Microsoft Foundry Models での OpenAI のAzure REST API プレビュー リファレンスの作成

この記事では、Azure OpenAI の推論 REST API エンドポイントについて詳しく説明します。

API specs

Azure OpenAI モデルとリソースの管理と操作は、3 つの主要な API サーフェスに分かれています。

  • Control plane
  • データ プレーン - 作成
  • データ プレーン - 推論

各 API surface/仕様は、Azure OpenAI 機能の異なるセットをカプセル化します。 各 API には、プレビューと安定または一般提供 (GA) API リリースの独自のセットがあります。 現在、プレビュー リリースは毎月のペースで行われる傾向があります。

Important

新しいプレビュー推論 API が追加されました。 詳細については、 API ライフサイクル ガイドを参照してください。

API 最新のプレビュー リリース 最新の GA リリース Specifications Description
Control plane 2025-07-01-preview 2025-06-01 Spec files コントロール プレーン API は、 リソースの作成モデルのデプロイ、その他の上位レベルのリソース管理タスクなどの操作に使用されます。 コントロール プレーンは、Azure Resource Manager、Bicep、Terraform、Azure CLIなどの機能を使用して、to do可能な機能も制御します。
Data plane v1 preview v1 Spec files データ プレーン API は、推論と作成の操作を制御します。

Authentication

Azure OpenAI には、認証に 2 つの方法があります。 API キーまたはMicrosoft Entra IDを使用できます。

  • API キー認証: この種類の認証の場合、すべての API 要求で、api-key HTTP ヘッダーに API キーを含める必要があります。 クイック スタートに、この種類の認証を使用して呼び出しを行う方法に関するガイダンスが用意されています。

  • Microsoft Entra ID 認証: Microsoft Entra トークンを使用して API 呼び出しを認証できます。 認証トークンは、Authorization ヘッダーとして要求に含まれます。 指定するトークンの前に Bearer を付ける必要があります (例: Bearer YOUR_AUTH_TOKEN)。 Microsoft Entra IDに関するハウツー ガイドを読>。

REST API のバージョン管理

サービス API は、api-version クエリ パラメーターを使用してバージョン管理されます。 すべてのバージョンは、YYYY-MM-DD 日付構造に従います。 For example:

POST https://YOUR_RESOURCE_NAME.openai.azure.com/openai/deployments/YOUR_DEPLOYMENT_NAME/chat/completions?api-version=2024-06-01

データ プレーンの作成

この記事の残りの部分では、Azure OpenAI データ プレーン推論仕様 (2025-04-01-preview) のプレビュー リリースについて説明します。

最新の GA API リリースに関するドキュメントをお探しの場合は、最新の GA データ プレーン推論 API に関する記事を参照してください

バッチ - リスト

GET {endpoint}/openai/batches?api-version=2025-04-01-preview

Azure OpenAI リソースによって所有されているすべてのバッチの一覧を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
after クエリ No 文字列 前の改ページ位置の自動修正の要求からの最後のイベントの識別子。
limit クエリ No 整数 取得するバッチの数。 既定値は 20 です。
$filter クエリ No 文字列 フィルター条件を記述する OData 式。
$orderby クエリ No 文字列 並べ替え順序を記述する OData 式。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json BatchesList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/batches?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "has_more": false,
    "data": [
      {
        "object": "batch",
        "id": "batch_72a2792ef7d24ba7b82c7fe4a37e379f",
        "endpoint": "/chat/completions",
        "errors": null,
        "input_file_id": "file-b4fa7277f3eb4722ac18b90f46102c3f",
        "completion_window": "24h",
        "status": "completed",
        "output_file_id": "file-f2ddaf43-b48b-46dd-b264-90da10c7a85b",
        "error_file_id": "file-c3b563b0-ebc7-47da-93e3-a2fa694aef0c",
        "created_at": 1646126127,
        "in_progress_at": 1646126130,
        "expires_at": 1646126170,
        "finalizing_at": 1646126134,
        "completed_at": 1646126136,
        "failed_at": null,
        "expired_at": null,
        "cancelling_at": null,
        "cancelled_at": null,
        "request_counts": {
          "total": 500,
          "completed": 400,
          "failed": 100
        },
        "first_id": "batch_abc123",
        "last_id": "batch_abc456",
        "metadata": {
          "batch_description": "Evaluation job"
        }
      }
    ],
    "object": "list"
  }
}

バッチの作成

POST {endpoint}/openai/batches?api-version=2025-04-01-preview

アップロードされた要求ファイルからバッチを作成して実行します。

応答には、ジョブの状態を含むエンキューされたジョブの詳細が含まれます。

完了すると、結果ファイルの ID が応答に追加されます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: application/json

Name タイプ Description Required Default
completion_window 文字列 バッチを処理する時間枠。 Yes
エンドポイント 文字列 バッチによって使用される API エンドポイント。 Yes
input_blob 文字列 バッチの入力として使用するAzure Storage BLOB の URL。 No
input_file_id 文字列 バッチの入力ファイルの ID。 No
メタデータ オブジェクト バッチにアタッチできるキーと値のペアのセット。 これは、バッチに関する追加情報を構造化された形式で格納する場合に役立ちます。 No
output_expires_after FileExpiresAfter ファイルの有効期限を定義します。 No
output_folder BatchOutputReference 出力を格納するAzure Storage フォルダー。 No

Responses

状態コード: 201

説明: バッチが正常に作成されました。

Content-Type Type Description
application/json Batch

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/batches?api-version=2025-04-01-preview

応答: 状態コード: 201

{
  "headers": {
    "deployment-enqueued-tokens": 1000,
    "deployment-maximum-enqueued-tokens": 5000
  },
  "body": {
    "object": "batch",
    "id": "batch_72a2792ef7d24ba7b82c7fe4a37e379f",
    "endpoint": "/chat/completions",
    "errors": null,
    "input_file_id": "file-b4fa7277f3eb4722ac18b90f46102c3f",
    "completion_window": "24h",
    "status": "validating",
    "output_file_id": null,
    "error_file_id": null,
    "created_at": 1646126127,
    "in_progress_at": null,
    "expires_at": null,
    "finalizing_at": null,
    "completed_at": null,
    "failed_at": null,
    "expired_at": null,
    "cancelling_at": null,
    "cancelled_at": null,
    "request_counts": {
      "total": 0,
      "completed": 0,
      "failed": 0
    },
    "metadata": {
      "batch_description": "Evaluation job"
    }
  }
}

バッチ - 取得

GET {endpoint}/openai/batches/{batch-id}?api-version=2025-04-01-preview

指定されたバッチ ID で指定された 1 つのバッチの詳細を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
batch-id パス Yes 文字列 バッチの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json Batch

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/batches/{batch-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "batch",
    "id": "batch_72a2792ef7d24ba7b82c7fe4a37e379f",
    "endpoint": "/chat/completions",
    "errors": null,
    "input_file_id": "file-b4fa7277f3eb4722ac18b90f46102c3f",
    "completion_window": "24h",
    "status": "completed",
    "output_file_id": "file-f2ddaf43-b48b-46dd-b264-90da10c7a85b",
    "error_file_id": "file-c3b563b0-ebc7-47da-93e3-a2fa694aef0c",
    "created_at": 1646126127,
    "in_progress_at": 1646126130,
    "expires_at": 1646126170,
    "finalizing_at": 1646126134,
    "completed_at": 1646126136,
    "failed_at": null,
    "expired_at": null,
    "cancelling_at": null,
    "cancelled_at": null,
    "request_counts": {
      "total": 500,
      "completed": 400,
      "failed": 100
    },
    "metadata": {
      "batch_description": "Evaluation job"
    }
  }
}

バッチ - キャンセル

POST {endpoint}/openai/batches/{batch-id}/cancel?api-version=2025-04-01-preview

指定されたバッチ ID で指定されたバッチの処理を取り消します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
batch-id パス Yes 文字列 バッチの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: バッチが正常に取り消されました。

Content-Type Type Description
application/json Batch

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/batches/{batch-id}/cancel?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "batch",
    "id": "batch_72a2792ef7d24ba7b82c7fe4a37e379f",
    "endpoint": "/chat/completions",
    "errors": null,
    "input_file_id": "file-b4fa7277f3eb4722ac18b90f46102c3f",
    "completion_window": "24h",
    "status": "cancelling",
    "output_file_id": null,
    "error_file_id": null,
    "created_at": 1646126127,
    "in_progress_at": 1646126130,
    "expires_at": 1646126170,
    "finalizing_at": null,
    "completed_at": null,
    "failed_at": null,
    "expired_at": null,
    "cancelling_at": 1646126138,
    "cancelled_at": null,
    "request_counts": {
      "total": 500,
      "completed": 100,
      "failed": 5
    },
    "metadata": {
      "batch_description": "Evaluation job"
    }
  }
}

格納済み入力候補 - 一覧表示

GET {endpoint}/openai/chat/completions?api-version=2025-04-01-preview

格納されている入力候補の一覧を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
メタデータ クエリ No 文字列 格納された完了の (キー、値) ペアでフィルター処理します。
モデル クエリ No 文字列 モデル名でフィルター処理します。
after クエリ No 文字列 前の改ページ位置の要求から最後に格納された完了の識別子。
limit クエリ No 整数 取得する格納済み完了の数。 既定値は 20 です。
順序 クエリ No 文字列 作成された時刻 (昇順または降順) による結果の順序。 既定値は desc です。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json StoredCompletionList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/chat/completions?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "list",
    "data": [
      {
        "id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
        "model": "gpt-4o-mini-2024-07-18",
        "created": 1738193475,
        "request_id": "e539c139-a97d-4ee1-bde9-3be3d5c6edb5",
        "usage": {
          "total_tokens": 25,
          "completion_tokens": 7,
          "prompt_tokens": 18
        },
        "seed": 123,
        "top_p": 1,
        "temperature": 1,
        "presence_penalty": 0,
        "frequency_penalty": 0,
        "system_fingerprint": "fp_f3927aa00d",
        "metadata": {
          "key_1": "val_1",
          "key_2": "val_2"
        },
        "choices": [
          {
            "index": 0,
            "message": {
              "content": "Hello, how are you?",
              "role": "assistant"
            },
            "finish_reason": "stop"
          }
        ]
      }
    ],
    "total": 1,
    "first_id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
    "last_id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
    "has_more": false
  }
}

格納済み入力候補 - 取得

GET {endpoint}/openai/chat/completions/{stored-completion-id}?api-version=2025-04-01-preview

格納されている完了 ID を指定して、格納されている完了を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
stored-completion-id パス Yes 文字列 格納されている完了の識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json StoredCompletionResponse

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/chat/completions/{stored-completion-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "chat.completion",
    "id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
    "model": "gpt-4o-mini-2024-07-18",
    "created": 1738193475,
    "request_id": "e539c139-a97d-4ee1-bde9-3be3d5c6edb5",
    "usage": {
      "total_tokens": 25,
      "completion_tokens": 7,
      "prompt_tokens": 18
    },
    "seed": 123,
    "top_p": 1,
    "temperature": 1,
    "presence_penalty": 0,
    "frequency_penalty": 0,
    "system_fingerprint": "fp_f3927aa00d",
    "metadata": {
      "key_1": "val_1",
      "key_2": "val_2"
    },
    "choices": [
      {
        "index": 0,
        "message": {
          "content": "Hello, how are you?",
          "role": "assistant"
        },
        "finish_reason": "stop"
      }
    ]
  }
}

格納済み入力候補 - 更新

POST {endpoint}/openai/chat/completions/{stored-completion-id}?api-version=2025-04-01-preview

格納されている完了を、指定された格納済み完了 ID で更新します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
stored-completion-id パス Yes 文字列 格納されている完了の識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: application/json

Name タイプ Description Required Default
メタデータ オブジェクト 追加情報のための任意のキーと値のペア。 No

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json StoredCompletion

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/chat/completions/{stored-completion-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "chat.completion",
    "id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
    "model": "gpt-4o-mini-2024-07-18",
    "created": 1738193475,
    "request_id": "e539c139-a97d-4ee1-bde9-3be3d5c6edb5",
    "usage": {
      "total_tokens": 25,
      "completion_tokens": 7,
      "prompt_tokens": 18
    },
    "seed": 123,
    "top_p": 1,
    "temperature": 1,
    "presence_penalty": 0,
    "frequency_penalty": 0,
    "system_fingerprint": "fp_f3927aa00d",
    "metadata": {
      "key_1": "val_1",
      "key_2": "val_2"
    },
    "choices": [
      {
        "index": 0,
        "message": {
          "content": "Hello, how are you?",
          "role": "assistant"
        },
        "finish_reason": "stop"
      }
    ]
  }
}

格納済み入力候補 - 削除

DELETE {endpoint}/openai/chat/completions/{stored-completion-id}?api-version=2025-04-01-preview

指定された格納済み完了 ID によって、格納されている完了を削除します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
stored-completion-id パス Yes 文字列 格納されている完了の識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 格納された完了が正常に削除されました。

Content-Type Type Description
application/json StoredCompletionDelete

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

DELETE {endpoint}/openai/chat/completions/{stored-completion-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "chat.completion.deleted",
    "id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
    "deleted": true
  }
}

格納済み入力候補 - Getmessages

GET {endpoint}/openai/chat/completions/{stored-completion-id}/messages?api-version=2025-04-01-preview

格納されている完了メッセージを、指定された格納済み完了 ID で取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
stored-completion-id パス Yes 文字列 格納されている完了の識別子。
after クエリ No 文字列 前の改ページ位置の要求から最後に格納された完了メッセージの識別子。
limit クエリ No 整数 取得する格納済み完了メッセージの数。 既定値は 20 です。
順序 クエリ No 文字列 メッセージ インデックスによる結果の順序 (昇順または降順)。 既定値は asc です。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json StoredCompletionMessages

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/chat/completions/{stored-completion-id}/messages?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "list",
    "data": [
      {
        "role": "user",
        "content": "Hello"
      }
    ],
    "total": 1,
    "first_id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
    "last_id": "chatcmpl-AvBCKqg2xqxVHCCEnUB4Bgj1Bjl7Y",
    "has_more": false
  }
}

評価 - リストを取得

GET {endpoint}/openai/evals?api-version=2025-04-01-preview

projectの評価を一覧表示します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
after クエリ No 文字列 前の改ページ調整要求からの最後の評価の識別子。
limit クエリ No 整数 取得する評価の数。
順序 クエリ No 文字列
使用可能な値: ascdesc
タイムスタンプによる評価の並べ替え順序。 昇順には asc を使用し、降順には desc を使用します。
order_by クエリ No 文字列
使用可能な値: created_atupdated_at
Evals は、作成時刻または最終更新時刻で並べ替えることができます。 作成時刻には created_at を使用し、最終更新時刻には updated_at を使用します。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価の一覧

Content-Type Type Description
application/json EvalList

Examples

GET {endpoint}/openai/evals?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "list",
    "data": [
      {
        "object": "eval",
        "id": "eval_6801694950848190b10968bb628b651d",
        "data_source_config": {
          "type": "custom",
          "schema": {
            "type": "object",
            "properties": {
              "item": {
                "type": "object",
                "properties": {
                  "question": {
                    "type": "string"
                  },
                  "A": {
                    "type": "string"
                  },
                  "B": {
                    "type": "string"
                  },
                  "C": {
                    "type": "string"
                  },
                  "D": {
                    "type": "string"
                  },
                  "answer": {
                    "type": "string"
                  }
                }
              }
            },
            "required": [
              "item"
            ]
          }
        },
        "testing_criteria": [
          {
            "name": "string check",
            "type": "string_check",
            "input": "{{sample.output_text}}",
            "reference": "{{item.answer}}",
            "operation": "eq"
          }
        ],
        "name": "Math Quiz",
        "created_at": 1744922953,
        "metadata": {}
      }
    ],
    "first_id": "eval_6801694950848190b10968bb628b651d",
    "last_id": "eval_6801694950848190b10968bb628b651d",
    "has_more": false
  }
}

評価 - 作成

POST {endpoint}/openai/evals?api-version=2025-04-01-preview

モデルのパフォーマンスをテストするために使用できる評価の構造を作成します。 評価は、一連のテスト基準とデータソースです。 評価を作成したら、さまざまなモデルとモデル パラメーターで実行できます。 複数の種類の採点者とデータソースをサポートしています。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 201

Description: OK

Content-Type Type Description
application/json Eval

Examples

POST {endpoint}/openai/evals?api-version=2025-04-01-preview

{
 "name": "Math Quiz",
 "data_source_config": {
  "type": "custom",
  "include_sample_schema": true,
  "item_schema": {
   "type": "object",
   "properties": {
    "question": {
     "type": "string"
    },
    "A": {
     "type": "string"
    },
    "B": {
     "type": "string"
    },
    "C": {
     "type": "string"
    },
    "D": {
     "type": "string"
    },
    "answer": {
     "type": "string"
    }
   }
  }
 },
 "testing_criteria": [
  {
   "type": "string_check",
   "reference": "{{item.answer}}",
   "input": "{{sample.output_text}}",
   "operation": "eq",
   "name": "string check"
  }
 ]
}

応答: 状態コード: 201

{
  "headers": {},
  "body": {
    "object": "eval",
    "id": "eval_6801694950848190b10968bb628b651d",
    "data_source_config": {
      "type": "custom",
      "schema": {
        "type": "object",
        "properties": {
          "item": {
            "type": "object",
            "properties": {
              "question": {
                "type": "string"
              },
              "A": {
                "type": "string"
              },
              "B": {
                "type": "string"
              },
              "C": {
                "type": "string"
              },
              "D": {
                "type": "string"
              },
              "answer": {
                "type": "string"
              }
            }
          }
        },
        "required": [
          "item"
        ]
      }
    },
    "testing_criteria": [
      {
        "name": "string check",
        "type": "string_check",
        "input": "{{sample.output_text}}",
        "reference": "{{item.answer}}",
        "operation": "eq"
      }
    ],
    "name": "Math Quiz",
    "created_at": 1744922953,
    "metadata": {}
  }
}

評価 - 削除

DELETE {endpoint}/openai/evals/{eval-id}?api-version=2025-04-01-preview

評価を削除します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 削除する評価の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価を正常に削除しました。

Content-Type Type Description
application/json オブジェクト

状態コード: 404

説明: 評価が見つかりません。

Content-Type Type Description
application/json Error

Examples

DELETE {endpoint}/openai/evals/{eval-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "eval.deleted",
    "deleted": true,
    "eval_id": "eval_6801694950848190b10968bb628b651d"
  }
}

状態コード: 404

{
  "headers": {},
  "body": {
    "code": "notFound",
    "message": "Evaluation with ID eval_6801694950848190b10968bb628b651d not found."
  }
}

評価 - 取得

GET {endpoint}/openai/evals/{eval-id}?api-version=2025-04-01-preview

ID による評価を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 取得する評価の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価

Content-Type Type Description
application/json Eval

Examples

GET {endpoint}/openai/evals/{eval-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "eval",
    "id": "eval_6801694950848190b10968bb628b651d",
    "data_source_config": {
      "type": "custom",
      "schema": {
        "type": "object",
        "properties": {
          "item": {
            "type": "object",
            "properties": {
              "question": {
                "type": "string"
              },
              "A": {
                "type": "string"
              },
              "B": {
                "type": "string"
              },
              "C": {
                "type": "string"
              },
              "D": {
                "type": "string"
              },
              "answer": {
                "type": "string"
              }
            }
          }
        },
        "required": [
          "item"
        ]
      }
    },
    "testing_criteria": [
      {
        "name": "string check",
        "type": "string_check",
        "input": "{{sample.output_text}}",
        "reference": "{{item.answer}}",
        "operation": "eq"
      }
    ],
    "name": "Math Quiz",
    "created_at": 1744922953,
    "metadata": {}
  }
}

評価 - 更新

POST {endpoint}/openai/evals/{eval-id}?api-version=2025-04-01-preview

評価の特定のプロパティを更新します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 更新する評価の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 更新された評価

Content-Type Type Description
application/json Eval

Examples

POST {endpoint}/openai/evals/{eval-id}?api-version=2025-04-01-preview

{
 "name": "Updated Math Quiz",
 "metadata": {
  "description": "Updated description"
 }
}

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "eval",
    "id": "eval_6801694950848190b10968bb628b651d",
    "data_source_config": {
      "type": "custom",
      "schema": {
        "type": "object",
        "properties": {
          "item": {
            "type": "object",
            "properties": {
              "question": {
                "type": "string"
              },
              "A": {
                "type": "string"
              },
              "B": {
                "type": "string"
              },
              "C": {
                "type": "string"
              },
              "D": {
                "type": "string"
              },
              "answer": {
                "type": "string"
              }
            }
          }
        },
        "required": [
          "item"
        ]
      }
    },
    "testing_criteria": [
      {
        "name": "string check",
        "type": "string_check",
        "input": "{{sample.output_text}}",
        "reference": "{{item.answer}}",
        "operation": "eq"
      }
    ],
    "name": "Updated Math Quiz",
    "created_at": 1744922953,
    "metadata": {
      "description": "Updated description"
    }
  }
}

評価 - Getrunlist

GET {endpoint}/openai/evals/{eval-id}/runs?api-version=2025-04-01-preview

評価の実行結果の一覧を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 実行を取得する評価の ID。
after クエリ No 文字列 前の改ページ要求からの最後の実行の識別子。
limit クエリ No 整数 取得する実行の数。
順序 クエリ No 文字列
使用可能な値: ascdesc
実行の並べ替え順序をタイムスタンプ別に指定します。 昇順には asc を使用し、降順には desc を使用します。 既定値は asc です。
状態 クエリ No 文字列
使用できる値: queuedin_progresscompletedcanceledfailed
状態別に実行をフィルター処理します。 "queued"、"in_progress"、"failed"、"completed"、"canceled" を使用します。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価の実行の一覧

Content-Type Type Description
application/json EvalRunList

Examples

GET {endpoint}/openai/evals/{eval-id}/runs?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "list",
    "data": [
      {
        "object": "eval.run",
        "id": "evalrun_68016a056f0481909b9774447bdd1aa3",
        "eval_id": "eval_6801694950848190b10968bb628b651d",
        "report_url": "https://ai.azure.com/resource/evaluation",
        "status": "queued",
        "model": "gpt-4o-mini",
        "name": "Math quiz",
        "created_at": 1744923141,
        "result_counts": {
          "total": 0,
          "errored": 0,
          "failed": 0,
          "passed": 0
        },
        "per_model_usage": null,
        "per_testing_criteria_results": null,
        "data_source": {
          "type": "completions",
          "source": {
            "type": "file_content",
            "content": [
              {
                "item": {
                  "question": "Find the degree for the given field extension Q(sqrt(2), sqrt(3), sqrt(18)) over Q.",
                  "A": "0",
                  "B": "4",
                  "C": "2",
                  "D": "6",
                  "answer": "B"
                }
              },
              {
                "item": {
                  "question": "Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5.",
                  "A": "8",
                  "B": "2",
                  "C": "24",
                  "D": "120",
                  "answer": "C"
                }
              }
            ]
          },
          "input_messages": {
            "template": [
              {
                "type": "message",
                "role": "system",
                "content": {
                  "text": "Answer the question's with A, B, C, or D."
                }
              },
              {
                "type": "message",
                "role": "user",
                "content": {
                  "text": "Question: {{item.question}} A: {{item.A}} B: {{item.B}} C: {{item.C}} D: {{item.D}}."
                }
              }
            ]
          },
          "model": "gpt-4o-mini",
          "sampling_params": {
            "seed": 42,
            "temperature": 1,
            "top_p": 1
          }
        },
        "error": null,
        "metadata": {}
      }
    ],
    "first_id": "evalrun_68016a056f0481909b9774447bdd1aa3",
    "last_id": "evalrun_68016a056f0481909b9774447bdd1aa3",
    "has_more": false
  }
}

評価 - 実行の作成

POST {endpoint}/openai/evals/{eval-id}/runs?api-version=2025-04-01-preview

新しい評価実行を作成します。 これは、採点を開始するエンドポイントです。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 実行を作成する評価の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 201

説明: 評価の実行が正常に作成されました

Content-Type Type Description
application/json EvalRun

状態コード: 400

説明: 要求が正しくありません (たとえば、eval オブジェクトが見つかりません)

Content-Type Type Description
application/json Error

Examples

POST {endpoint}/openai/evals/{eval-id}/runs?api-version=2025-04-01-preview

{
 "name": "Math quiz",
 "data_source": {
  "type": "completions",
  "source": {
   "type": "file_content",
   "content": [
    {
     "item": {
      "question": "Find the degree for the given field extension Q(sqrt(2), sqrt(3), sqrt(18)) over Q.",
      "A": "0",
      "B": "4",
      "C": "2",
      "D": "6",
      "answer": "B"
     }
    },
    {
     "item": {
      "question": "Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5.",
      "A": "8",
      "B": "2",
      "C": "24",
      "D": "120",
      "answer": "C"
     }
    }
   ]
  },
  "input_messages": {
   "type": "template",
   "template": [
    {
     "type": "message",
     "role": "system",
     "content": {
      "text": "Answer the question's with A, B, C, or D."
     }
    },
    {
     "type": "message",
     "role": "user",
     "content": {
      "text": "Question: {{item.question}} A: {{item.A}} B: {{item.B}} C: {{item.C}} D: {{item.D}}."
     }
    }
   ]
  },
  "model": "gpt-4o-mini",
  "sampling_params": {
   "temperature": 1,
   "top_p": 1,
   "seed": 42
  }
 }
}

応答: 状態コード: 201

{
  "headers": {},
  "body": {
    "object": "eval.run",
    "id": "evalrun_68016a056f0481909b9774447bdd1aa3",
    "eval_id": "eval_6801694950848190b10968bb628b651d",
    "report_url": "https://ai.azure.com/resource/evaluation",
    "status": "queued",
    "model": "gpt-4o-mini",
    "name": "Math quiz",
    "created_at": 1744923141,
    "result_counts": {
      "total": 0,
      "errored": 0,
      "failed": 0,
      "passed": 0
    },
    "per_model_usage": null,
    "per_testing_criteria_results": null,
    "data_source": {
      "type": "completions",
      "source": {
        "type": "file_content",
        "content": [
          {
            "item": {
              "question": "Find the degree for the given field extension Q(sqrt(2), sqrt(3), sqrt(18)) over Q.",
              "A": "0",
              "B": "4",
              "C": "2",
              "D": "6",
              "answer": "B"
            }
          },
          {
            "item": {
              "question": "Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5.",
              "A": "8",
              "B": "2",
              "C": "24",
              "D": "120",
              "answer": "C"
            }
          }
        ]
      },
      "input_messages": {
        "type": "template",
        "template": [
          {
            "type": "message",
            "role": "system",
            "content": {
              "text": "Answer the question's with A, B, C, or D."
            }
          },
          {
            "type": "message",
            "role": "user",
            "content": {
              "text": "Question: {{item.question}} A: {{item.A}} B: {{item.B}} C: {{item.C}} D: {{item.D}}."
            }
          }
        ]
      },
      "model": "gpt-4o-mini",
      "sampling_params": {
        "seed": 42,
        "temperature": 1,
        "top_p": 1
      }
    },
    "error": null,
    "metadata": {}
  }
}

評価 - 実行の削除

DELETE {endpoint}/openai/evals/{eval-id}/runs/{run-id}?api-version=2025-04-01-preview

評価実行を削除します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 実行を削除する評価の ID。
run-id パス Yes 文字列 削除する実行の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価実行が正常に削除されました

Content-Type Type Description
application/json オブジェクト

状態コード: 404

説明: 実行が見つかりません

Content-Type Type Description
application/json Error

Examples

DELETE {endpoint}/openai/evals/{eval-id}/runs/{run-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "eval.deleted",
    "deleted": true,
    "run_id": "evalrun_68016a056f0481909b9774447bdd1aa3"
  }
}

状態コード: 404

{
  "headers": {},
  "body": {
    "code": "notFound",
    "message": "Evaluation Run with ID evalrun_68016a056f0481909b9774447bdd1aa3 not found."
  }
}

評価 - 実行の取得

GET {endpoint}/openai/evals/{eval-id}/runs/{run-id}?api-version=2025-04-01-preview

ID で評価実行を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 実行を取得する評価の ID。
run-id パス Yes 文字列 取得する実行の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価の実行

Content-Type Type Description
application/json EvalRun

Examples

GET {endpoint}/openai/evals/{eval-id}/runs/{run-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "eval.run",
    "id": "evalrun_68016a056f0481909b9774447bdd1aa3",
    "eval_id": "eval_6801694950848190b10968bb628b651d",
    "report_url": "https://ai.azure.com/resource/evaluation",
    "status": "queued",
    "model": "gpt-4o-mini",
    "name": "Math quiz",
    "created_at": 1744923141,
    "result_counts": {
      "total": 0,
      "errored": 0,
      "failed": 0,
      "passed": 0
    },
    "per_model_usage": null,
    "per_testing_criteria_results": null,
    "data_source": {
      "type": "completions",
      "source": {
        "type": "file_content",
        "content": [
          {
            "item": {
              "question": "Find the degree for the given field extension Q(sqrt(2), sqrt(3), sqrt(18)) over Q.",
              "A": "0",
              "B": "4",
              "C": "2",
              "D": "6",
              "answer": "B"
            }
          },
          {
            "item": {
              "question": "Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5.",
              "A": "8",
              "B": "2",
              "C": "24",
              "D": "120",
              "answer": "C"
            }
          }
        ]
      },
      "input_messages": {
        "type": "template",
        "template": [
          {
            "type": "message",
            "role": "system",
            "content": {
              "text": "Answer the question's with A, B, C, or D."
            }
          },
          {
            "type": "message",
            "role": "user",
            "content": {
              "text": "Question: {{item.question}} A: {{item.A}} B: {{item.B}} C: {{item.C}} D: {{item.D}}."
            }
          }
        ]
      },
      "model": "gpt-4o-mini",
      "sampling_params": {
        "seed": 42,
        "temperature": 1,
        "top_p": 1
      }
    },
    "error": null,
    "metadata": {}
  }
}

評価 - 実行の取り消し

POST {endpoint}/openai/evals/{eval-id}/runs/{run-id}?api-version=2025-04-01-preview

継続的な評価実行を取り消します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 実行を取り消したい評価の ID。
run-id パス Yes 文字列 取り消す実行の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 取り消された eval run オブジェクト

Content-Type Type Description
application/json EvalRun

Examples

POST {endpoint}/openai/evals/{eval-id}/runs/{run-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "eval.run",
    "id": "evalrun_68016a056f0481909b9774447bdd1aa3",
    "eval_id": "eval_6801694950848190b10968bb628b651d",
    "report_url": "https://ai.azure.com/resource/evaluation",
    "status": "canceled",
    "model": "gpt-4o-mini",
    "name": "Math quiz",
    "created_at": 1744923141,
    "result_counts": {
      "total": 0,
      "errored": 0,
      "failed": 0,
      "passed": 0
    },
    "per_model_usage": null,
    "per_testing_criteria_results": null,
    "data_source": {
      "type": "completions",
      "source": {
        "type": "file_content",
        "content": [
          {
            "item": {
              "question": "Find the degree for the given field extension Q(sqrt(2), sqrt(3), sqrt(18)) over Q.",
              "A": "0",
              "B": "4",
              "C": "2",
              "D": "6",
              "answer": "B"
            }
          },
          {
            "item": {
              "question": "Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5.",
              "A": "8",
              "B": "2",
              "C": "24",
              "D": "120",
              "answer": "C"
            }
          }
        ]
      },
      "input_messages": {
        "type": "template",
        "template": [
          {
            "type": "message",
            "role": "system",
            "content": {
              "text": "Answer the question's with A, B, C, or D."
            }
          },
          {
            "type": "message",
            "role": "user",
            "content": {
              "text": "Question: {{item.question}} A: {{item.A}} B: {{item.B}} C: {{item.C}} D: {{item.D}}."
            }
          }
        ]
      },
      "model": "gpt-4o-mini",
      "sampling_params": {
        "seed": 42,
        "temperature": 1,
        "top_p": 1
      }
    },
    "error": null,
    "metadata": {}
  }
}

評価 - Getrunoutputitems

GET {endpoint}/openai/evals/{eval-id}/runs/{run-id}/output_items?api-version=2025-04-01-preview

評価実行の出力項目の一覧を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 実行を取得する評価の ID。
run-id パス Yes 文字列 出力項目を取得する実行の ID。
after クエリ No 文字列 前の改ページ調整要求からの最後の出力項目の識別子。
limit クエリ No 整数 取得する出力項目の数。
状態 クエリ No 文字列
使用可能な値: failpass
状態で出力項目をフィルタリングする。 failedを使用して失敗した出力項目でフィルター処理するか、passを使用して渡された出力項目でフィルター処理します。
順序 クエリ No 文字列
使用可能な値: ascdesc
タイムスタンプによる出力項目の並べ替え順序。 昇順には asc を使用し、降順には desc を使用します。 既定値は asc です。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価実行の出力項目の一覧

Content-Type Type Description
application/json EvalRunOutputItemList

Examples

GET {endpoint}/openai/evals/{eval-id}/runs/{run-id}/output_items?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "list",
    "data": [
      {
        "object": "eval.run.output_item",
        "id": "outputitem_68017251ff3881908bf5096bf4cd91c6",
        "created_at": 1744925265,
        "run_id": "evalrun_68016a056f0481909b9774447bdd1aa3",
        "eval_id": "eval_6801694950848190b10968bb628b651d",
        "status": "fail",
        "datasource_item_id": 1,
        "datasource_item": {
          "question": "Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5.",
          "A": "8",
          "B": "2",
          "C": "24",
          "D": "120",
          "answer": "C"
        },
        "results": [
          {
            "name": "string check-63b1fffa-bee6-4c37-ae77-ed46e3dce2b7",
            "sample": null,
            "passed": false,
            "score": 0
          }
        ],
        "sample": {
          "input": [
            {
              "role": "system",
              "content": "Answer the question's with A, B, C, or D.",
              "function_call": null
            },
            {
              "role": "user",
              "content": "Question: Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5. A: 8 B: 2 C: 24 D: 120.",
              "function_call": null
            }
          ],
          "output": [
            {
              "role": "assistant",
              "content": "To find the index of the subgroup generated by the permutation \\( p = (1, 2, 5, 4)(2, 3) \\) in \\( S_5 \\), we first need to determine the order of \\( p \\).\n\n1. The cycle \\( (1, 2, 5, 4) \\) has length 4.\n2. The cycle \\( (2, 3) \\) has length 2.\n3. The least common multiple (LCM) of these lengths gives the order of the permutation \\( p \\).\n\nCalculating the LCM:\n\\[\n\\text{lcm}(4, 2) = 4\n\\]\nThus, the order of \\( p \\) is 4.\n\nNext, we find the size of the subgroup \\( \\langle p \\rangle \\):\n- The size of \\( \\langle p \\rangle \\) is equal to the order of \\( p \\), which is 4.\n\nNow, we know the size of \\( S_5 \\):\n\\[\n|S_5| = 5! = 120\n\\]\n\nFinally, to find the index of \\( \\langle p \\rangle \\) in \\( S_5 \\), we use the formula:\n\\[\n\\text{index} = \\frac{|S_5|}{|\\langle p \\rangle|} = \\frac{120}{4} = 30\n\\]\n\nHowever, the available answer choices do not include 30. There may be a mistake because the question expects an answer among A, B, C, and D.\n\nLet\u00e2\u20ac\u2122s assume we misunderstood \\( \\langle p \\rangle \\), and instead, we can deduce based solely on given answers looking for a relation to \\( S_5 \\) without delving into detailed subgroup tracks. \n\nSince \\( S_5 \\) has 30 different elements in a subgroup configuration, the closest answer physically relating as long as \\( p \\) covers two elements effectively would logically fit an answer of 120 / 60 which has no direct relationship.\n\nGiven the option choices and specific rank formulations as often made regarding elements in specific construct the subgroup at best reflects around a viable ratio of parts allowed through available indices. \n\nThus, after reasoning through which aligns most structurally geometrically yielding across all configurations possible integrated yet arrives leading \\( p \\) through neighborhood distributions leaving reflections outstanding:\n\n\n\nThe correct answer is:\n**C: 24**\n\nHowever per the discussion migrating \\( p \\) may leave various pathways leading ultimately toward that framing in modeling. Always a good suggestion confirming the elements group generating any possible in outputs rationales. ",
              "function_call": null
            }
          ],
          "finish_reason": "stop",
          "model": "gpt-4o-mini-2024-07-18",
          "usage": {
            "total_tokens": 627,
            "completion_tokens": 546,
            "prompt_tokens": 81,
            "cached_tokens": 0
          },
          "error": null,
          "temperature": 1,
          "top_p": 1,
          "seed": 42
        }
      },
      {
        "object": "eval.run.output_item",
        "id": "outputitem_6801724f54888190942ec66a197309dd",
        "created_at": 1744925263,
        "run_id": "evalrun_68016a056f0481909b9774447bdd1aa3",
        "eval_id": "eval_6801694950848190b10968bb628b651d",
        "status": "fail",
        "datasource_item_id": 0,
        "datasource_item": {
          "question": "Find the degree for the given field extension Q(sqrt(2), sqrt(3), sqrt(18)) over Q.",
          "A": "0",
          "B": "4",
          "C": "2",
          "D": "6",
          "answer": "B"
        },
        "results": [
          {
            "name": "string check-63b1fffa-bee6-4c37-ae77-ed46e3dce2b7",
            "sample": null,
            "passed": false,
            "score": 0
          }
        ],
        "sample": {
          "input": [
            {
              "role": "system",
              "content": "Answer the question's with A, B, C, or D.",
              "function_call": null
            },
            {
              "role": "user",
              "content": "Question: Find the degree for the given field extension Q(sqrt(2), sqrt(3), sqrt(18)) over Q. A: 0 B: 4 C: 2 D: 6.",
              "function_call": null
            }
          ],
          "output": [
            {
              "role": "assistant",
              "content": "To find the degree of the field extension \\(\\mathbb{Q}(\\sqrt{2}, \\sqrt{3}, \\sqrt{18})\\) over \\(\\mathbb{Q}\\), we can simplify \\(\\sqrt{18}\\) since \\(\\sqrt{18} = \\sqrt{9 \\cdot 2} = 3\\sqrt{2}\\). \n\nThus, the field can be rewritten as \\(\\mathbb{Q}(\\sqrt{2}, \\sqrt{3})\\).\n\nNext, we calculate the degree of the extensions step by step:\n\n1. The extension \\(\\mathbb{Q}(\\sqrt{2})\\) over \\(\\mathbb{Q}\\) has degree 2, since \\(\\sqrt{2}\\) is not in \\(\\mathbb{Q}\\) and satisfies the polynomial \\(x^2 - 2\\).\n\n2. Next, we consider the extension \\(\\mathbb{Q}(\\sqrt{2}, \\sqrt{3})\\) over \\(\\mathbb{Q}(\\sqrt{2})\\). The element \\(\\sqrt{3}\\) is also not in \\(\\mathbb{Q}(\\sqrt{2})\\) and satisfies the polynomial \\(x^2 - 3\\), which is irreducible over \\(\\mathbb{Q}(\\sqrt{2})\\). Hence, the degree of the extension \\(\\mathbb{Q}(\\sqrt{2}, \\sqrt{3})\\) over \\(\\mathbb{Q}(\\sqrt{2})\\) is also 2.\n\nNow we can combine these degrees:\n\n\\[\n[\\mathbb{Q}(\\sqrt{2}, \\sqrt{3}) : \\mathbb{Q}] = [\\mathbb{Q}(\\sqrt{2}, \\sqrt{3}) : \\mathbb{Q}(\\sqrt{2})] \\times [\\mathbb{Q}(\\sqrt{2}) : \\mathbb{Q}] = 2 \\times 2 = 4.\n\\]\n\nThus, the degree of the field extension \\(\\mathbb{Q}(\\sqrt{2}, \\sqrt{3}, \\sqrt{18})\\) over \\(\\mathbb{Q}\\) is 4.\n\nTherefore, the answer is:\n\n**B: 4**.",
              "function_call": null
            }
          ],
          "finish_reason": "stop",
          "model": "gpt-4o-mini-2024-07-18",
          "usage": {
            "total_tokens": 556,
            "completion_tokens": 487,
            "prompt_tokens": 69,
            "cached_tokens": 0
          },
          "error": null,
          "temperature": 1,
          "top_p": 1,
          "seed": 42
        }
      }
    ],
    "first_id": "outputitem_68017251ff3881908bf5096bf4cd91c6",
    "last_id": "outputitem_6801724f54888190942ec66a197309dd",
    "has_more": false
  }
}

評価 - Getrunoutputitem

GET {endpoint}/openai/evals/{eval-id}/runs/{run-id}/output_items/{output-item-id}?api-version=2025-04-01-preview

評価実行出力項目を ID で取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
eval-id パス Yes 文字列 実行を取得する評価の ID。
run-id パス Yes 文字列 取得する実行の ID。
output-item-id パス Yes 文字列 取得する出力項目の ID。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 評価実行の出力項目

Content-Type Type Description
application/json EvalRunOutputItem

Examples

GET {endpoint}/openai/evals/{eval-id}/runs/{run-id}/output_items/{output-item-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "eval.run.output_item",
    "id": "outputitem_68017251ff3881908bf5096bf4cd91c6",
    "created_at": 1744925265,
    "run_id": "evalrun_68016a056f0481909b9774447bdd1aa3",
    "eval_id": "eval_6801694950848190b10968bb628b651d",
    "status": "fail",
    "datasource_item_id": 1,
    "datasource_item": {
      "question": "Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5.",
      "A": "8",
      "B": "2",
      "C": "24",
      "D": "120",
      "answer": "C"
    },
    "results": [
      {
        "name": "string check-63b1fffa-bee6-4c37-ae77-ed46e3dce2b7",
        "sample": null,
        "passed": false,
        "score": 0
      }
    ],
    "sample": {
      "input": [
        {
          "role": "system",
          "content": "Answer the question's with A, B, C, or D."
        },
        {
          "role": "user",
          "content": "Question: Let p = (1, 2, 5, 4)(2, 3) in S_5 . Find the index of <p> in S_5. A: 8 B: 2 C: 24 D: 120."
        }
      ],
      "output": [
        {
          "role": "assistant",
          "content": "To find the index of the subgroup generated by the permutation \\( p = (1, 2, 5, 4)(2, 3) \\) in \\( S_5 \\), we first need to determine the order of \\( p \\).\n\n1. The cycle \\( (1, 2, 5, 4) \\) has length 4.\n2. The cycle \\( (2, 3) \\) has length 2.\n3. The least common multiple (LCM) of these lengths gives the order of the permutation \\( p \\).\n\nCalculating the LCM:\n\\[\n\\text{lcm}(4, 2) = 4\n\\]\nThus, the order of \\( p \\) is 4.\n\nNext, we find the size of the subgroup \\( \\langle p \\rangle \\):\n- The size of \\( \\langle p \\rangle \\) is equal to the order of \\( p \\), which is 4.\n\nNow, we know the size of \\( S_5 \\):\n\\[\n|S_5| = 5! = 120\n\\]\n\nFinally, to find the index of \\( \\langle p \\rangle \\) in \\( S_5 \\), we use the formula:\n\\[\n\\text{index} = \\frac{|S_5|}{|\\langle p \\rangle|} = \\frac{120}{4} = 30\n\\]\n\nHowever, the available answer choices do not include 30. There may be a mistake because the question expects an answer among A, B, C, and D.\n\nLet\u00e2\u20ac\u2122s assume we misunderstood \\( \\langle p \\rangle \\), and instead, we can deduce based solely on given answers looking for a relation to \\( S_5 \\) without delving into detailed subgroup tracks. \n\nSince \\( S_5 \\) has 30 different elements in a subgroup configuration, the closest answer physically relating as long as \\( p \\) covers two elements effectively would logically fit an answer of 120 / 60 which has no direct relationship.\n\nGiven the option choices and specific rank formulations as often made regarding elements in specific construct the subgroup at best reflects around a viable ratio of parts allowed through available indices. \n\nThus, after reasoning through which aligns most structurally geometrically yielding across all configurations possible integrated yet arrives leading \\( p \\) through neighborhood distributions leaving reflections outstanding:\n\n\n\nThe correct answer is:\n**C: 24**\n\nHowever per the discussion migrating \\( p \\) may leave various pathways leading ultimately toward that framing in modeling. Always a good suggestion confirming the elements group generating any possible in outputs rationales. "
        }
      ],
      "finish_reason": "stop",
      "model": "gpt-4o-mini-2024-07-18",
      "usage": {
        "total_tokens": 627,
        "completion_tokens": 546,
        "prompt_tokens": 81,
        "cached_tokens": 0
      },
      "error": null,
      "temperature": 1,
      "top_p": 1,
      "seed": 42
    }
  }
}

ファイル - リスト

GET {endpoint}/openai/files?api-version=2025-04-01-preview

Azure OpenAI リソースによって所有されているすべてのファイルの一覧を取得します。

これには、微調整モデルのトレーニングまたは検証のための "微調整" 目的のファイルや、サービスによって生成されたファイル (対応する微調整ジョブのさまざまなメトリックを含む "微調整結果" など) を含むファイルなど、ユーザーがアップロードしたコンテンツが含まれます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
purpose クエリ No 文字列
使用可能な値: fine-tunefine-tune-resultsassistantsassistants_outputbatchbatch_outputevalsevals_output
指定された目的のファイルのみを返します。
limit クエリ No 整数 返されるオブジェクトの数の制限。 制限の範囲は 1 から 10,000 で、既定値は 10,000 です。
after クエリ No 文字列 改ページで使うカーソル。 after は、リスト内での場所を定義するオブジェクト ID です。 たとえば、リスト取得要求を行い、obj_foo で終わる 100 個のオブジェクトを受け取った場合、後続の呼び出しに after=obj_foo を含めて、リストの次のページをフェッチできます。
順序 クエリ No 文字列 オブジェクトの created_at タイムスタンプによる並べ替え順序。 asc は昇順に使用し、desc は降順に使用します。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json FileList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/files?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "data": [
      {
        "status": "processed",
        "bytes": 140,
        "purpose": "fine-tune",
        "filename": "puppy.jsonl",
        "id": "file-181a1cbdcdcf4677ada87f63a0928099",
        "created_at": 1646126127,
        "object": "file"
      },
      {
        "status": "processed",
        "bytes": 32423,
        "purpose": "fine-tune-results",
        "filename": "results.csv",
        "id": "file-181a1cbdcdcf4677ada87f63a0928099",
        "created_at": 1646126127,
        "object": "file"
      }
    ],
    "object": "list"
  }
}

ファイル - アップロード

POST {endpoint}/openai/files?api-version=2025-04-01-preview

ローカル コンピューターからデータをアップロードして、新しいファイル エンティティを作成します。 アップロードされたファイルは、たとえば、微調整されたモデルのトレーニングや評価に使用できます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: multipart/form-data

Name タイプ Description Required Default
アンカー 文字列 絶対有効期限を生成する基準となる時刻に対するアンカーを定義します。 No
seconds 整数 相対有効期限を秒単位で取得します。
範囲: [1209600 - 2592000]. No
ファイル 文字列 OpenAI にアップロードするファイルAzure取得または設定します。 Yes
purpose 文字列 アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

Responses

状態コード: 201

説明: ファイルが正常に作成されました。

Content-Type Type Description
application/json File

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/files?api-version=2025-04-01-preview

応答: 状態コード: 201

{
  "headers": {
    "location": "https://aoairesource.openai.azure.com/openai/files/file-181a1cbdcdcf4677ada87f63a0928099"
  },
  "body": {
    "status": "pending",
    "purpose": "fine-tune",
    "filename": "puppy.jsonl",
    "id": "file-181a1cbdcdcf4677ada87f63a0928099",
    "created_at": 1646126127,
    "object": "file"
  }
}

ファイル - インポート

POST {endpoint}/openai/files/import?api-version=2025-04-01-preview

指定された URL からデータをインポートして、新しいファイル エンティティを作成します。 アップロードされたファイルは、たとえば、微調整されたモデルのトレーニングや評価に使用できます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: application/json

Name タイプ Description Required Default
content_url 文字列 ドキュメントをダウンロードする URL (BLOB の SAS URL、または GET 要求でアクセスできるその他の外部 URL を指定できます)。 Yes
expires_after FileExpiresAfter ファイルの有効期限を定義します。 No
filename 文字列 アップロードする JSON 行 ファイルの名前
purpose が "微調整" に設定されている場合、各行は、トレーニング例を表す "prompt" フィールドと "completion" フィールドを含む JSON レコードです。 Yes
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

Responses

状態コード: 201

説明: ファイルが正常に作成されました。

Content-Type Type Description
application/json File

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/files/import?api-version=2025-04-01-preview

応答: 状態コード: 201

{
  "headers": {
    "location": "https://aoairesource.openai.azure.com/openai/files/file-181a1cbdcdcf4677ada87f63a0928099"
  },
  "body": {
    "status": "pending",
    "purpose": "fine-tune",
    "filename": "puppy.jsonl",
    "id": "file-181a1cbdcdcf4677ada87f63a0928099",
    "created_at": 1646126127,
    "object": "file"
  }
}

ファイル - 取得

GET {endpoint}/openai/files/{file-id}?api-version=2025-04-01-preview

状態、サイズ、目的など、指定されたファイル ID で指定された 1 つのファイルの詳細を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
file-id パス Yes 文字列 ファイルの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json File

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/files/{file-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "status": "processed",
    "bytes": 140,
    "purpose": "fine-tune",
    "filename": "puppy.jsonl",
    "id": "file-181a1cbdcdcf4677ada87f63a0928099",
    "created_at": 1646126127,
    "object": "file"
  }
}

ファイル - 削除

DELETE {endpoint}/openai/files/{file-id}?api-version=2025-04-01-preview

指定されたファイル ID を持つファイルを削除します。

また、ファイルが使用された場合 (微調整ジョブのトレーニング ファイルなど) も削除できます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
file-id パス Yes 文字列 ファイルの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: ファイルが正常に削除されました。

Content-Type Type Description
application/json FileDelete

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

DELETE {endpoint}/openai/files/{file-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "object": "file",
    "deleted": true,
    "id": "file-181a1cbdcdcf4677ada87f63a0928099"
  }
}

ファイル - コンテンツを取得する

GET {endpoint}/openai/files/{file-id}/content?api-version=2025-04-01-preview

指定されたファイル ID で指定されたファイルの内容を取得します。

ファイルは、ユーザーがコンテンツをアップロードしたり、微調整ジョブの結果メトリックのようにサービスによって生成されたりすることができます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
file-id パス Yes 文字列 ファイルの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/octet-stream 文字列
application/json 文字列

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/octet-stream ErrorResponse
application/json ErrorResponse

Examples

GET {endpoint}/openai/files/{file-id}/content?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": "raw file content"
}

微調整 - 一覧表示

GET {endpoint}/openai/fine_tuning/jobs?api-version=2025-04-01-preview

Azure OpenAI リソースが所有するすべての微調整ジョブの一覧を取得します。各微調整ジョブに対して返される詳細には、基本モデル、トレーニングおよび検証ファイル、ハイパー パラメーター、タイム スタンプ、状態、イベントの識別子以外にも含まれます。

イベントは、ジョブの状態が変わると作成されます。 たとえば、実行中または完了時、および結果がアップロードされるタイミングなどです。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
after クエリ No 文字列 前の改ページ位置の自動修正の要求からの最後のイベントの識別子。
limit クエリ No 整数 取得するイベントの数。 既定値は 20 です。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json FineTuningJobList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/fine_tuning/jobs?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "has_more": false,
    "data": [
      {
        "hyperparameters": {
          "n_epochs": -1
        },
        "integrations": [
          {
            "type": "wandb",
            "wandb": {
              "project": "custom-wandb-project",
              "tags": [
                "project:tag",
                "lineage"
              ]
            }
          }
        ],
        "method": {
          "type": "supervised"
        },
        "status": "succeeded",
        "model": "curie",
        "fine_tuned_model": "curie.ft-72a2792ef7d24ba7b82c7fe4a37e379f",
        "training_file": "file-181a1cbdcdcf4677ada87f63a0928099",
        "result_files": [
          "file-181a1cbdcdcf4677ada87f63a0928099"
        ],
        "finished_at": 1646127311,
        "trained_tokens": 2342,
        "id": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
        "created_at": 1646126127,
        "object": "fine_tuning.job"
      }
    ],
    "object": "list"
  }
}

微調整 - 作成

POST {endpoint}/openai/fine_tuning/jobs?api-version=2025-04-01-preview

特定のトレーニング ファイルから指定したモデルを微調整するジョブを作成します。

応答には、ジョブの状態やハイパー パラメーターなど、エンキューされたジョブの詳細が含まれます。

微調整されたモデルの名前は、完了すると応答に追加されます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: application/json

Name タイプ Description Required Default
hyperparameters FineTuningHyperParameters 微調整ジョブで使用されるハイパー パラメーター設定。 No
integrations アレイ 微調整ジョブをサポートする統合の構成の一覧。 多くの統合が計画されているため、必ず統合の種類を確認して操作してください。 No
メソッド オブジェクト 監視対象の微調整に使用されるメソッド No
└─ type FineTuningMethodType No
モデル 文字列 この微調整に使用される基本モデルの識別子 (model-id)。 Yes
seed 整数 微調整ジョブに使用されるシード。 同じシードとジョブ パラメーターを渡すと同じ結果が得られますが、まれに異なる場合があります。 シードが指定されていない場合は生成されます。 No
suffix 文字列 微調整されたモデルを識別するために使用されるサフィックス。 サフィックスには最大 40 文字 (a-z、A-Z、0-9、-、_) を指定することができ、微調整したモデル名に追加されます。 No
training_file 文字列 この微調整されたモデルのトレーニングに使用されるファイル ID (ファイル ID)。 Yes
validation_file 文字列 トレーニング中に微調整されたモデルを評価するために使用されるファイル ID (ファイル ID)。 No

Responses

状態コード: 201

説明: ファインチューニングが正常に作成されました。

Content-Type Type Description
application/json FineTuningJob

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/fine_tuning/jobs?api-version=2025-04-01-preview

応答: 状態コード: 201

{
  "headers": {
    "location": "https://aoairesource.openai.azure.com/openai/fine_tuning/jobs/ft-72a2792ef7d24ba7b82c7fe4a37e379f"
  },
  "body": {
    "hyperparameters": {
      "n_epochs": -1
    },
    "integrations": [
      {
        "type": "wandb",
        "wandb": {
          "project": "custom-wandb-project",
          "tags": [
            "project:tag",
            "lineage"
          ]
        }
      }
    ],
    "method": {
      "type": "supervised"
    },
    "status": "pending",
    "model": "curie",
    "training_file": "file-181a1cbdcdcf4677ada87f63a0928099",
    "id": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "fine_tuning.job"
  }
}

微調整 - 取得

GET {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}?api-version=2025-04-01-preview

指定された微調整 ID で指定された 1 つの微調整ジョブの詳細を取得します。

詳細には、基本モデル、トレーニングおよび検証ファイル、ハイパー パラメーター、タイムスタンプ、状態、イベントが含まれます。

イベントは、ジョブの状態 (実行中や完了など) が変更されたとき、および結果がアップロードされたときに作成されます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
fine-tuning-job-id パス Yes 文字列 微調整ジョブの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json FineTuningJob

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "hyperparameters": {
      "n_epochs": -1
    },
    "integrations": [
      {
        "type": "wandb",
        "wandb": {
          "project": "custom-wandb-project",
          "tags": [
            "project:tag",
            "lineage"
          ]
        }
      }
    ],
    "method": {
      "type": "supervised"
    },
    "status": "succeeded",
    "model": "curie",
    "fine_tuned_model": "curie.ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "training_file": "file-181a1cbdcdcf4677ada87f63a0928099",
    "result_files": [
      "file-181a1cbdcdcf4677ada87f63a0928099"
    ],
    "finished_at": 1646127311,
    "trained_tokens": 2342,
    "id": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "fine_tuning.job"
  }
}

微調整 - 削除

DELETE {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}?api-version=2025-04-01-preview

所定の微調整 ID で指定された微調整ジョブを削除します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
fine-tuning-job-id パス Yes 文字列 微調整ジョブの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 204

説明: 微調整が正常に削除されました。

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

DELETE {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}?api-version=2025-04-01-preview

応答: 状態コード: 204

{
  "headers": {}
}

微調整 - キャンセル

POST {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/cancel?api-version=2025-04-01-preview

指定された微調整 ID で指定された微調整ジョブの処理を取り消します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
fine-tuning-job-id パス Yes 文字列 微調整ジョブの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 微調整が正常に取り消されました

Content-Type Type Description
application/json FineTuningJob

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/cancel?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "hyperparameters": {
      "n_epochs": -1
    },
    "integrations": [
      {
        "type": "wandb",
        "wandb": {
          "project": "custom-wandb-project",
          "tags": [
            "project:tag",
            "lineage"
          ]
        }
      }
    ],
    "method": {
      "type": "supervised"
    },
    "status": "cancelled",
    "model": "curie",
    "training_file": "file-181a1cbdcdcf4677ada87f63a0928099",
    "finished_at": 1646127311,
    "id": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "fine_tuning.job"
  }
}

微調整 - チェックポイントを取得する

GET {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/checkpoints?api-version=2025-04-01-preview

所定の微調整 ID で指定された微調整ジョブのチェックポイントを取得します。

チェックポイントは、トレーニング中に成功したエポックの最後に作成されます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
fine-tuning-job-id パス Yes 文字列 微調整ジョブの識別子。
after クエリ No 文字列 前の改ページ調整要求の最後のチェックポイント ID の識別子。
limit クエリ No 整数 取得するチェックポイントの数。 既定値は 10 です。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json FineTuningJobCheckpointList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/checkpoints?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "has_more": false,
    "data": [
      {
        "id": "ftckpt_qtZ5Gyk4BLq1SfLFWp3RtO3P",
        "created_at": 1646126127,
        "fine_tuned_model_checkpoint": "gpt-35-turbo-0613.ft-72a2792ef7d24ba7b82c7fe4a37e379f:ckpt-step-10",
        "step_number": 10,
        "metrics": {
          "step": 10,
          "train_loss": 0.478,
          "train_mean_token_accuracy": 0.924,
          "valid_loss": 10.112,
          "valid_mean_token_accuracy": 0.145,
          "full_valid_loss": 0.567,
          "full_valid_mean_token_accuracy": 0.944
        },
        "fine_tuning_job_id": "ftjob-72a2792ef7d24ba7b82c7fe4a37e379f",
        "object": "fine_tuning.job.checkpoint"
      },
      {
        "id": "ftckpt_frtXDR2453C4FG67t03MLPs5",
        "created_at": 1646126137,
        "fine_tuned_model_checkpoint": "gpt-35-turbo-0613.ft-72a2792ef7d24ba7b82c7fe4a37e379f:ckpt-step-20",
        "step_number": 20,
        "metrics": {
          "step": 20,
          "train_loss": 0.355,
          "train_mean_token_accuracy": 0.947,
          "valid_loss": 11.32,
          "valid_mean_token_accuracy": 0.122,
          "full_valid_loss": 0.317,
          "full_valid_mean_token_accuracy": 0.955
        },
        "fine_tuning_job_id": "ftjob-72a2792ef7d24ba7b82c7fe4a37e379f",
        "object": "fine_tuning.job.checkpoint"
      },
      {
        "id": "ftckpt_agLk6Gio560ORp14gl123fgd",
        "created_at": 1646126147,
        "fine_tuned_model_checkpoint": "gpt-35-turbo-0613.ft-72a2792ef7d24ba7b82c7fe4a37e379f:ckpt-step-30",
        "step_number": 30,
        "metrics": {
          "step": 30,
          "train_loss": 0.155,
          "train_mean_token_accuracy": 0.975,
          "valid_loss": 9.31,
          "valid_mean_token_accuracy": 0.092,
          "full_valid_loss": 0.114,
          "full_valid_mean_token_accuracy": 0.963
        },
        "fine_tuning_job_id": "ftjob-72a2792ef7d24ba7b82c7fe4a37e379f",
        "object": "fine_tuning.job.checkpoint"
      }
    ],
    "object": "list"
  }
}

微調整 - イベントを取得する

GET {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/events?api-version=2025-04-01-preview

所定の微調整 ID で指定された微調整ジョブのイベントを取得します。

イベントは、ジョブの状態 (実行中や完了など) が変更されたとき、および結果がアップロードされたときに作成されます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
fine-tuning-job-id パス Yes 文字列 微調整ジョブの識別子。
after クエリ No 文字列 前の改ページ位置の自動修正の要求からの最後のイベントの識別子。
limit クエリ No 整数 取得するイベントの数。 既定値は 20 です。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json FineTuningJobEventList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/events?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "has_more": false,
    "data": [
      {
        "id": "ftevent-363dcd7cb4c74539bc53293c1dceef05",
        "created_at": 1646126127,
        "level": "info",
        "message": "Job enqueued. Waiting for jobs ahead to complete.",
        "type": "message",
        "object": "fine_tuning.job.event"
      },
      {
        "id": "ftevent-8c2a44999790437cb3230e543fa2cf0f",
        "created_at": 1646126169,
        "level": "info",
        "message": "Job started.",
        "type": "message",
        "object": "fine_tuning.job.event"
      },
      {
        "id": "ftevent-2d47d651d2f3484c8187c88c00078147",
        "created_at": 1646126192,
        "level": "info",
        "message": "Job succeeded.",
        "type": "message",
        "object": "fine_tuning.job.event"
      }
    ],
    "object": "list"
  }
}

微調整 - 一時停止

POST {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/pause?api-version=2025-04-01-preview

指定された微調整 ID で指定された微調整ジョブの処理を一時停止します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
fine-tuning-job-id パス Yes 文字列 微調整ジョブの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 微調整が正常に一時停止されました。

Content-Type Type Description
application/json FineTuningJob

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/pause?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "hyperparameters": {
      "n_epochs": -1
    },
    "integrations": [
      {
        "type": "wandb",
        "wandb": {
          "project": "custom-wandb-project",
          "tags": [
            "project:tag",
            "lineage"
          ]
        }
      }
    ],
    "method": {
      "type": "supervised"
    },
    "status": "paused",
    "model": "curie",
    "training_file": "file-181a1cbdcdcf4677ada87f63a0928099",
    "finished_at": 1646127311,
    "id": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "fine_tuning.job"
  }
}

微調整 - 再開

POST {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/resume?api-version=2025-04-01-preview

指定された微調整 ID で指定された微調整ジョブの処理を再開します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
fine-tuning-job-id パス Yes 文字列 微調整ジョブの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

説明: 微調整が正常に再開されました。

Content-Type Type Description
application/json FineTuningJob

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/fine_tuning/jobs/{fine-tuning-job-id}/resume?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "hyperparameters": {
      "n_epochs": -1
    },
    "integrations": [
      {
        "type": "wandb",
        "wandb": {
          "project": "custom-wandb-project",
          "tags": [
            "project:tag",
            "lineage"
          ]
        }
      }
    ],
    "method": {
      "type": "supervised"
    },
    "status": "resuming",
    "model": "curie",
    "training_file": "file-181a1cbdcdcf4677ada87f63a0928099",
    "finished_at": 1646127311,
    "id": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "fine_tuning.job"
  }
}

インジェスト ジョブ - 一覧表示

GET {endpoint}/openai/ingestion/jobs?api-version=2025-04-01-preview

インジェスト ジョブを一覧表示します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。
mgmt-user-token ヘッダー No 文字列 ワークスペースをaccessするために使用されるトークン (ユーザー コンピューティング ジョブにのみ必要)。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json IngestionJobList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/ingestion/jobs?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "body": {
    "value": [
      {
        "jobId": "usercompute-ingestion-job",
        "kind": "UserCompute",
        "workspaceId": "/subscriptions/f375b912-331c-4fc5-8e9f-2d7205e3e036/resourceGroups/adrama-copilot-demo/providers/Microsoft.MachineLearningServices/workspaces/adrama-rag-dev"
      },
      {
        "jobId": "syscompute-ingestion-job",
        "kind": "SystemCompute"
      }
    ]
  }
}

インジェスト ジョブ - 作成

PUT {endpoint}/openai/ingestion/jobs/{job-id}?api-version=2025-04-01-preview

指定したジョブ ID を使用してインジェスト ジョブを作成します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。
job-id パス Yes 文字列 作成されるジョブの ID。
mgmt-user-token ヘッダー No 文字列 ワークスペースをaccessするために使用されるトークン (ユーザー コンピューティング ジョブにのみ必要)。
aml-user-token ヘッダー No 文字列 ワークスペース内のジョブ内のリソースをaccessするために使用されるトークン (ユーザー コンピューティング ジョブにのみ必要)。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: application/json

Name タイプ Description Required Default
jobId 文字列 No
kind IngestionJobType ジョブの種類。 Yes

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json IngestionJob

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

PUT {endpoint}/openai/ingestion/jobs/{job-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {
    "operation-location": "https://aoairesource.openai.azure.com/openai/ingestion/jobs/ingestion-job/runs/72a2792ef7d24ba7b82c7fe4a37e379f?api-version=2025-04-01-preview"
  },
  "body": {
    "kind": "SystemCompute",
    "jobId": "ingestion-job",
    "searchServiceConnection": {
      "kind": "EndpointWithManagedIdentity",
      "endpoint": "https://aykame-dev-search.search.windows.net"
    },
    "datasource": {
      "kind": "Storage",
      "connection": {
        "kind": "EndpointWithManagedIdentity",
        "endpoint": "https://mystorage.blob.core.windows.net/",
        "resourceId": "/subscriptions/1234567-abcd-1234-5678-1234abcd/resourceGroups/my-resource/providers/Microsoft.Storage/storageAccounts/mystorage"
      },
      "containerName": "container",
      "chunking": {
        "maxChunkSizeInTokens": 2048
      },
      "embeddings": [
        {
          "connection": {
            "kind": "RelativeConnection"
          },
          "deploymentName": "Ada"
        }
      ]
    },
    "dataRefreshIntervalInHours": 24,
    "completionAction": "keepAllAssets"
  }
}
PUT {endpoint}/openai/ingestion/jobs/{job-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {
    "operation-location": "https://aoairesource.openai.azure.com/openai/ingestion/jobs/ingestion-job/runs/72a2792ef7d24ba7b82c7fe4a37e379f?api-version=2025-04-01-preview"
  },
  "body": {
    "kind": "UserCompute",
    "jobId": "ingestion-job",
    "workspaceId": "/subscriptions/f375b912-331c-4fc5-8e9f-2d7205e3e036/resourceGroups/adrama-copilot-demo/providers/Microsoft.MachineLearningServices/workspaces/adrama-rag-dev",
    "compute": {
      "kind": "ServerlessCompute"
    },
    "target": {
      "kind": "AzureAISearch",
      "connectionId": "/subscriptions/f375b912-331c-4fc5-8e9f-2d7205e3e036/resourceGroups/adrama-copilot-demo/providers/Microsoft.MachineLearningServices/workspaces/adrama-rag-dev/connections/search-connection"
    },
    "datasource": {
      "kind": "Dataset",
      "datasetId": "azureml://locations/centraluseuap/workspaces/83317fe6-efa6-4e4a-b020-d0edd11ec382/data/PlainText/versions/1",
      "datasetType": "uri_folder"
    }
  }
}

インジェスト ジョブ - 取得

GET {endpoint}/openai/ingestion/jobs/{job-id}?api-version=2025-04-01-preview

指定したジョブ ID の詳細を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。
job-id パス Yes 文字列 ジョブの ID。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json IngestionJob

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/ingestion/jobs/{job-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "body": {
    "kind": "UserCompute",
    "jobId": "ingestion-job",
    "workspaceId": "/subscriptions/f375b912-331c-4fc5-8e9f-2d7205e3e036/resourceGroups/adrama-copilot-demo/providers/Microsoft.MachineLearningServices/workspaces/adrama-rag-dev",
    "compute": {
      "kind": "ServerlessCompute"
    },
    "target": {
      "kind": "AzureAISearch",
      "connectionId": "/subscriptions/f375b912-331c-4fc5-8e9f-2d7205e3e036/resourceGroups/adrama-copilot-demo/providers/Microsoft.MachineLearningServices/workspaces/adrama-rag-dev/connections/search-connection"
    },
    "datasource": {
      "kind": "Dataset",
      "datasetId": "azureml://locations/centraluseuap/workspaces/83317fe6-efa6-4e4a-b020-d0edd11ec382/data/PlainText/versions/1",
      "datasetType": "uri_folder"
    }
  }
}

Ingestionjobruns - 一覧表示

GET {endpoint}/openai/ingestion/jobs/{job-id}/runs?api-version=2025-04-01-preview

指定したジョブ ID の実行を一覧表示します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。
job-id パス Yes 文字列 ジョブの ID。
mgmt-user-token ヘッダー No 文字列 ワークスペースをaccessするために使用されるトークン (ユーザー コンピューティング ジョブにのみ必要)。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json IngestionJobRunList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/ingestion/jobs/{job-id}/runs?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "body": {
    "value": [
      {
        "jobId": "ingestion-job",
        "runId": "f375b912-331c-4fc5-8e9f-2d7205e3e036",
        "status": "succeeded"
      },
      {
        "jobId": "ingestion-job",
        "runId": "07f2d192-add7-4202-a2e3-858c2577f4fd",
        "status": "failed"
      },
      {
        "jobId": "ingestion-job",
        "runId": "5ef7a436-1147-4cbb-82e0-3d502bcc6a7b",
        "status": "running"
      }
    ]
  }
}

Ingestionjobruns - 取得

GET {endpoint}/openai/ingestion/jobs/{job-id}/runs/{run-id}?api-version=2025-04-01-preview

指定したジョブ ID の一部として、指定した実行 ID の詳細を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。
job-id パス Yes 文字列 ジョブの ID。
run-id パス Yes 文字列 実行の ID。
mgmt-user-token ヘッダー No 文字列 ワークスペースをaccessするために使用されるトークン (ユーザー コンピューティング ジョブにのみ必要)。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json IngestionJobRun

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/ingestion/jobs/{job-id}/runs/{run-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "body": {
    "jobId": "ingestion-job",
    "runId": "5ef7a436-1147-4cbb-82e0-3d502bcc6a7b",
    "status": "running",
    "progress": {
      "stageProgress": [
        {
          "name": "Preprocessing",
          "totalItems": 14,
          "processedItems": 0,
          "state": "notRunning"
        },
        {
          "name": "Indexing",
          "state": "notRunning"
        }
      ]
    }
  }
}

モデル - リスト

GET {endpoint}/openai/models?api-version=2025-04-01-preview

Azure OpenAI リソースからアクセスできるすべてのモデルの一覧を取得します。

これには、基本モデルと、Azure OpenAI リソースが所有するすべての正常に完了した微調整されたモデルが含まれます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json ModelList

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/models?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "data": [
      {
        "status": "succeeded",
        "capabilities": {
          "fine_tune": true,
          "inference": true,
          "completion": true,
          "chat_completion": false,
          "embeddings": false
        },
        "lifecycle_status": `generally-available`,
        "deprecation": {
          "fine_tune": 1677662127,
          "inference": 1709284527
        },
        "id": "curie",
        "created_at": 1646126127,
        "object": "model"
      },
      {
        "status": "succeeded",
        "model": "curie",
        "fine_tune": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
        "capabilities": {
          "fine_tune": false,
          "inference": true,
          "completion": true,
          "chat_completion": false,
          "embeddings": false
        },
        "lifecycle_status": `generally-available`,
        "deprecation": {
          "inference": 1709284527
        },
        "id": "curie.ft-72a2792ef7d24ba7b82c7fe4a37e379f",
        "created_at": 1646126127,
        "object": "model"
      }
    ],
    "object": "list"
  }
}

モデル - 取得

GET {endpoint}/openai/models/{model-id}?api-version=2025-04-01-preview

指定された modelId で指定されたモデルの詳細を取得します。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
model-id パス Yes 文字列 モデルの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json Model

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

GET {endpoint}/openai/models/{model-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "status": "succeeded",
    "capabilities": {
      "fine_tune": true,
      "inference": true,
      "completion": true,
      "chat_completion": false,
      "embeddings": false
    },
    "lifecycle_status": `generally-available`,
    "deprecation": {
      "fine_tune": 1677662127,
      "inference": 1709284527
    },
    "id": "curie",
    "created_at": 1646126127,
    "object": "model"
  }
}
GET {endpoint}/openai/models/{model-id}?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "status": "succeeded",
    "model": "curie",
    "fine_tune": "ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "capabilities": {
      "fine_tune": false,
      "inference": true,
      "completion": true,
      "chat_completion": false,
      "embeddings": false
    },
    "lifecycle_status": `generally-available`,
    "deprecation": {
      "inference": 1709284527
    },
    "id": "curie.ft-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "model"
  }
}

ファイルのアップロード - 開始

POST {endpoint}/openai/uploads?api-version=2025-04-01-preview

中間 Upload オブジェクトが作成され、パーツを追加できます。 現在、Upload のサイズは合計で最大 9 GB にすることができ、作成されてから 2 時間後に有効期限が切れます。

Upload が完了すると、アップロードされたすべてのパーツを含む File オブジェクトが生成されます。 この File オブジェクトは、他のファイルと同様に、プラットフォーム全体で使用できます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: application/json

Name タイプ Description Required Default
bytes 整数 Yes
filename 文字列 Yes
mime_type 文字列 Yes
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

Request Body

Content-Type: text/json

Name タイプ Description Required Default
bytes 整数 Yes
filename 文字列 Yes
mime_type 文字列 Yes
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

Request Body

Content-Type: application/*+json

Name タイプ Description Required Default
bytes 整数 Yes
filename 文字列 Yes
mime_type 文字列 Yes
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

Responses

状態コード: 200

説明: アップロードが正常に作成されました。

Content-Type Type Description
application/json UploadResource

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/uploads?api-version=2025-04-01-preview

{
 "purpose": "fine-tune",
 "filename": "training_data_v21.jsonl",
 "bytes": 2097152,
 "mime_type": "application/json"
}

応答: 状態コード: 200

{
  "headers": {
    "location": "https://aoairesource.openai.azure.com/openai/uploads/runs/fine-tune-72a2792ef7d24ba7b82c7fe4a37e379f"
  },
  "body": {
    "bytes": 2097152,
    "filename": "training_data_v21.jsonl",
    "purpose": "fine-tune",
    "status": "pending",
    "expires_at": 1646133327,
    "file": {
      "status": "pending",
      "bytes": 140,
      "purpose": "fine-tune",
      "filename": "puppy.jsonl",
      "id": "file-181a1cbdcdcf4677ada87f63a0928099",
      "created_at": 1646126127,
      "object": "file"
    },
    "id": "fine-tune-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "upload"
  }
}

ファイルのアップロード - キャンセル

POST {endpoint}/openai/uploads/{upload-id}/cancel?api-version=2025-04-01-preview

Upload を取り消し、アップロードされたすべてのパーツが非同期的に削除されます。

Upload が取り消された後、Parts を追加することはできません。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
upload-id パス Yes 文字列 アップロードの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json UploadResource

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/uploads/{upload-id}/cancel?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "bytes": 2097152,
    "filename": "training_data_v21.jsonl",
    "purpose": "fine-tune",
    "status": "cancelled",
    "expires_at": 1646133327,
    "file": {
      "status": "pending",
      "bytes": 140,
      "purpose": "fine-tune",
      "filename": "puppy.jsonl",
      "id": "file-181a1cbdcdcf4677ada87f63a0928099",
      "created_at": 1646126127,
      "object": "file"
    },
    "id": "fine-tune-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "upload"
  }
}

ファイルのアップロード - 完了

POST {endpoint}/openai/uploads/{upload-id}/complete?api-version=2025-04-01-preview

これで Upload が完了し、返された Upload オブジェクトには、プラットフォーム全体で使用できる入れ子になった File オブジェクトが含まれます。

Parts ID の順序付きリストを指定することで、Parts の順序を定義できます。

アップロードされた合計バイト数は、Upload オブジェクトの作成時に最初に指定したサイズと一致している必要があります。

この操作の後は、Upload が完了した後に追加の Parts を追加することはできません。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
upload-id パス Yes 文字列 アップロードの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: application/json

Name タイプ Description Required Default
md5 文字列 No
part_ids アレイ Yes

Request Body

Content-Type: text/json

Name タイプ Description Required Default
md5 文字列 No
part_ids アレイ Yes

Request Body

Content-Type: application/*+json

Name タイプ Description Required Default
md5 文字列 No
part_ids アレイ Yes

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json UploadResource

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/uploads/{upload-id}/complete?api-version=2025-04-01-preview

{
 "part_ids": [
  "LnmictL3p0u4LH/ko343nw==",
  "LmmictL3p0u4LH/ko343nw=="
 ]
}

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "bytes": 2097152,
    "filename": "training_data_v21.jsonl",
    "purpose": "fine-tune",
    "status": "completed",
    "expires_at": 1646133327,
    "file": {
      "status": "processed",
      "bytes": 140,
      "purpose": "fine-tune",
      "filename": "puppy.jsonl",
      "id": "file-181a1cbdcdcf4677ada87f63a0928099",
      "created_at": 1646126127,
      "object": "file"
    },
    "id": "fine-tune-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "upload"
  }
}

ファイルのアップロード - 部分

POST {endpoint}/openai/uploads/{upload-id}/parts?api-version=2025-04-01-preview

アップロード オブジェクトにパーツを追加します。各パーツは、アップロードするファイルのセグメントを表します。

各 Parts は、File の目的に基づいて、ファイルアップロードの標準サイズ制限に達することができます。 9 GB の Uplaod サイズ制限に達するまで、Parts の追加を続けることができます。

URI Parameters

Name In Required タイプ Description
エンドポイント パス Yes 文字列
url
サポートされている Azure OpenAI エンドポイント (プロトコルとホスト名:https://aoairesource.openai.azure.comなど)。"aoairesource" を Azure OpenAI リソース名) に置き換えます。 https://{your-resource-name}.openai.azure.com
upload-id パス Yes 文字列 アップロードの識別子。
api-version クエリ Yes 文字列 要求された API バージョン。

Request Header

トークン ベースの認証または API キーを使用します。 トークン ベースの認証を使用した認証が推奨され、より安全です。

Name Required タイプ Description
Authorization True 文字列 Example:Authorization: Bearer {Azure_OpenAI_Auth_Token}

Azure CLIを使用して認証トークンを生成するには: az account get-access-token --resource https://cognitiveservices.azure.com

Type: oauth2
承認 URL: https://login.microsoftonline.com/common/oauth2/v2.0/authorize
スコープ: https://ai.azure.com/.default
api-key True 文字列 ここで OpenAI API キー Azure指定します

Request Body

Content-Type: multipart/form-data

Name タイプ Description Required Default
データ 文字列 Yes

Responses

状態コード: 200

Description: Success

Content-Type Type Description
application/json UploadPartResource

状態コード: 既定値

説明: エラーが発生しました。

Content-Type Type Description
application/json ErrorResponse

Examples

POST {endpoint}/openai/uploads/{upload-id}/parts?api-version=2025-04-01-preview

応答: 状態コード: 200

{
  "headers": {},
  "body": {
    "upload_id": "fine-tune-72a2792ef7d24ba7b82c7fe4a37e379f",
    "azure_block_id": "LnmictL3p0u4LH/ko343nw==",
    "id": "uplprt-72a2792ef7d24ba7b82c7fe4a37e379f",
    "created_at": 1646126127,
    "object": "upload.part"
  }
}

Components

AzureAISearchIndex

Azure AI 検索インデックス。

Name タイプ Description Required Default
connectionId 文字列 Azure AI 検索 インデックスを指す接続の ID。 No
kind TargetType 対象の型。 Yes

BaseConnection

リソースへの接続。

BaseConnection の識別子

このコンポーネントでは、プロパティ kind を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
kind ConnectionType 接続の種類。 Yes

Batch

バッチの値を定義します。

Name タイプ Description Required Default
cancelled_at 整数 このバッチが取り消されたときのタイムスタンプ (UNIX エポック)。 No
cancelling_at 整数 このバッチが取り消しを開始したときのタイムスタンプ (UNIX エポック)。 No
completed_at 整数 このバッチが完了したときのタイムスタンプ (UNIX エポック)。 No
completion_window 文字列 バッチを処理する時間枠。 No
created_at 整数 このバッチが作成された時点のタイムスタンプ(UNIXエポック)。 No
エンドポイント 文字列 バッチによって使用される API エンドポイント。 No
error_blob 文字列 エラーが発生した要求の出力を含む BLOB URL。 No
error_file_id 文字列 エラーのある要求の出力を含むファイルの ID。 No
errors BatchErrors 失敗したバッチの場合、エラーの原因に関する詳細情報が含まれます。 No
expired_at 整数 このバッチの有効期限が切れた時点のタイムスタンプ (unix エポック)。 No
expires_at 整数 このバッチの有効期限が切れる時点のタイムスタンプ (unix エポック)。 No
failed_at 整数 このバッチが失敗した時点のタイムスタンプ (UNIX エポック)。 No
finalizing_at 整数 このバッチが終了処理を開始した時点のタイムスタンプ (UNIX エポック)。 No
id 文字列 この項目の ID。 Yes
in_progress_at 整数 このバッチが進行状態になった時点のタイムスタンプ (UNIX エポック)。 No
input_blob 文字列 バッチの入力ファイルを含む BLOB URL。 No
input_file_id 文字列 バッチの入力ファイルの ID。 Yes
メタデータ オブジェクト バッチにアタッチできるキーと値のペアのセット。 これは、バッチに関する追加情報を構造化された形式で格納する場合に役立ちます。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
output_blob 文字列 正常に実行された要求の出力を含む BLOB URL。 No
output_file_id 文字列 正常に実行された要求の出力を含むファイルの ID。 No
request_counts BatchRequestCounts バッチ内のさまざまな状態の要求の数。 No
状態 BatchStatus バッチの状態。 No

BatchCreateRequest

バッチを作成する要求を定義します。

Name タイプ Description Required Default
completion_window 文字列 バッチを処理する時間枠。 Yes
エンドポイント 文字列 バッチによって使用される API エンドポイント。 Yes
input_blob 文字列 バッチの入力として使用するAzure Storage BLOB の URL。 No
input_file_id 文字列 バッチの入力ファイルの ID。 No
メタデータ オブジェクト バッチにアタッチできるキーと値のペアのセット。 これは、バッチに関する追加情報を構造化された形式で格納する場合に役立ちます。 No
output_expires_after FileExpiresAfter ファイルの有効期限を定義します。 No
output_folder BatchOutputReference 出力を格納するAzure Storage フォルダー。 No

BatchErrorData

バッチ内の失敗のエラー情報。

Name タイプ Description Required Default
コード 文字列 エラーの種類を識別するエラー コード。 No
文字列 該当する場合、エラーが発生した入力ファイルの行番号 (null 値になる場合があります)。 No
メッセージ 文字列 エラーの詳細を提供する、人間が判読できるメッセージ。 No
param 文字列 該当する場合、エラーの原因となったパラメーターの名前 (null 値になる場合があります)。 No

BatchErrors

失敗したバッチの場合、エラーの原因に関する詳細情報が含まれます。

Name タイプ Description Required Default
データ BatchErrorData バッチ内の失敗のエラー情報。 No
オブジェクト 文字列 エラー オブジェクトの型。 これは常に 'list' です。 No

BatchRequestCounts

バッチ内のさまざまな状態の要求の数。

Name タイプ Description Required Default
completed 整数 バッチ内で正常に完了した要求の数。 No
failed 整数 バッチ内の失敗した要求の数。 No
total 整数 バッチ内の要求の合計数。 No

BatchOutputReference

出力を格納するAzure Storage フォルダー。

Name タイプ Description Required Default
区切り記号 文字列 Optional. 既定では、フォルダー パスで使用される区切り記号 /。 No
url 文字列 バッチ出力が保存されるAzure Storage フォルダーの URL。 No

BatchStatus

バッチの状態。

Property Value
Description バッチの状態。
Type 文字列
Values validating
failed
in_progress
finalizing
completed
expired
cancelling
cancelled

BatchesList

バッチの一覧を表します。

Name タイプ Description Required Default
データ アレイ 項目の一覧。 No
first_id 文字列 返されるバッチの一覧の最初のバッチの ID。 No
has_more boolean リストに返される要素より多くの要素が含まれているかどうかを示す値。 No
last_id 文字列 返されるバッチの一覧の最後のバッチの ID。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No

Capabilities

基本モデルまたは微調整モデルの機能。

Name タイプ Description Required Default
chat_completion boolean モデルがチャット完了をサポートしているかどうかを示す値。 Yes
completion boolean モデルが完了をサポートしているかどうかを示す値。 Yes
embeddings boolean モデルが埋め込みをサポートしているかどうかを示す値。 Yes
fine_tune boolean モデルを微調整に使用できるかどうかを示す値。 Yes
推論 boolean モデルをデプロイできるかどうかを示す値。 Yes

Choice

Name タイプ Description Required Default
finish_reason 文字列 モデルがトークンの生成を停止した理由。 No
インデックス 整数 選択肢のリスト内の選択肢のインデックス。 No
logprobs オブジェクト 選択肢のログ確率情報。 No
メッセージ ChoiceMessage No

ChoiceMessage

Name タイプ Description Required Default
コンテンツ 文字列 メッセージの内容。 No
ロール 文字列 メッセージを作成しているエンティティのロール。 No
tool_calls アレイ 関連するツール呼び出しの一覧。 No

ChunkingSettings

Chunking settings

Name タイプ Description Required Default
maxChunkSizeInTokens 整数 No

CompletionAction

完了アクション。

Property Value
Description 完了アクション。
Type 文字列
Values cleanUpTempAssets
keepAllAssets

ComputeType

コンピューティングの種類。

Property Value
Description コンピューティングの種類。
Type 文字列
Values ServerlessCompute
CustomCompute

ConnectionStringConnection

Connection string接続。

Name タイプ Description Required Default
connectionString 文字列 Connection string No
kind ConnectionType 接続の種類。 Yes

ConnectionType

接続の種類。

Property Value
Description 接続の種類。
Type 文字列
Values EndpointWithKey
ConnectionString
EndpointWithManagedIdentity
WorkspaceConnection
RelativeConnection

CosmosDBIndex

CosmosDB Index.

Name タイプ Description Required Default
collectionName 文字列 Cosmos DB コレクションの名前。 No
connectionId 文字列 Cosmos DB を指す接続の ID。 No
databaseName 文字列 Cosmos DB データベースの名前。 No
kind TargetType 対象の型。 Yes

CrawlingSettings

Crawling settings

Name タイプ Description Required Default
maxCrawlDepth 整数 No
maxCrawlTimeInMins 整数 No
maxDownloadTimeInMins 整数 No
maxFileSize 整数 No
maxFiles 整数 No
maxRedirects 整数 No

CreateEvalCompletionsRunDataSource

モデル サンプリング構成を記述する CompletionsRunDataSource オブジェクト。

Name タイプ Description Required Default
input_messages オブジェクト No
└─ item_reference 文字列 'item' 名前空間内の変数への参照 (例: 'item.name')。 No
└─ template アレイ プロンプトまたはコンテキストを形成するチャット メッセージの一覧。 'item' 名前空間への変数参照を含めることができます (例: {{item.name}})。 No
└─ type 列挙型 入力メッセージの種類。 template または item_reference のいずれかです。
使用可能な値: templateitem_reference
No
モデル 文字列 入力候補の生成に使用するモデルの名前 (例: 'o3-mini')。 No
sampling_params オブジェクト No
└─ max_completion_tokens 整数 生成された出力内のトークンの最大数。 No
└─ seed 整数 サンプリング中にランダム性を初期化するシード値。 No 42
└─ temperature number 温度が高いほど、出力のランダム性が高くなります。 No 1
└─ top_p number 核サンプリングの温度に代わるもの。1.0 にはすべてのトークンが含まれます。 No 1
ソース オブジェクト Yes
└─ type 列挙型 ソースの種類。 EvalJsonlFileContentSourceEvalJsonlFileIdSourceEvalStoredCompletionsSource のいずれかを指定できます。
使用可能な値: EvalJsonlFileContentSourceEvalJsonlFileIdSourceEvalStoredCompletionsSource
No
列挙型 実行データ ソースの種類。 常に completions です。
使用可能な値: completions
Yes

CreateEvalCustomDataSourceConfig

評価実行に使用されるデータ ソースのスキーマを定義する CustomDataSourceConfig オブジェクト。 このスキーマは、データの形状を次のように定義するために使用されます。テスト条件を設定するために使用されるデータ、および実行を作成する際に必要となるデータ。

Name タイプ Description Required Default
include_sample_schema boolean 評価で、(データ ソースから応答を生成することで) サンプル名前空間にデータを設定することを必要とするかどうか No False
item_schema オブジェクト データ ソース内の各行の json スキーマ。 Yes
列挙型 データ ソースの種類。 常に custom です。
使用可能な値: custom
Yes

CreateEvalItem

プロンプトまたはコンテキストを構成するチャット メッセージ。 "item" 名前空間 ({{item.name}} など) への変数参照を含めることができます。

Name タイプ Description Required Default
コンテンツ ['string', 'InputTextContent', 'object'] モデルへのテキスト入力 - テンプレート文字列を含めることができます。 Yes
ロール 列挙型 メッセージ入力の役割。 userassistantsystem、または developer のいずれかです。
使用可能な値: userassistantsystemdeveloper
Yes
列挙型 メッセージ入力の種類。 常に message です。
使用可能な値: message
No

CreateEvalJsonlRunDataSource

eval に一致する JSONL ファイルを指定する JsonlRunDataSource オブジェクト

Name タイプ Description Required Default
ソース オブジェクト Yes
└─ content アレイ jsonl ファイルの内容。 No
└─ id 文字列 ファイルの識別子。 No
└─ type 列挙型 jsonl ソースの型。 常に file_id です。
使用可能な値: file_id
No
列挙型 データ ソースの種類。 常に jsonl です。
使用可能な値: jsonl
Yes

CreateEvalLabelModelGrader

モデルを使用して評価の各項目にラベルを割り当てる LabelModelGrader オブジェクト。

Name タイプ Description Required Default
入力 アレイ プロンプトまたはコンテキストを形成するチャット メッセージの一覧。 "item" 名前空間 ({{item.name}} など) への変数参照を含めることができます。 Yes
labels アレイ 各評価項目を分類するためのラベル。 Yes
モデル 文字列 評価に使用するモデル。 構造化された出力をサポートする必要があります。 Yes
name 文字列 グレーダーの名前。 Yes
passing_labels アレイ 合格結果を示すラベル。 ラベルのサブセットである必要があります。 Yes
列挙型 オブジェクトの種類。これは常に label_model です。
使用可能な値: label_model
Yes

CreateEvalLogsDataSourceConfig

格納された補完クエリのメタデータプロパティを指定するデータソース構成。 これは通常、 usecase=chatbotprompt-version=v2などのメタデータです。

Name タイプ Description Required Default
メタデータ オブジェクト ログ データ ソースのメタデータ フィルター。 No
列挙型 データ ソースの種類。 常に logs です。
使用可能な値: logs
Yes

CreateEvalRequest

Name タイプ Description Required Default
data_source_config オブジェクト 評価実行に使用されるデータ ソースのスキーマを定義する CustomDataSourceConfig オブジェクト。 このスキーマは、データの形状を次のように定義するために使用されます。テスト条件を設定するために使用されるデータ、および実行を作成する際に必要となるデータ。 Yes
└─ include_sample_schema boolean 評価で、(データ ソースから応答を生成することで) サンプル名前空間にデータを設定することを必要とするかどうか No False
└─ item_schema オブジェクト データ ソース内の各行の json スキーマ。 No
└─ metadata オブジェクト ログ データ ソースのメタデータ フィルター。 No
└─ type 列挙型 データ ソースの種類。 常に logs です。
使用可能な値: logs
No
メタデータ Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 No
name 文字列 評価の名前。 No
testing_criteria アレイ このグループにおけるすべての評価実行のグレーダーの一覧。 Yes

CreateEvalRunRequest

Name タイプ Description Required Default
data_source オブジェクト eval に一致する JSONL ファイルを指定する JsonlRunDataSource オブジェクト Yes
└─ input_messages オブジェクト No
└─ item_reference 文字列 'item' 名前空間内の変数への参照 (例: 'item.name')。 No
└─ template アレイ プロンプトまたはコンテキストを形成するチャット メッセージの一覧。 'item' 名前空間への変数参照を含めることができます (例: {{item.name}})。 No
└─ type 列挙型 入力メッセージの種類。 template または item_reference のいずれかです。
使用可能な値: templateitem_reference
No
└─ model 文字列 入力候補の生成に使用するモデルの名前 (例: 'o3-mini')。 No
└─ sampling_params オブジェクト No
└─ max_completion_tokens 整数 生成された出力内のトークンの最大数。 No
└─ seed 整数 サンプリング中にランダム性を初期化するシード値。 No 42
└─ temperature number 温度が高いほど、出力のランダム性が高くなります。 No 1
└─ top_p number 核サンプリングの温度に代わるもの。1.0 にはすべてのトークンが含まれます。 No 1
└─ source オブジェクト No
└─ type 列挙型 ソースの種類。 EvalJsonlFileContentSourceEvalJsonlFileIdSourceEvalStoredCompletionsSource のいずれかを指定できます。
使用可能な値: EvalJsonlFileContentSourceEvalJsonlFileIdSourceEvalStoredCompletionsSource
No
└─ type 列挙型 実行データ ソースの種類。 常に completions です。
使用可能な値: completions
No
メタデータ Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 No
name 文字列 実行の名前。 No

CreateEvalStoredCompletionsRunDataSource

フィルターのセットを記述する StoredCompletionsRunDataSource 構成

Name タイプ Description Required Default
created_after 整数 この時刻より後に作成された項目をフィルター処理するためのオプションの Unix タイムスタンプ。 No
created_before 整数 この時刻より前に作成された項目をフィルター処理するためのオプションの Unix タイムスタンプ。 No
limit 整数 返される項目の最大数 (省略可能)。 No
メタデータ Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 No
モデル 文字列 選択可能なフィルター用モデル (例: "gpt-4o")。 No
列挙型 データ ソースの種類。 常に stored_completions です。
使用可能な値: stored_completions
Yes

CustomCompute

Custom compute.

Name タイプ Description Required Default
computeId 文字列 カスタム コンピューティングの ID No
kind ComputeType コンピューティングの種類。 Yes

DeploymentConnection

相対デプロイ接続。

Name タイプ Description Required Default
kind ConnectionType 接続の種類。 Yes

Deprecation

モデルのさまざまなユース ケースの非推奨の日付を定義します。

通常、基本モデルでは、作成後 1 年間の微調整がサポートされます。 推論は通常、ベースモデルまたは微調整されたモデルの作成から 2 年後にサポートされます。 正確な日付はプロパティで指定されます。

Name タイプ Description Required Default
fine_tune 整数 このモデルの微調整サポートの終了日。 モデルを微調整するために null になります。 No
推論 整数 このモデルの推論サポートの終了日。 Yes

DpoHyperparamsRequest

微調整の DPO 方法のハイパーパラメーター

Name タイプ Description Required Default
batch_size 整数 トレーニングに使用するバッチ サイズ。 バッチ サイズは、単一のフォワード パスおよびバックワード パスのトレーニングに使用されるトレーニング例の数です。 一般に、大規模なデータセットでは、バッチ サイズが大きいほど適切に機能する傾向があることがわかりました。既定値とこのプロパティの最大値は、基本モデルに固有です。 No
ベータ number DPO ベータ値。 No
l2_multiplier number L2 正則化乗数 No
learning_rate_multiplier number トレーニングに使用する学習率乗数。 微調整の学習率は、事前トレーニングに使用された元の学習率にこの値を掛けた値です。 学習率が高いほど、大きいバッチ サイズでのパフォーマンスが向上する傾向があります。 0.02 から 0.2 の範囲の値を試して、最適な結果が得られるものを確認することをお勧めします。 No
n_epochs 整数 モデルをトレーニングするエポックの数。 エポックとは、トレーニング データセット全体での 1 つの完全なサイクルを指します。 No

DpoMethod

DPO メソッドの微調整の要求

Name タイプ Description Required Default
dpo DpoMethodConfigRequest dpo 微調整方法の構成。 DPO 固有のハイパーパラメーターを含む Yes
FineTuningMethodType Yes

DpoMethodConfigRequest

dpo 微調整方法の構成。 DPO 固有のハイパーパラメーターを含む

Name タイプ Description Required Default
hyperparameters DpoHyperparamsRequest 微調整の DPO 方法のハイパーパラメーター No

DpoMethodRequest

DPO メソッドの微調整の要求

Name タイプ Description Required Default
dpo DpoMethodConfigRequest dpo 微調整方法の構成。 DPO 固有のハイパーパラメーターを含む Yes
FineTuningMethodType Yes

EndpointKeyConnection

エンドポイント キー接続。

Name タイプ Description Required Default
エンドポイント 文字列 Endpoint No
キー 文字列 Key No
kind ConnectionType 接続の種類。 Yes

EndpointMIConnection

エンドポイント マネージド ID 接続。

Name タイプ Description Required Default
エンドポイント 文字列 Endpoint No
kind ConnectionType 接続の種類。 Yes
resourceId 文字列 Resource Id No

Error

Microsoft REST ガイドラインで定義されているエラーの内容。

Name タイプ Description Required Default
コード ErrorCode Microsoft REST ガイドラインで定義されているエラー コード。 Yes
details アレイ エラーの詳細 (使用可能な場合)。 No
innererror InnerError Microsoft REST ガイドラインで定義されている内部エラー。 No
メッセージ 文字列 このエラーのメッセージ。 Yes
ターゲット 文字列 エラーが発生した場所 (使用可能な場合)。 No

ErrorCode

Microsoft REST ガイドラインで定義されているエラー コード。

Property Value
Description Microsoft REST ガイドラインで定義されているエラー コード。
Type 文字列
Values conflict
invalidPayload
forbidden
notFound
unexpectedEntityState
itemDoesAlreadyExist
serviceUnavailable
internalFailure
quotaExceeded
jsonlValidationFailed
fileImportFailed
tooManyRequests
unauthorized
contentFilter

ErrorResponse

Microsoft REST ガイドラインで定義されているエラー応答。

Name タイプ Description Required Default
エラー Error Microsoft REST ガイドラインで定義されているエラーの内容。 Yes

Eval

データ ソース構成とテスト条件を持つ Eval オブジェクト。 Eval は、LLM 統合のために実行するタスクを表します。 Like:

  • チャットボットの品質を向上させる
  • チャットボットがカスタマー サポートをどの程度適切に処理しているかを確認する
  • o3-miniがgpt-4oよりも私のユースケースで優れているかどうかを確認する
Name タイプ Description Required Default
created_at 整数 eval が作成されたときの Unix タイムスタンプ (秒単位)。 Yes
data_source_config オブジェクト itemのスキーマを指定し、必要に応じてsampleの名前空間を指定するCustomDataSourceConfig。 応答スキーマは以下のデータの形状を定義します:- テスト基準を定義するために使用されるデータ、および- 実行を作成する際に必要なデータ。 Yes
└─ metadata Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 No
└─ schema オブジェクト 実行データ ソース項目の json スキーマ。 JSON スキーマを作成する方法については 、こちらをご覧ください No
└─ type 列挙型 データ ソースの種類。 常に stored_completions です。
使用可能な値: stored_completions
No
id 文字列 評価の一意識別子。 Yes
メタデータ Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 Yes
name 文字列 評価の名前。 Yes
オブジェクト 列挙型 オブジェクトの種類。
使用可能な値: eval
Yes
testing_criteria アレイ テスト条件の一覧。 Yes

EvalApiError

Eval API からのエラー応答を表すオブジェクト。

Name タイプ Description Required Default
コード 文字列 エラー コード。 Yes
メッセージ 文字列 エラー メッセージ。 Yes

EvalCustomDataSourceConfig

itemのスキーマを指定し、必要に応じてsampleの名前空間を指定するCustomDataSourceConfig。 応答スキーマは以下のデータの形状を定義します:- テスト基準を定義するために使用されるデータ、および- 実行を作成する際に必要なデータ。

Name タイプ Description Required Default
スキーマ オブジェクト 実行データ ソース項目の json スキーマ。 JSON スキーマを作成する方法については 、こちらをご覧ください Yes
列挙型 データ ソースの種類。 常に custom です。
使用可能な値: custom
Yes

EvalFileRunDataSource

ファイル ID を持つ FileRunDataSource 構成

Name タイプ Description Required Default
file_id 文字列 ファイルの識別子。 Yes
列挙型 データ ソースの種類。 常に file です。
使用可能な値: file
Yes

EvalItem

命令フォロー階層を示すロールを含むモデルへのメッセージ入力。 developerロールまたはsystemロールで指定された命令は、userロールで指定された命令よりも優先されます。 assistant ロールを持つメッセージは、以前の操作でモデルによって生成されたものと見なされます。

Name タイプ Description Required Default
コンテンツ オブジェクト モデルへのテキスト入力。 Yes
└─ text 文字列 モデルからのテキスト出力。 No
└─ type 列挙型 出力テキストの型。 常に output_text です。
使用可能な値: output_text
No
ロール 列挙型 メッセージ入力の役割。 userassistantsystem、または developer のいずれかです。
使用可能な値: userassistantsystemdeveloper
Yes
列挙型 メッセージ入力の種類。 常に message です。
使用可能な値: message
No

EvalJsonlFileContentSource

Name タイプ Description Required Default
コンテンツ アレイ jsonl ファイルの内容。 Yes
列挙型 jsonl ソースの型。 常に file_content です。
使用可能な値: file_content
Yes

EvalJsonlFileIdSource

Name タイプ Description Required Default
id 文字列 ファイルの識別子。 Yes
列挙型 jsonl ソースの型。 常に file_id です。
使用可能な値: file_id
Yes

EvalLabelModelGrader

モデルを使用して評価の各項目にラベルを割り当てる LabelModelGrader オブジェクト。

Name タイプ Description Required Default
入力 アレイ Yes
labels アレイ 評価の各項目に割り当てるラベル。 Yes
モデル 文字列 評価に使用するモデル。 構造化された出力をサポートする必要があります。 Yes
name 文字列 グレーダーの名前。 Yes
passing_labels アレイ 合格結果を示すラベル。 ラベルのサブセットである必要があります。 Yes
列挙型 オブジェクトの種類。これは常に label_model です。
使用可能な値: label_model
Yes

EvalList

評価のリストを表すオブジェクト。

Name タイプ Description Required Default
データ アレイ 評価オブジェクトの配列。 Yes
first_id 文字列 データ配列内の最初の評価の識別子。 Yes
has_more boolean 選択可能な評価がこれ以上存在するかどうかを示します。 Yes
last_id 文字列 データ配列の最後の評価の識別子。 Yes
オブジェクト 列挙型 このオブジェクトの種類。 常に "list" に設定されます。
使用可能な値: list
Yes

EvalRun

評価実行を表すスキーマ。

Name タイプ Description Required Default
created_at 整数 評価実行が作成されたときの Unix タイムスタンプ (秒単位)。 Yes
data_source オブジェクト eval に一致する JSONL ファイルを指定する JsonlRunDataSource オブジェクト Yes
└─ input_messages オブジェクト No
└─ item_reference 文字列 'item' 名前空間内の変数への参照 (例: 'item.name')。 No
└─ template アレイ プロンプトまたはコンテキストを形成するチャット メッセージの一覧。 'item' 名前空間への変数参照を含めることができます (例: {{item.name}})。 No
└─ type 列挙型 入力メッセージの種類。 template または item_reference のいずれかです。
使用可能な値: templateitem_reference
No
└─ model 文字列 入力候補の生成に使用するモデルの名前 (例: 'o3-mini')。 No
└─ sampling_params オブジェクト No
└─ max_completion_tokens 整数 生成された出力内のトークンの最大数。 No
└─ seed 整数 サンプリング中にランダム性を初期化するシード値。 No 42
└─ temperature number 温度が高いほど、出力のランダム性が高くなります。 No 1
└─ top_p number 核サンプリングの温度に代わるもの。1.0 にはすべてのトークンが含まれます。 No 1
└─ source オブジェクト No
└─ type 列挙型 ソースの種類。 EvalJsonlFileContentSourceEvalJsonlFileIdSourceEvalStoredCompletionsSource のいずれかを指定できます。
使用可能な値: EvalJsonlFileContentSourceEvalJsonlFileIdSourceEvalStoredCompletionsSource
No
└─ type 列挙型 実行データ ソースの種類。 常に completions です。
使用可能な値: completions
No
エラー EvalApiError Eval API からのエラー応答を表すオブジェクト。 Yes
eval_id 文字列 関連付けられている評価の識別子。 Yes
id 文字列 評価実行の一意識別子。 Yes
メタデータ Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 Yes
モデル 文字列 評価されるモデル (該当する場合)。 Yes
name 文字列 評価実行の名前。 Yes
オブジェクト 列挙型 オブジェクトの型。 Always "eval.run".
使用可能な値: eval.run
Yes
per_model_usage アレイ 評価実行中の各モデルの使用状況の統計情報。 Yes
per_testing_criteria_results アレイ 評価の実行中に適用されるテスト条件ごとの結果。 Yes
report_url 文字列 UI ダッシュボードに表示される評価実行レポートの URL。 Yes
result_counts オブジェクト 評価実行の結果を要約するカウンター。 Yes
└─ errored 整数 エラーが発生した出力項目の数。 No
└─ failed 整数 評価に合格しなかった出力項目の数。 No
└─ passed 整数 評価に合格した出力項目の数。 No
└─ total 整数 実行された出力項目の合計数。 No
状態 文字列 評価実行の状態。 Yes

EvalRunList

評価の実行の一覧を表すオブジェクト。

Name タイプ Description Required Default
データ アレイ eval 実行オブジェクトの配列。 Yes
first_id 文字列 データ配列で最初に実行された評価の識別子。 Yes
has_more boolean 選択可能な評価がこれ以上存在するかどうかを示します。 Yes
last_id 文字列 データ配列で最後に実行された評価の識別子。 Yes
オブジェクト 列挙型 このオブジェクトの種類。 常に "list" に設定されます。
使用可能な値: list
Yes

EvalRunOutputItem

評価実行出力項目を表すスキーマ。

Name タイプ Description Required Default
created_at 整数 評価実行が作成されたときの Unix タイムスタンプ (秒単位)。 Yes
datasource_item オブジェクト 入力データ ソース項目の詳細。 Yes
datasource_item_id 整数 データ ソース項目の識別子。 Yes
eval_id 文字列 評価グループの識別子。 Yes
id 文字列 評価実行出力項目のユニークな識別子。 Yes
オブジェクト 列挙型 オブジェクトの型。 Always "eval.run.output_item".
使用可能な値: eval.run.output_item
Yes
results アレイ 評価実行の結果の一覧。 Yes
run_id 文字列 この出力項目に関連付けられている評価実行の識別子。 Yes
サンプル オブジェクト 評価実行の入力と出力を含むサンプル。 Yes
└─ error EvalApiError Eval API からのエラー応答を表すオブジェクト。 No
└─ finish_reason 文字列 サンプルの生成が完了した理由。 No
└─ input アレイ 入力メッセージの配列。 No
└─ max_completion_tokens 整数 完了に許可されるトークンの最大数。 No
└─ model 文字列 サンプルの生成に使用されるモデル。 No
└─ output アレイ 出力メッセージの配列。 No
└─ seed 整数 サンプルの生成に使用されるシード。 No
└─ temperature number 使用されるサンプリング温度。 No
└─ top_p number サンプリングに使用されるtop_p値。 No
└─ usage オブジェクト サンプルのトークン使用状況の詳細。 No
└─ cached_tokens 整数 キャッシュから取得されたトークンの数。 No
└─ completion_tokens 整数 生成された完了トークンの数。 No
└─ prompt_tokens 整数 使用されるプロンプト トークンの数。 No
└─ total_tokens 整数 使用されたトークンの合計数。 No
状態 文字列 評価実行の状態。 Yes

EvalRunOutputItemList

評価実行の出力項目の一覧を表すオブジェクト。

Name タイプ Description Required Default
データ アレイ eval 実行出力項目オブジェクトの配列。 Yes
first_id 文字列 データ配列内の最初の評価実行出力項目の識別子。 Yes
has_more boolean 使用可能な eval 実行出力項目が増えるかどうかを示します。 Yes
last_id 文字列 データ配列内の最後の eval 実行出力項目の識別子。 Yes
オブジェクト 列挙型 このオブジェクトの種類。 常に "list" に設定されます。
使用可能な値: list
Yes

EvalScoreModelGrader

モデルを使用して入力にスコアを割り当てる ScoreModelGrader オブジェクト。

Name タイプ Description Required Default
入力 アレイ 入力テキスト。 これには、テンプレート文字列が含まれる場合があります。 Yes
モデル 文字列 評価に使用するモデル。 Yes
name 文字列 グレーダーの名前。 Yes
pass_threshold number スコアのしきい値。 No
範囲 アレイ スコアの範囲。 既定値は [0, 1] です。 No
sampling_params オブジェクト モデルのサンプリング パラメーター。 No
列挙型 オブジェクトの種類。これは常に score_model です。
使用可能な値: score_model
Yes

EvalStoredCompletionsDataSourceConfig

格納されている完了クエリのメタデータプロパティを指定するStoredCompletionsDataSourceConfigの構成。 これは通常、 usecase=chatbotprompt-version=v2などのメタデータです。このデータ ソース構成によって返されるスキーマは、evals で使用可能な変数を定義するために使用されます。 item このデータ ソース構成を使用する場合、 sample は両方とも定義されます。

Name タイプ Description Required Default
メタデータ Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 No
スキーマ オブジェクト 実行データ ソース項目の json スキーマ。 JSON スキーマを作成する方法については 、こちらをご覧ください Yes
列挙型 データ ソースの種類。 常に stored_completions です。
使用可能な値: stored_completions
Yes

EvalStoredCompletionsSource

フィルターのセットを記述する StoredCompletionsRunDataSource 構成

Name タイプ Description Required Default
created_after 整数 この時刻より後に作成された項目をフィルター処理するためのオプションの Unix タイムスタンプ。 No
created_before 整数 この時刻より前に作成された項目をフィルター処理するためのオプションの Unix タイムスタンプ。 No
limit 整数 返される項目の最大数 (省略可能)。 No
メタデータ Metadata オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。 No
モデル 文字列 選択可能なフィルター用モデル (例: "gpt-4o")。 No
列挙型 ソースの種類。 常に stored_completions です。
使用可能な値: stored_completions
Yes

EvalStringCheckGrader

指定した操作を使用して入力と参照の間で文字列比較を実行する StringCheckGrader オブジェクト。

Name タイプ Description Required Default
入力 文字列 入力テキスト。 これには、テンプレート文字列が含まれる場合があります。 Yes
name 文字列 グレーダーの名前。 Yes
操作 列挙型 実行する文字列チェック操作。 eqnelike、または ilike のいずれかです。
使用可能な値: eqnelikeilike
Yes
参照 文字列 参照テキスト。 これには、テンプレート文字列が含まれる場合があります。 Yes
列挙型 オブジェクトの種類。これは常に string_check です。
使用可能な値: string_check
Yes

EvalTextSimilarityGrader

類似性メトリックに基づいてテキストを評価する TextSimilarityGrader オブジェクト。

Name タイプ Description Required Default
evaluation_metric 列挙型 使用する評価メトリック。 fuzzy_matchbleugleumeteorrouge_1rouge_2rouge_3rouge_4rouge_5、またはrouge_lのいずれか。
使用できる値: fuzzy_matchbleugleumeteorrouge_1rouge_2rouge_3rouge_4rouge_5rouge_l
Yes
入力 文字列 採点評価中のテキスト。 Yes
name 文字列 グレーダーの名前。 No
pass_threshold number それ以上の値が合格点を示す浮動小数点スコア。 Yes
参照 文字列 採点対象のテキスト。 Yes
列挙型 グレーダーの種類。
使用可能な値: text_similarity
Yes

EventType

コンテンツ フィルター結果の重大度を定義します。

Property Value
Description コンテンツ フィルター結果の重大度を定義します。
Type 文字列
Values message
metrics

File

ファイルとは、トレーニングと検証に使用できるドキュメントです。 また、結果の詳細を含むサービス生成ドキュメントにすることもできます。

Name タイプ Description Required Default
bytes 整数 使用可能な場合は、このファイルのサイズ (null 値を指定できます)。 JavaScript 整数との互換性を確保するために、2^53-1 より大きいファイル サイズはサポートされていません。 No
created_at 整数 このジョブまたは項目が作成されたときのタイムスタンプ (UNIX エポック)。 No
expires_at 整数 ファイルの有効期限が切れた UNIX タイムスタンプ (1970 年 1 月 1 日以降に経過した秒数)。 No
filename 文字列 ファイルの名前。 Yes
id 文字列 この項目の ID。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes
状態 FileState ファイルの状態。 No
status_details 文字列 このファイルの処理に失敗した場合の詳細を含むエラー メッセージ。 Deprecated. No

FileCreate

OpenAI で使用できる外部コンテンツ URL からインポートするドキュメントAzure定義します。

Name タイプ Description Required Default
content_url 文字列 ドキュメントをダウンロードする URL (BLOB の SAS URL、または GET 要求でアクセスできるその他の外部 URL を指定できます)。 Yes
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

FileDelete

ファイル削除の応答を定義します。

Name タイプ Description Required Default
deleted boolean ファイルが削除されたかどうかを取得するかどうかを示す値。 Yes
id 文字列 削除されたファイル ID。 Yes
オブジェクト 文字列 ファイルが削除された場合。 Yes

FileDetails

ファイル詳細は、バッチサービスによって使用されるドキュメントで、ファイルのBLOB詳細を取得するために使用されます。

Name タイプ Description Required Default
filename 文字列 ファイルの名前。 Yes
id 文字列 この項目の ID。 No
パス 文字列 コンテナー内のファイルへの相対パス。 No
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes
storage_account_uri 文字列 このファイルのstorage アカウントの URL。 No
storage_container_name 文字列 このファイル BLOB の Storage コンテナー名。 No

FileExpiresAfter

ファイルの有効期限を定義します。

Name タイプ Description Required Default
アンカー FileExpiryAnchor 絶対有効期限を生成する基準となる時刻に対するアンカーを定義します。 No
seconds 整数 相対有効期限 (秒単位)。 範囲: [1209600 - 2592000]. No

FileExpiryAnchor

絶対有効期限を生成する基準となる時刻に対するアンカーを定義します。

Property Value
Description 絶対有効期限を生成する基準となる時刻に対するアンカーを定義します。
Type 文字列
Values created_at

FileImport

OpenAI で使用できる外部コンテンツ URL からインポートするドキュメントAzure定義します。

Name タイプ Description Required Default
content_url 文字列 ドキュメントをダウンロードする URL (BLOB の SAS URL、または GET 要求でアクセスできるその他の外部 URL を指定できます)。 Yes
expires_after FileExpiresAfter ファイルの有効期限を定義します。 No
filename 文字列 アップロードする JSON 行 ファイルの名前 purpose が "微調整" に設定されている場合、各行は、トレーニング例を表す "prompt" フィールドと "completion" フィールドを含む JSON レコードです。 Yes
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

FileList

ファイルの一覧を表します。

Name タイプ Description Required Default
データ アレイ 項目の一覧。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No

FileState

ファイルの状態。

Property Value
Description ファイルの状態。
Type 文字列
Values uploaded
pending
running
processed
error
deleting
deleted

FineTuneMethod

FineTuneMethod の識別子

このコンポーネントでは、プロパティ type を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
FineTuningMethodType Yes

FineTuneMethodRequest

FineTuneMethodRequest の識別子

このコンポーネントでは、プロパティ type を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
FineTuningMethodType Yes

FineTuningHyperParameters

微調整ジョブで使用されるハイパー パラメーター設定。

Name タイプ Description Required Default
batch_size 整数 トレーニングに使用するバッチ サイズ。 バッチ サイズは、単一のフォワード パスおよびバックワード パスのトレーニングに使用されるトレーニング例の数です。 一般的に、大規模なデータセットでは、バッチ サイズが大きいほど適切に機能する傾向があることがわかっています。 このプロパティの既定値と最大値は、基本モデルに固有です。 No
learning_rate_multiplier number トレーニングに使用する学習率乗数。 微調整の学習率は、事前トレーニングに使用された元の学習率にこの値を掛けた値です。 学習率が高いほど、大きいバッチ サイズでのパフォーマンスが向上する傾向があります。 0.02 から 0.2 の範囲の値を試して、最適な結果が得られるものを確認することをお勧めします。 No
n_epochs 整数 モデルをトレーニングするエポックの数。 エポックとは、トレーニング データセット全体での 1 つの完全なサイクルを指します。 No

FineTuningJob

微調整タスクの値を定義します。

Name タイプ Description Required Default
created_at 整数 このジョブまたは項目が作成されたときのタイムスタンプ (UNIX エポック)。 No
エラー FineTuningJobError 失敗したジョブを微調整する場合は、失敗の原因に関する詳細情報が含まれます。 No
estimated_finish 整数 微調整ジョブが完了すると推定される場合の Unix タイムスタンプ (秒単位)。 微調整ジョブが実行されていない場合、値は null になります。 No
fine_tuned_model 文字列 結果として得られる微調整されたモデルの識別子 (model-id)。 このプロパティは、正常に完了した微調整の実行に対してのみ設定されます。 推論用のデプロイを作成するには、この識別子を使用します。 No
finished_at 整数 このジョブまたは項目が正常に完了したときのタイムスタンプ (UNIX エポック)。 No
hyperparameters FineTuningHyperParameters 微調整ジョブで使用されるハイパー パラメーター設定。 No
id 文字列 この項目の ID。 No
integrations アレイ 微調整ジョブをサポートする統合の構成の一覧。 多くの統合が計画されているため、必ず統合の種類を確認して操作してください。 No
メソッド オブジェクト 監視対象の微調整に使用されるメソッド No
└─ type FineTuningMethodType No
モデル 文字列 微調整に使用される基本モデルの識別子 (model-id)。 Yes
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
organisation_id 文字列 この微調整ジョブの組織 ID。 Azure OpenAI では使用されません。OpenAI のみの互換性。 No
result_files アレイ トレーニングと評価のメトリックを csv 形式で含む結果ファイル ID (ファイル ID)。 このファイルは、正常に完了した微調整の実行でのみ使用できます。 No
seed 整数 微調整ジョブに使用されるシード。 No
状態 FineTuningState 微調整オブジェクトまたは微調整ジョブの状態。 No
suffix 文字列 微調整されたモデルを識別するために使用されるサフィックス。 サフィックスには最大 40 文字 (a-z、A-Z、0-9、-、_) を指定することができ、微調整したモデル名に追加されます。 No
trained_tokens 整数 この微調整ジョブによって処理された課金対象トークンの合計数。 No
training_file 文字列 トレーニングに使用されるファイル。 Yes
validation_file 文字列 トレーニング中に微調整されたモデルを評価するために使用されるファイル。 No

FineTuningJobCheckpoint

Name タイプ Description Required Default
created_at 整数 チェックポイントが作成されたときの UNIX タイムスタンプ (秒単位)。 Yes
fine_tuned_model_checkpoint 文字列 作成される微調整されたチェックポイント モデルの名前。 Yes
fine_tuning_job_id 文字列 このチェックポイントが作成された微調整ジョブの名前。 Yes
id 文字列 このチェックポイントの ID。 No
メトリクス オブジェクト 微調整ジョブ中のステップ番号のメトリック。 Yes
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
step_number 整数 チェックポイントが作成されたステップ番号。 Yes

FineTuningJobCheckpointList

チェックポイントの一覧を表します。

Name タイプ Description Required Default
データ アレイ 項目の一覧。 No
has_more boolean リストに返される要素より多くの要素が含まれているかどうかを示す値。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No

FineTuningJobCreation

微調整タスクの値を定義します。

Name タイプ Description Required Default
hyperparameters FineTuningHyperParameters 微調整ジョブで使用されるハイパー パラメーター設定。 No
integrations アレイ 微調整ジョブをサポートする統合の構成の一覧。 多くの統合が計画されているため、必ず統合の種類を確認して操作してください。 No
メソッド オブジェクト 監視対象の微調整に使用されるメソッド No
└─ type FineTuningMethodType No
モデル 文字列 この微調整に使用される基本モデルの識別子 (model-id)。 Yes
seed 整数 微調整ジョブに使用されるシード。 同じシードとジョブ パラメーターを渡すと同じ結果が得られますが、まれに異なる場合があります。 シードが指定されていない場合は生成されます。 No
suffix 文字列 微調整されたモデルを識別するために使用されるサフィックス。 サフィックスには最大 40 文字 (a-z、A-Z、0-9、-、_) を指定することができ、微調整したモデル名に追加されます。 No
training_file 文字列 この微調整されたモデルのトレーニングに使用されるファイル ID (ファイル ID)。 Yes
validation_file 文字列 トレーニング中に微調整されたモデルを評価するために使用されるファイル ID (ファイル ID)。 No

FineTuningJobError

失敗したジョブを微調整する場合は、失敗の原因に関する詳細情報が含まれます。

Name タイプ Description Required Default
コード 文字列 コンピューターが読み取り可能なエラー コード。 No
メッセージ 文字列 人間が判読できるエラー メッセージ。 No
param 文字列 無効なパラメーター (通常は training_file または validation_file)。 エラーがパラメーター固有でない場合、このフィールドは null 値になります。 No

FineTuningJobEvent

Name タイプ Description Required Default
created_at 整数 このイベントが作成されたときのタイムスタンプ (UNIX エポック)。 Yes
データ このイベントのマシン読み取り可能なデータ。 No
id 文字列 このイベントの ID。 No
レベル LogLevel イベントの詳細レベル。 Yes
メッセージ 文字列 イベントを説明するメッセージ。 これは、エンキュー、開始、失敗、完了、アップロードされた結果などの他のイベントなど、状態の変化です。 Yes
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
EventType コンテンツ フィルター結果の重大度を定義します。 Yes

FineTuningJobEventList

イベントの一覧を表します。

Name タイプ Description Required Default
データ アレイ 項目の一覧。 No
has_more boolean リストに返される要素より多くの要素が含まれているかどうかを示す値。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No

FineTuningJobList

微調整の一覧を表します。

Name タイプ Description Required Default
データ アレイ 項目の一覧。 No
has_more boolean リストに返される要素より多くの要素が含まれているかどうかを示す値。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No

FineTuningMethodType

Property Value
Type 文字列
Values dpo
supervised

FineTuningState

微調整オブジェクトまたは微調整ジョブの状態。

Property Value
Description 微調整オブジェクトまたは微調整ジョブの状態。
Type 文字列
Values created
pending
running
pausing
paused
resuming
succeeded
cancelled
failed

FunctionCallFunction

Name タイプ Description Required Default
arguments 文字列 モデルが関数に渡すことを予期する引数。 No
name 文字列 関数の名前です。 No

FunctionDefinition

Name タイプ Description Required Default
description 文字列 モデルが関数をいつどのように呼び出すかを選ぶために使われる、関数の動作の説明。 No
name 文字列 呼び出す関数の名前。 No
parameters 関数が受け入れるパラメーター (JSON スキーマ オブジェクトとして記述)。 No

GenericEmbeddingSettings

接続埋め込み設定

Name タイプ Description Required Default
接続 BaseConnection リソースへの接続。 No
deploymentName 文字列 No
modelName 文字列 No

IngestionError

インジェスト エラーの詳細

Name タイプ Description Required Default
innerErrors アレイ No
メッセージ 文字列 No

IngestionJob

ジョブの詳細を表します。

IngestionJob の識別子

このコンポーネントでは、プロパティ kind を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
jobId 文字列 No
kind IngestionJobType ジョブの種類。 Yes

IngestionJobList

インジェスト ジョブの一覧を表します。

Name タイプ Description Required Default
nextLink 文字列 次のページへのリンク。 No
value アレイ 項目の一覧。 No

IngestionJobProgress

Name タイプ Description Required Default
stageProgress アレイ Yes

IngestionJobRun

ジョブ実行の詳細

Name タイプ Description Required Default
エラー IngestionError インジェスト エラーの詳細 No
jobId 文字列 No
progress IngestionJobProgress No
runId 文字列 No
状態 OperationState ジョブまたはアイテムの状態。 No
warnings アレイ No

IngestionJobRunList

取り込みジョブ実行履歴の一覧を表します。

Name タイプ Description Required Default
nextLink 文字列 次のページへのリンク。 No
value アレイ 項目の一覧。 No

IngestionJobStageProgress

Name タイプ Description Required Default
name 文字列 No
processedItems 整数 No
状態 OperationState ジョブまたはアイテムの状態。 No
subStageProgress アレイ No
totalItems 整数 No

IngestionJobSystemCompute

Name タイプ Description Required Default
completionAction CompletionAction 完了アクション。 No
dataRefreshIntervalInHours 整数 No
datasource SystemComputeDatasource No
jobId 文字列 No
kind IngestionJobType ジョブの種類。 Yes
searchServiceConnection BaseConnection リソースへの接続。 No

IngestionJobType

ジョブの種類。

Property Value
Description ジョブの種類。
Type 文字列
Values SystemCompute
UserCompute

IngestionJobUserCompute

Name タイプ Description Required Default
コンピューティング JobCompute ジョブのコンピューティング設定。 No
dataRefreshIntervalInHours 整数 No
datasource UserComputeDatasource No
jobId 文字列 No
kind IngestionJobType ジョブの種類。 Yes
ターゲット TargetIndex 作成するインデックスに関する情報。 No
workspaceId 文字列 No

InnerError

Microsoft REST ガイドラインで定義されている内部エラー。

Name タイプ Description Required Default
コード InnerErrorCode Microsoft REST ガイドラインで定義されている内部エラー コード。 No
innererror InnerError Microsoft REST ガイドラインで定義されている内部エラー。 No

InnerErrorCode

Microsoft REST ガイドラインで定義されている内部エラー コード。

Property Value
Description Microsoft REST ガイドラインで定義されている内部エラー コード。
Type 文字列
Values invalidPayload

InputMessage

Name タイプ Description Required Default
コンテンツ 文字列 メッセージの内容。 No
id 文字列 メッセージの ID。 No
name 文字列 参加者の名前 (省略可能)。 同じロールの参加者を区別するためのモデル情報を指定します。 No
ロール 文字列 メッセージを作成しているエンティティのロール。 No

InputTextContent

モデルへのテキスト入力。

Name タイプ Description Required Default
SMS 送信 文字列 モデルへのテキスト入力。 Yes
列挙型 入力項目の種類。 常に input_text です。
使用可能な値: input_text
Yes

IntegrationTypes

使用可能な微調整統合の一覧。

Property Value
Description 使用可能な微調整統合の一覧。
Type 文字列
Values wandb

JobCompute

ジョブのコンピューティング設定。

JobCompute の識別子

このコンポーネントでは、プロパティ kind を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
kind ComputeType コンピューティングの種類。 Yes

LifeCycleStatus

モデルのライフ サイクルの状態。

注: モデルは preview から generally-availableに昇格できますが、 generally-available から previewに昇格することはできません。

Property Value
Description モデルのライフ サイクルの状態。 注: モデルは preview から generally-availableに昇格できますが、 generally-available から previewに昇格することはできません。
Type 文字列
Values preview
generally-available
deprecating
deprecated

LogLevel

イベントの詳細レベル。

Property Value
Description イベントの詳細レベル。
Type 文字列
Values info
warning
error

Logprob

Name タイプ Description Required Default
logprob number 上位 20 個の最も可能性の高いトークン内にある場合の、このトークンのログ確率。 それ以外の場合、トークンの可能性が非常に低いことを示すために、-9999.0 値が使用されます。 No
トークン 文字列 The token. No

Metadata

オブジェクトにアタッチできる 16 個のキーと値のペアのセット。 これは、構造化された形式でオブジェクトに関する追加情報を格納したり、API またはダッシュボードを使用してオブジェクトのクエリを実行したりする場合に役立ちます。 キーは、最大長が 64 文字の文字列です。 値は、最大長が 512 文字の文字列です。

このコンポーネントに対して定義されているプロパティはありません。

Model

モデルは、基本モデルであるか、成功したファインチューニングジョブの結果のいずれかです。

Name タイプ Description Required Default
capabilities Capabilities 基本モデルまたは微調整モデルの機能。 Yes
created_at 整数 このジョブまたは項目が作成されたときのタイムスタンプ (UNIX エポック)。 No
deprecation Deprecation モデルのさまざまなユース ケースの非推奨の日付を定義します。 通常、基本モデルでは、作成後 1 年間の微調整がサポートされます。推論は、通常、基本モデルまたは微調整されたモデルの作成から 2 年後にサポートされます。 正確な日付はプロパティで指定されます。 Yes
fine_tune 文字列 微調整モデルの場合の微調整ジョブ ID (微調整 ID)。それ以外の場合は null です。 No
id 文字列 この項目の ID。 No
lifecycle_status LifeCycleStatus モデルのライフ サイクルの状態。 注: モデルは "プレビュー" から generally-available に昇格できますが、 generally-available から "プレビュー" に昇格することはできません。 Yes
モデル 文字列 これが微調整モデルの場合の基本モデル ID (model-id)。それ以外の場合は null です。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
状態 FineTuningState 微調整オブジェクトまたは微調整ジョブの状態。 No

ModelList

モデルの一覧を表します。

Name タイプ Description Required Default
データ アレイ 項目の一覧。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No

OperationState

ジョブまたはアイテムの状態。

Property Value
Description ジョブまたはアイテムの状態。
Type 文字列
Values notRunning
running
succeeded
failed

Order

ファイルの目的を定義します。

Property Value
Description ファイルの目的を定義します。
Type 文字列
Values desc
asc

OrderByOrder

Property Value
Type 文字列
Values asc
desc

PineconeIndex

Pinecone Index.

Name タイプ Description Required Default
connectionId 文字列 パインコンを指す接続の ID。 No
kind TargetType 対象の型。 Yes

Purpose

アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。

Property Value
Description アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。
Type 文字列
Values fine-tune
fine-tune-results
assistants
assistants_output
batch
batch_output
evals
evals_output

ServerlessCompute

Serverless compute.

Name タイプ Description Required Default
instanceCount 整数 ジョブを実行するインスタンスの数。 No
kind ComputeType コンピューティングの種類。 Yes
sku 文字列 SKU Level No

StoredCompletion

Name タイプ Description Required Default
choices アレイ チャット入力候補の選択肢の一覧。 n が 1 より大きい場合は複数の値を指定できます。 No
created 整数 チャット入力候補が作成されたときの Unix タイムスタンプ (秒単位)。 No
frequency_penalty number -2.0 から 2.0 の数値。 値を正にすると、これまでのテキストに存在する頻度に基づいて新しいトークンにペナルティが課せられ、モデルが同じ行を逐語的に繰り返す可能性が低下します。 No
id 文字列 格納されている完了の ID。 No
input_user 文字列 この要求の入力ユーザー。 No
メタデータ オブジェクト 追加情報のための任意のキーと値のペア。 No
モデル 文字列 使用するモデルの ID。 No
presence_penalty number -2.0 から 2.0 の数値。 正の値を指定すると、これまでのテキストに出現するかどうかに基づいて新しいトークンにペナルティが課せられ、モデルが新しいトピックを扱う可能性が高まります。 No
request_id 文字列 OpenAI API 要求の一意識別子。 サポートに問い合わせる場合は、この要求 ID を含めてください。 No
seed 整数 指定されている場合、このシステムは、同じ seed とパラメーターを使用して繰り返された要求に対して同じ結果を返せるように、決定論的にサンプリングを行うように最善を尽くします。 決定主義は保証されておらず、以下を参照する必要があります。
system_fingerprint 応答パラメーターを使用してバックエンドの変更を監視します。 No
service_tier 文字列 要求の処理に使用する待機時間レベルを指定します。 No
system_fingerprint 文字列 このフィンガープリントは、モデルが実行されるバックエンド構成を表します。 seed 要求パラメーターと組み合わせて使うと、決定性に影響を与える可能性のあるバックエンドの変更がいつ行われたかを把握できます。 No
temperature number 使用するサンプリング温度 (0 から 2)。 0.8 のような大きい値にすると、出力はよりランダムになり、0.2 のような小さい値にすると、出力はより集中的および決定論的になります。 一般に、これと top_p の両方ではなく、いずれかを変更することをお勧めします。 No
tool_choice 文字列 モデルによって呼び出されるツールを制御します (ある場合)。 No
tools アレイ モデルによって呼び出されるツールの一覧。 現在のところ、関数のみがツールとしてサポートされています。 これを使用し、モデルによって JSON 入力が生成される関数の一覧を提供します。 最大 128 個の関数がサポートされています。 No
top_p number 温度によるサンプリングに代わる核サンプリングと呼ばれるもので、モデルは top_p の確率質量を持つトークンの結果を考慮します。 したがって、0.1 は、上位 10% の確率質量を構成するトークンのみが考慮されることを意味します。 一般に、これと temperature の両方ではなく、いずれかを変更することをお勧めします。 No
使用 Usage No

StoredCompletionDelete

Name タイプ Description Required Default
deleted boolean 格納された完了が削除されたかどうかを取得するかどうかを示す値。 No
id 文字列 削除された格納済みの完了 ID。 No
オブジェクト 文字列 格納された完了が削除された場合。 Yes

StoredCompletionList

Name タイプ Description Required Default
データ アレイ 取得したオブジェクトの一覧。 No
first_id 文字列 取得した list の最初の ID No
has_more boolean has_more プロパティは、改ページ位置の変更に使用され、追加の結果があることを示します。 No
last_id 文字列 取得された list の最後の ID No
オブジェクト 文字列 オブジェクトの種類。これは常に list です No
total 整数 アイテムの合計数。 No

StoredCompletionMessages

Name タイプ Description Required Default
データ アレイ 取得したオブジェクトの一覧。 No
first_id 文字列 取得した list の最初の ID No
has_more boolean has_more プロパティは、改ページ位置の変更に使用され、追加の結果があることを示します。 No
last_id 文字列 取得された list の最後の ID No
オブジェクト 文字列 オブジェクトの種類。これは常に list です No
total 整数 アイテムの合計数。 No

StoredCompletionResponse

Name タイプ Description Required Default
choices アレイ チャット入力候補の選択肢の一覧。 n が 1 より大きい場合は複数の値を指定できます。 No
created 整数 チャット入力候補が作成されたときの Unix タイムスタンプ (秒単位)。 No
frequency_penalty number -2.0 から 2.0 の数値。 値を正にすると、これまでのテキストに存在する頻度に基づいて新しいトークンにペナルティが課せられ、モデルが同じ行を逐語的に繰り返す可能性が低下します。 No
id 文字列 格納されている完了の ID。 No
input_user 文字列 この要求の入力ユーザー。 No
メタデータ オブジェクト 追加情報のための任意のキーと値のペア。 No
モデル 文字列 使用するモデルの ID。 No
オブジェクト 文字列 このオブジェクトの種類。 No
presence_penalty number -2.0 から 2.0 の数値。 正の値を指定すると、これまでのテキストに出現するかどうかに基づいて新しいトークンにペナルティが課せられ、モデルが新しいトピックを扱う可能性が高まります。 No
request_id 文字列 OpenAI API 要求の一意識別子。 サポートに問い合わせる場合は、この要求 ID を含めてください。 No
seed 整数 指定した場合、システムは決定論的にサンプリングするように最善を尽くします。これにより、同じ seedパラメーターを持つ要求が繰り返し実行され、同じ結果が返されます。 決定性は保証されないため、バックエンドでの変更を監視するには、system_fingerprint 応答パラメーターを参照する必要があります。 No
service_tier 文字列 要求の処理に使用する待機時間レベルを指定します。 No
system_fingerprint 文字列 このフィンガープリントは、モデルが実行されるバックエンド構成を表します。 seed 要求パラメーターと組み合わせて使うと、決定性に影響を与える可能性のあるバックエンドの変更がいつ行われたかを把握できます。 No
temperature number 使用するサンプリング温度 (0 から 2)。 0.8 のような値が大きいほど出力はランダムになり、0.2 のような値を小さくすると、より集中して決定論的になります。 一般に、これと top_p の両方ではなく、いずれかを変更することをお勧めします。 No
tool_choice 文字列 モデルによって呼び出されるツールを制御します (ある場合)。 No
tools アレイ モデルによって呼び出されるツールの一覧。 現在のところ、関数のみがツールとしてサポートされています。 これを使用し、モデルによって JSON 入力が生成される関数の一覧を提供します。 最大 128 個の関数がサポートされています。 No
top_p number 温度によるサンプリングに代わる核サンプリングと呼ばれるもので、モデルは top_p の確率質量を持つトークンの結果を考慮します。 したがって、0.1 は、上位 10% の確率質量を構成するトークンのみが考慮されることを意味します。 一般に、これと temperature の両方ではなく、いずれかを変更することをお勧めします。 No
使用 Usage No

SupervisedMethod

監視対象の微調整に使用されるメソッド

Name タイプ Description Required Default
FineTuningMethodType Yes

SupervisedMethodRequest

監視対象の微調整に使用されるメソッド

Name タイプ Description Required Default
FineTuningMethodType Yes

SystemComputeDatasource

SystemComputeDatasource の識別因子

このコンポーネントでは、プロパティ kind を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
kind SystemComputeDatasourceType データソースの型。 Yes

SystemComputeDatasourceType

データソースの型。

Property Value
Description データソースの型。
Type 文字列
Values Storage
Urls

SystemComputeStorage

Storage account

Name タイプ Description Required Default
chunking ChunkingSettings Chunking settings No
接続 BaseConnection リソースへの接続。 No
containerName 文字列 container name No
embeddings アレイ No
kind SystemComputeDatasourceType データソースの型。 Yes

SystemComputeUrl

Urls

Name タイプ Description Required Default
chunking ChunkingSettings Chunking settings No
接続 BaseConnection リソースへの接続。 No
containerName 文字列 container name No
crawling CrawlingSettings Crawling settings No
embeddings アレイ No
kind SystemComputeDatasourceType データソースの型。 Yes
urls アレイ No

TargetIndex

作成するインデックスに関する情報。

TargetIndex の識別子

このコンポーネントでは、プロパティ kind を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
kind TargetType 対象の型。 Yes

TargetType

対象の型。

Property Value
Description 対象の型。
Type 文字列
Values AzureAISearch
CosmosDB
Pinecone

Tool

Name タイプ Description Required Default
関数 FunctionDefinition No
文字列 ツール呼び出しの種類。 これは、この種類のツール呼び出しに対して常に function になります。 No

ToolCall

Name タイプ Description Required Default
関数 FunctionCallFunction No
id 文字列 ツール呼び出しの ID。 No
文字列 出力が必要なツール呼び出しの種類。 現時点では、これは常に function です。 No

TypeDiscriminator

オブジェクトの型を定義します。

Property Value
Description オブジェクトの型を定義します。
Type 文字列
Values list
fine_tuning.job
file
fine_tuning.job.event
fine_tuning.job.checkpoint
model
upload
upload.part
batch
wandb

UpdateStoredCompletionRequest

Name タイプ Description Required Default
メタデータ オブジェクト 追加情報のための任意のキーと値のペア。 No

UploadFileCompleteBody

Name タイプ Description Required Default
md5 文字列 No
part_ids アレイ Yes

UploadFileStartBody

Name タイプ Description Required Default
bytes 整数 Yes
filename 文字列 Yes
mime_type 文字列 Yes
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 Yes

UploadPartResource

Name タイプ Description Required Default
azure_block_id 文字列 No
created_at 整数 このジョブまたは項目が作成されたときのタイムスタンプ (UNIX エポック)。 No
id 文字列 この項目の ID。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
upload_id 文字列 No

UploadResource

Name タイプ Description Required Default
bytes 整数 No
created_at 整数 このジョブまたは項目が作成されたときのタイムスタンプ (UNIX エポック)。 No
expires_at 整数 No
ファイル File ファイルとは、トレーニングと検証に使用できるドキュメントです。 また、結果の詳細を含むサービス生成ドキュメントにすることもできます。 No
filename 文字列 No
id 文字列 この項目の ID。 No
オブジェクト TypeDiscriminator オブジェクトの型を定義します。 No
purpose Purpose アップロードするドキュメントの目的。 微調整には "微調整" を使用します。 これにより、アップロードされたファイルの形式を検証できます。 No
状態 UploadStatus No

UploadStatus

Property Value
Type 文字列
Values pending
expired
completed
cancelled

Usage

Name タイプ Description Required Default
completion_tokens 整数 生成された入力候補内のトークンの数。 No
prompt_tokens 整数 プロンプト内のトークンの数。 No
total_tokens 整数 要求内で使われたトークンの合計数 (プロンプトと入力候補の和)。 No

UserComputeDataset

Storage account

Name タイプ Description Required Default
chunking ChunkingSettings Chunking settings No
datasetId 文字列 No
datasetType 文字列 No
embeddings アレイ No
kind UserComputeDatasourceType データソースの型。 Yes

UserComputeDatasource

UserComputeDatasource の識別因子

このコンポーネントでは、プロパティ kind を使用して、さまざまな型を区別します。

Type Value Schema
Name タイプ Description Required Default
kind UserComputeDatasourceType データソースの型。 Yes

UserComputeDatasourceType

データソースの型。

Property Value
Description データソースの型。
Type 文字列
Values Dataset
Urls

UserComputeUrl

Urls

Name タイプ Description Required Default
chunking ChunkingSettings Chunking settings No
crawling CrawlingSettings Crawling settings No
embeddings アレイ No
kind UserComputeDatasourceType データソースの型。 Yes
urls アレイ No

WandBIntegrationRequest

Name タイプ Description Required Default
エンティティ 文字列 No
name 文字列 No
プロジェクト 文字列 Yes
tags アレイ No

WandBIntegrationRequestWrapper

Name タイプ Description Required Default
IntegrationTypes 使用可能な微調整統合の一覧。 Yes
wandb WandBIntegrationRequest Yes

WorkspaceConnection

AML ワークスペース接続。

Name タイプ Description Required Default
connectionId 文字列 ConnectionId No
kind ConnectionType 接続の種類。 Yes

WorkspaceConnectionEmbeddingSettings

埋め込みモデルへの接続 ID

Name タイプ Description Required Default
connectionId 文字列 No
deploymentName 文字列 No
modelName 文字列 No

Next steps

モデルと REST API を使用した微調整について説明します。 OpenAI をAzureする基になるモデルの詳細について説明します。