Max request (tokens) | Training data (up to) | Modality |
+|  Model ID  | Standard training regions | Global training | Max request (tokens) | Training data (up to) | Modality |
 |  --- | --- | :---: | :---: | :---: | --- |
 | `gpt-35-turbo` <br> (1106) | East US2 <br> North Central US <br> Sweden Central <br> Switzerland West | - | Input: 16,385<br> Output: 4,096 |  Sep 2021 | Text to text |
 | `gpt-35-turbo` <br> (0125)  | East US2 <br> North Central US <br> Sweden Central <br> Switzerland West | - | 16,385 | Sep 2021 | Text to text |
@@ -30,7 +30,7 @@ ms.custom:
 | `o4-mini` <br> (2025-04-16) | East US2 <br> Sweden Central | - | Input: 128,000 <br> Output: 16,384 <br> Training example context length: 65,536 | May 2024 | Text to text |
 
 > [!NOTE]
-> Global training (in preview) provides [more affordable](https://aka.ms/aoai-pricing) training per token, but doesn't offer [data residency](https://aka.ms/data-residency). It's currently available to Azure OpenAI resources in the following regions, with more regions coming soon:
+> Global training provides [more affordable](https://aka.ms/aoai-pricing) training per token, but doesn't offer [data residency](https://aka.ms/data-residency). It's currently available to Azure OpenAI resources in the following regions:
 >
 >- Australia East
 >- Brazil South
diff --git a/articles/ai-foundry/openai/includes/fine-tuning-python.md b/articles/ai-foundry/openai/includes/fine-tuning-python.md
@@ -139,14 +139,13 @@ After you upload your training and validation files, you're ready to start the f
 
 The following Python code shows an example of how to create a new fine-tune job with the Python SDK:
 
-In this example we are also passing the seed parameter. The seed controls the reproducibility of the job. Passing in the same seed and job parameters should produce the same results, but may differ in rare cases. If a seed isn't specified, one will be generated for you.
-
 ```python
 response = client.fine_tuning.jobs.create(
     training_file=training_file_id,
     validation_file=validation_file_id,
-    model="gpt-4.1-2025-04-14", # Enter base model name. Note that in Azure OpenAI the model name contains dashes and cannot contain dot/period characters.
-    seed = 105  # seed parameter controls reproducibility of the fine-tuning job. If no seed is specified one will be generated automatically.
+    model="gpt-4.1-2025-04-14", # Enter base model name.
+    suffix="my-model", # Custom suffix for naming the resulting model. Note that in Azure OpenAI the model cannot contain dot/period characters.
+    seed=105, # seed parameter controls reproducibility of the fine-tuning job. If no seed is specified one will be generated automatically.
 )
 
 job_id = response.id
@@ -159,9 +158,24 @@ print("Status:", response.id)
 print(response.model_dump_json(indent=2))
 ```
 
+If you are fine tuning a model that supports [Global Training](../concepts/models.md#fine-tuning-models), you can specify the training type by using the `extra_body` named argument:
+
+```python
+response = client.fine_tuning.jobs.create(
+    training_file=training_file_id,
+    validation_file=validation_file_id,
+    model="gpt-4.1-2025-04-14",
+    suffix="my-model",
+    seed=105,
+    extra_body={ "trainingType": "globalstandard" }
+)
+
+job_id = response.id
+```
+
 You can also pass additional optional parameters like hyperparameters to take greater control of the fine-tuning process. For initial training we recommend using the automatic defaults that are present without specifying these parameters. 
 
-The current supported hyperparameters for fine-tuning are:
+The current supported hyperparameters for Supervised Fine-Tuning are:
 
 |**Name**| **Type**| **Description**|
 |---|---|---|
@@ -170,7 +184,7 @@ The current supported hyperparameters for fine-tuning are:
 |`n_epochs` | integer | The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset. |
 |`seed` | integer |	The seed controls the reproducibility of the job. Passing in the same seed and job parameters should produce the same results, but may differ in rare cases. If a seed isn't specified, one will be generated for you. |
 
-To set custom hyperparameters with the 1.x version of the OpenAI Python API:
+To set custom hyperparameters with the 1.x version of the OpenAI Python API, provide them as part of the `method`:
 
 ```python
 from openai import OpenAI
@@ -182,13 +196,24 @@ client = OpenAI(
 
 client.fine_tuning.jobs.create(
   training_file="file-abc123", 
-  model="gpt-4.1-2025-04-14", # Enter base model name. Note that in Azure OpenAI the model name contains dashes and cannot contain dot/period characters.
-  hyperparameters={
-    "n_epochs":2
+  model="gpt-4.1-2025-04-14",
+  suffix="my-model",
+  seed=105,
+  method={
+    "type": "supervised", # In this case, the job will be using Supervised Fine Tuning.
+    "supervised": {
+      "hyperparameters": {
+        "n_epochs": 2
+      }
+    }
   }
 )
 ```
 
+> [!NOTE]
+> See the guides for [Direct Preference Optimization](../how-to/fine-tuning-direct-preference-optimization.md) and [Reinforcement Fine-Tuning](../how-to/reinforcement-fine-tuning.md) to learn more about their supported hyperparameters.
+
+
 ## Check fine-tuning job status
 
 ```python
diff --git a/articles/ai-foundry/openai/includes/fine-tuning-rest.md b/articles/ai-foundry/openai/includes/fine-tuning-rest.md
@@ -119,7 +119,7 @@ curl -X POST $AZURE_OPENAI_ENDPOINT/openai/v1/files \
 
 ## Create a customized model
 
-After you uploaded your training and validation files, you're ready to start the fine-tuning job. The following code shows an example of how to [create a new fine-tuning job](/rest/api/azureopenai/fine-tuning/create?view=rest-azureopenai-2023-12-01-preview&tabs=HTTP&preserve-view=true) with the REST API.
+After you uploaded your training and validation files, you're ready to start the fine-tuning job. The following code shows an example of how to [create a new fine-tuning job](/rest/api/azureopenai/fine-tuning/create?view=rest-azureopenai-2024-10-21&tabs=HTTP&preserve-view=true) with the REST API.
 
 In this example we are also passing the seed parameter. The seed controls the reproducibility of the job. Passing in the same seed and job parameters should produce the same results, but can differ in rare cases. If a seed is not specified, one will be generated for you.
 
@@ -129,15 +129,30 @@ curl -X POST $AZURE_OPENAI_ENDPOINT/openai/v1/fine_tuning/jobs \
   -H "api-key: $AZURE_OPENAI_API_KEY" \
   -d '{
     "model": "gpt-4.1-2025-04-14",
-    "training_file": "<TRAINING_FILE_ID>", 
+    "training_file": "<TRAINING_FILE_ID>",
     "validation_file": "<VALIDATION_FILE_ID>",
     "seed": 105
 }'
 ```
 
-You can also pass additional optional parameters like [hyperparameters](/rest/api/azureopenai/fine-tuning/create?view=rest-azureopenai-2023-12-01-preview&tabs=HTTP#finetuninghyperparameters&preserve-view=true) to take greater control of the fine-tuning process. For initial training we recommend using the automatic defaults that are present without specifying these parameters.
+If you are fine tuning a model that supports [Global Training](../concepts/models.md#fine-tuning-models), you can specify the training type by using the `extra_body` named argument and using api-version `2025-04-01-preview`:
+
+```bash
+curl -X POST $AZURE_OPENAI_ENDPOINT/openai/fine_tuning/jobs?api-version=2025-04-01-preview \
+  -H "Content-Type: application/json" \
+  -H "api-key: $AZURE_OPENAI_API_KEY" \
+  -d '{
+    "model": "gpt-4.1-2025-04-14",
+    "training_file": "<TRAINING_FILE_ID>",
+    "validation_file": "<VALIDATION_FILE_ID>",
+    "seed": 105,
+    "trainingType": "globalstandard"
+}'
+```
+
+You can also pass additional optional parameters like [hyperparameters](/rest/api/azureopenai/fine-tuning/create?view=rest-azureopenai-2024-10-21&tabs=HTTP#finetuninghyperparameters&preserve-view=true) to take greater control of the fine-tuning process. For initial training we recommend using the automatic defaults that are present without specifying these parameters.
 
-The current supported hyperparameters for fine-tuning are:
+The current supported hyperparameters for Supervised Fine-Tuning are:
 
 |**Name**| **Type**| **Description**|
 |---|---|---|
@@ -146,6 +161,9 @@ The current supported hyperparameters for fine-tuning are:
 |`n_epochs` | integer | The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset. |
 |`seed` | integer | The seed controls the reproducibility of the job. Passing in the same seed and job parameters should produce the same results, but may differ in rare cases. If a seed isn't specified, one will be generated for you. |
 
+> [!NOTE]
+> See the guides for [Direct Preference Optimization](../how-to/fine-tuning-direct-preference-optimization.md) and [Reinforcement Fine-Tuning](../how-to/reinforcement-fine-tuning.md) to learn more about their supported hyperparameters.
+
 ## Check the status of your customized model
 
 After you start a fine-tune job, it can take some time to complete. Your job might be queued behind other jobs in the system. Training your model can take minutes or hours depending on the model and dataset size. The following example uses the REST API to check the status of your fine-tuning job. The example retrieves information about your job by using the job ID returned from the previous example:
