Merge pull request #114 from Azure-Samples/keyupdate

Readme improvements
Azure-Samples · May 19, 2024 · 3baa19b · 3baa19b
2 parents 5cd6cfb + d0e7fba
commit 3baa19b
Show file tree

Hide file tree

Showing 15 changed files with 354 additions and 222 deletions.
diff --git a/README.md b/README.md
diff --git a/docs/README.md b/docs/README.md
diff --git a/docs/deploy_existing.md b/docs/deploy_existing.md
@@ -0,0 +1,24 @@
+
+# Deploying with existing Azure resources
+
+If you already have existing Azure resources, or if you want to specify the exact name of new Azure Resource, you can do so by setting `azd` environment values.
+You should set these values before running `azd up`. Once you've set them, return to the [deployment steps](../README.md#deployment).
+
+* [Resource group](#resource-group)
+* [Azure OpenAI resource](#azure-openai-resource)
+
+## Resource group
+
+1. Run `azd env set AZURE_RESOURCE_GROUP {Name of existing resource group}`
+1. Run `azd env set AZURE_LOCATION {Location of existing resource group}`
+
+## Azure OpenAI resource
+
+If you already have an OpenAI resource and would like to re-use it, run `azd env set` to specify the values for the existing OpenAI resource.
+
+```shell
+azd env set AZURE_OPENAI_RESOURCE {name of OpenAI resource}
+azd env set AZURE_OPENAI_RESOURCE_GROUP {name of resource group that it's inside}
+azd env set AZURE_OPENAI_RESOURCE_GROUP_LOCATION {location for that group}
+azd env set AZURE_OPENAI_SKU_NAME {name of the SKU, defaults to "S0"}
+```
diff --git a/docs/local_docker.md b/docs/local_docker.md
@@ -0,0 +1,19 @@
+# Local development server with Docker
+
+In addition to the `Dockerfile` that's used in production, this repo includes a `docker-compose.yaml` for
+local development which creates a volume for the app code. That allows you to make changes to the code
+and see them instantly.
+
+1. Install [Docker Desktop](https://www.docker.com/products/docker-desktop/). If you opened this inside Github Codespaces or a Dev Container in VS Code, installation is not needed. ⚠️ If you're on an Apple M1/M2, you won't be able to run `docker` commands inside a Dev Container; either use Codespaces or do not open the Dev Container.
+
+2. Make sure that the `.env` file exists. The `azd up` deployment step should have created it.
+
+3. Store a key for the OpenAI resource in the `.env` file. You can get the key from the Azure Portal, or from the output of `./infra/getkey.sh`. The key should be stored in the `.env` file as `AZURE_OPENAI_KEY`. This is necessary because Docker containers don't have access to your user Azure credentials.
+
+4. Start the services with this command:
+
+    ```shell
+    docker-compose up --build
+    ```
+
+5. Click 'http://0.0.0.0:50505' in the terminal, which should open a new tab in the browser. You may need to navigate to 'http://localhost:50505' if that URL doesn't work.
diff --git a/docs/local_ollama.md b/docs/local_ollama.md
@@ -0,0 +1,19 @@
+# Using a local LLM server
+
+You may want to save costs by developing against a local LLM server, such as
+[llamafile](https://github.com/Mozilla-Ocho/llamafile/). Note that a local LLM
+will generally be slower and not as sophisticated.
+
+Once you've got your local LLM running and serving an OpenAI-compatible endpoint, define `LOCAL_OPENAI_ENDPOINT` in your `.env` file.
+
+For example, to point at a local llamafile server running on its default port:
+
+```shell
+LOCAL_OPENAI_ENDPOINT="http://localhost:8080/v1"
+```
+
+If you're running inside a dev container, use this local URL instead:
+
+```shell
+LOCAL_OPENAI_ENDPOINT="http://host.docker.internal:8080/v1"
+```
diff --git a/docs/screenshot_chatapp.png b/docs/screenshot_chatapp.png
diff --git a/infra/aca.bicep b/infra/aca.bicep
@@ -45,7 +45,8 @@ module app 'core/host/container-app-upsert.bicep' = {
         value: 'true'
       }
       {
-        name: 'AZURE_OPENAI_CLIENT_ID'
+        // DefaultAzureCredential will look for an environment variable with this name:
+        name: 'AZURE_CLIENT_ID'
         value: acaIdentity.properties.clientId
       }
     ]

diff --git a/infra/core/ai/cognitiveservices.bicep b/infra/core/ai/cognitiveservices.bicep
@@ -1,23 +1,41 @@
+metadata description = 'Creates an Azure Cognitive Services instance.'
 param name string
 param location string = resourceGroup().location
 param tags object = {}
-
+@description('The custom subdomain name used to access the API. Defaults to the value of the name parameter.')
 param customSubDomainName string = name
+param disableLocalAuth bool = false
 param deployments array = []
 param kind string = 'OpenAI'
+
+@allowed([ 'Enabled', 'Disabled' ])
 param publicNetworkAccess string = 'Enabled'
 param sku object = {
   name: 'S0'
 }
+@allowed([ 'None', 'AzureServices' ])
+param bypass string = 'None'
+
+var networkAcls = {
+  defaultAction: 'Allow'
+}
+
+var networkAclsWithBypass = {
+  defaultAction: 'Allow'
+  bypass: bypass
+}
 
-resource account 'Microsoft.CognitiveServices/accounts@2023-05-01' = {
+resource account 'Microsoft.CognitiveServices/accounts@2023-10-01-preview' = {
   name: name
   location: location
   tags: tags
   kind: kind
   properties: {
     customSubDomainName: customSubDomainName
     publicNetworkAccess: publicNetworkAccess
+    // Document Intelligence (FormRecognizer) does not support bypass in network acls
+    networkAcls: kind == 'FormRecognizer' ? networkAcls : networkAclsWithBypass
+    disableLocalAuth: disableLocalAuth
   }
   sku: sku
 }
@@ -40,4 +58,3 @@ output endpoint string = account.properties.endpoint
 output id string = account.id
 output name string = account.name
 output skuName string = account.sku.name
-output key string = account.listKeys().key1
diff --git a/infra/main.bicep b/infra/main.bicep
@@ -23,6 +23,7 @@ param openAiResourceGroupLocation string = ''
 param openAiSkuName string = ''
 param openAiDeploymentCapacity int = 30
 param openAiApiVersion string = ''
+param disableKeyBasedAuth bool = true
 
 var resourceToken = toLower(uniqueString(subscription().id, name, location))
 var tags = { 'azd-env-name': name }
@@ -47,6 +48,7 @@ module openAi 'core/ai/cognitiveservices.bicep' = {
     name: !empty(openAiResourceName) ? openAiResourceName : '${resourceToken}-cog'
     location: !empty(openAiResourceGroupLocation) ? openAiResourceGroupLocation : location
     tags: tags
+    disableLocalAuth: disableKeyBasedAuth
     sku: {
       name: !empty(openAiSkuName) ? openAiSkuName : 'S0'
     }
@@ -60,7 +62,7 @@ module openAi 'core/ai/cognitiveservices.bicep' = {
         }
         sku: {
           name: 'Standard'
-          capacity: 30
+          capacity: openAiDeploymentCapacity
         }
       }
     ]

diff --git a/infra/main.parameters.json b/infra/main.parameters.json
@@ -31,6 +31,9 @@
       },
       "acaExists": {
         "value": "${SERVICE_ACA_RESOURCE_EXISTS=false}"
+      },
+      "disableKeyBasedAuth": {
+        "value": "${DISABLE_KEY_BASED_AUTH=true}"
       }
     }
   }
diff --git a/src/quartapp/__init__.py b/src/quartapp/__init__.py
@@ -6,9 +6,9 @@
 
 def create_app():
     if os.getenv("RUNNING_IN_PRODUCTION"):
-        logging.basicConfig(level=logging.INFO)
+        logging.basicConfig(level=logging.WARNING)
     else:
-        logging.basicConfig(level=logging.DEBUG)
+        logging.basicConfig(level=logging.INFO)
 
     app = Quart(__name__)
 

diff --git a/src/quartapp/chat.py b/src/quartapp/chat.py
@@ -36,22 +36,13 @@ async def configure_openai():
             current_app.logger.info("Using Azure OpenAI with key")
             client_args["api_key"] = os.getenv("AZURE_OPENAI_KEY")
         else:
-            if client_id := os.getenv("AZURE_OPENAI_CLIENT_ID"):
-                # Authenticate using a user-assigned managed identity on Azure
-                # See aca.bicep for value of AZURE_OPENAI_CLIENT_ID
-                current_app.logger.info(
-                    "Using Azure OpenAI with managed identity for client ID %s",
-                    client_id,
-                )
-                default_credential = azure.identity.aio.ManagedIdentityCredential(client_id=client_id)
-            else:
-                # Authenticate using the default Azure credential chain
-                # See https://docs.microsoft.com/azure/developer/python/azure-sdk-authenticate#defaultazurecredential
-                # This will *not* work inside a Docker container.
-                current_app.logger.info("Using Azure OpenAI with default credential")
-                default_credential = azure.identity.aio.DefaultAzureCredential(
-                    exclude_shared_token_cache_credential=True
-                )
+            # Authenticate using the default Azure credential chain
+            # See https://docs.microsoft.com/azure/developer/python/azure-sdk-authenticate#defaultazurecredential
+            # This will *not* work inside a local Docker container.
+            # If using managed user-assigned identity, make sure that AZURE_CLIENT_ID is set
+            # to the client ID of the user-assigned identity.
+            current_app.logger.info("Using Azure OpenAI with default credential")
+            default_credential = azure.identity.aio.DefaultAzureCredential(exclude_shared_token_cache_credential=True)
             client_args["azure_ad_token_provider"] = azure.identity.aio.get_bearer_token_provider(
                 default_credential, "https://cognitiveservices.azure.com/.default"
             )
@@ -91,7 +82,6 @@ async def response_stream():
         )
         try:
             async for event in await chat_coroutine:
-                current_app.logger.info(event)
                 yield json.dumps(event.model_dump(), ensure_ascii=False) + "\n"
         except Exception as e:
             current_app.logger.error(e)

diff --git a/src/requirements.txt b/src/requirements.txt
@@ -6,7 +6,7 @@
 #
 aiofiles==23.2.1
     # via quart
-aiohttp==3.9.3
+aiohttp==3.9.5
     # via quartapp (pyproject.toml)
 aiosignal==1.3.1
     # via aiohttp
@@ -21,9 +21,9 @@ attrs==23.2.0
     # via aiohttp
 azure-core==1.30.1
     # via azure-identity
-azure-identity==1.15.0
+azure-identity==1.16.0
     # via quartapp (pyproject.toml)
-blinker==1.7.0
+blinker==1.8.2
     # via
     #   flask
     #   quart
@@ -41,7 +41,7 @@ click==8.1.7
     #   flask
     #   quart
     #   uvicorn
-cryptography==42.0.5
+cryptography==42.0.7
     # via
     #   azure-identity
     #   msal
@@ -54,7 +54,7 @@ frozenlist==1.4.1
     # via
     #   aiohttp
     #   aiosignal
-gunicorn==21.2.0
+gunicorn==22.0.0
     # via quartapp (pyproject.toml)
 h11==0.14.0
     # via
@@ -82,11 +82,11 @@ idna==3.7
     #   httpx
     #   requests
     #   yarl
-itsdangerous==2.1.2
+itsdangerous==2.2.0
     # via
     #   flask
     #   quart
-jinja2==3.1.3
+jinja2==3.1.4
     # via
     #   flask
     #   quart
@@ -105,7 +105,7 @@ multidict==6.0.5
     # via
     #   aiohttp
     #   yarl
-openai==1.16.2
+openai==1.30.1
     # via quartapp (pyproject.toml)
 packaging==24.0
     # via
@@ -117,12 +117,14 @@ priority==2.0.0
     # via hypercorn
 pycparser==2.22
     # via cffi
-pydantic==2.6.4
+pydantic==2.7.1
     # via openai
-pydantic-core==2.16.3
+pydantic-core==2.18.2
     # via pydantic
 pyjwt[crypto]==2.8.0
-    # via msal
+    # via
+    #   msal
+    #   pyjwt
 python-dotenv==1.0.1
     # via
     #   quartapp (pyproject.toml)
@@ -144,7 +146,7 @@ sniffio==1.3.1
     #   anyio
     #   httpx
     #   openai
-tqdm==4.66.2
+tqdm==4.66.4
     # via openai
 typing-extensions==4.11.0
     # via
@@ -162,7 +164,7 @@ watchfiles==0.21.0
     # via uvicorn
 websockets==12.0
     # via uvicorn
-werkzeug==3.0.2
+werkzeug==3.0.3
     # via
     #   flask
     #   quart

diff --git a/tests/snapshots/test_app/test_chat_stream_text/result.json b/tests/snapshots/test_app/test_chat_stream_text/result.json
@@ -1,9 +1,9 @@
-{"id": "", "choices": [], "created": 0, "model": "", "object": "chat.completion.chunk", "system_fingerprint": null, "prompt_filter_results": [{"prompt_index": 0, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}]}
-{"id": "test-123", "choices": [{"delta": {"content": null, "function_call": null, "role": "assistant", "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
-{"id": "test-123", "choices": [{"delta": {"content": "The", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
-{"id": "test-123", "choices": [{"delta": {"content": " capital", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
-{"id": "test-123", "choices": [{"delta": {"content": " of", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
-{"id": "test-123", "choices": [{"delta": {"content": " France", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
-{"id": "test-123", "choices": [{"delta": {"content": " is", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
-{"id": "test-123", "choices": [{"delta": {"content": " Paris.", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
-{"id": "test-123", "choices": [{"delta": {"content": null, "function_call": null, "role": null, "tool_calls": null}, "finish_reason": "stop", "index": 0, "logprobs": null, "content_filter_results": {}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null}
+{"id": "", "choices": [], "created": 0, "model": "", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null, "prompt_filter_results": [{"prompt_index": 0, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}]}
+{"id": "test-123", "choices": [{"delta": {"content": null, "function_call": null, "role": "assistant", "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}
+{"id": "test-123", "choices": [{"delta": {"content": "The", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}
+{"id": "test-123", "choices": [{"delta": {"content": " capital", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}
+{"id": "test-123", "choices": [{"delta": {"content": " of", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}
+{"id": "test-123", "choices": [{"delta": {"content": " France", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}
+{"id": "test-123", "choices": [{"delta": {"content": " is", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}
+{"id": "test-123", "choices": [{"delta": {"content": " Paris.", "function_call": null, "role": null, "tool_calls": null}, "finish_reason": null, "index": 0, "logprobs": null, "content_filter_results": {"hate": {"filtered": false, "severity": "safe"}, "self_harm": {"filtered": false, "severity": "safe"}, "sexual": {"filtered": false, "severity": "safe"}, "violence": {"filtered": false, "severity": "safe"}}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}
+{"id": "test-123", "choices": [{"delta": {"content": null, "function_call": null, "role": null, "tool_calls": null}, "finish_reason": "stop", "index": 0, "logprobs": null, "content_filter_results": {}}], "created": 1703462735, "model": "gpt-35-turbo", "object": "chat.completion.chunk", "system_fingerprint": null, "usage": null}