Refactors to scripts

Author: pamelafox

Diff for: app/backend/app.py

@@ -395,7 +395,7 @@ async def setup_clients():
     AZURE_SEARCH_INDEX = os.environ["AZURE_SEARCH_INDEX"]
     # Shared by all OpenAI deployments
     OPENAI_HOST = os.getenv("OPENAI_HOST", "azure")
-    OPENAI_CHATGPT_MODEL = os.environ["AZURE_OPENAI_CHATGPT_MODEL"]
+    OPENAI_CHATGPT_MODEL = "phi3.5:latest"
     OPENAI_EMB_MODEL = os.getenv("AZURE_OPENAI_EMB_MODEL_NAME", "text-embedding-ada-002")
     OPENAI_EMB_DIMENSIONS = int(os.getenv("AZURE_OPENAI_EMB_DIMENSIONS", 1536))
     # Used with Azure OpenAI deployments
@@ -448,8 +448,16 @@ async def setup_clients():
     # The managed identity is setup in the infra/ folder.
     azure_credential: Union[AzureDeveloperCliCredential, ManagedIdentityCredential]
     if RUNNING_ON_AZURE:
-        current_app.logger.info("Setting up Azure credential using ManagedIdentityCredential")
-        azure_credential = ManagedIdentityCredential()
+        if AZURE_CLIENT_ID := os.getenv("AZURE_CLIENT_ID"):
+            # ManagedIdentityCredential should use AZURE_CLIENT_ID if set in env, but its not working for some reason,
+            # so we explicitly pass it in as the client ID here. This is necessary for user-assigned managed identities.
+            current_app.logger.info(
+                "Setting up Azure credential using ManagedIdentityCredential with client_id %s", AZURE_CLIENT_ID
+            )
+            azure_credential = ManagedIdentityCredential(client_id=AZURE_CLIENT_ID)
+        else:
+            current_app.logger.info("Setting up Azure credential using ManagedIdentityCredential")
+            azure_credential = ManagedIdentityCredential()
     elif AZURE_TENANT_ID:
         current_app.logger.info(
             "Setting up Azure credential using AzureDeveloperCliCredential with tenant_id %s", AZURE_TENANT_ID
@@ -705,9 +713,10 @@ def create_app():
     # Log levels should be one of https://docs.python.org/3/library/logging.html#logging-levels
     # Set root level to WARNING to avoid seeing overly verbose logs from SDKS
     logging.basicConfig(level=logging.WARNING)
-    # Set the app logger level to INFO by default
-    default_level = "INFO"
-    app.logger.setLevel(os.getenv("APP_LOG_LEVEL", default_level))
+    # Set our own logger levels to INFO by default
+    app_level = os.getenv("APP_LOG_LEVEL", "INFO")
+    app.logger.setLevel(os.getenv("APP_LOG_LEVEL", app_level))
+    logging.getLogger("ragapp").setLevel(app_level)
 
     if allowed_origin := os.getenv("ALLOWED_ORIGIN"):
         app.logger.info("ALLOWED_ORIGIN is set, enabling CORS for %s", allowed_origin)

Diff for: app/backend/approaches/chatreadretrieveread.py

@@ -51,7 +51,7 @@ def __init__(
         self.content_field = content_field
         self.query_language = query_language
         self.query_speller = query_speller
-        self.chatgpt_token_limit = get_token_limit(chatgpt_model)
+        self.chatgpt_token_limit = get_token_limit(chatgpt_model, default_to_minimum=True)
 
     @property
     def system_message_chat_conversation(self):

Diff for: app/backend/approaches/retrievethenread.py

@@ -66,7 +66,7 @@ def __init__(
         self.content_field = content_field
         self.query_language = query_language
         self.query_speller = query_speller
-        self.chatgpt_token_limit = get_token_limit(chatgpt_model)
+        self.chatgpt_token_limit = get_token_limit(chatgpt_model, default_to_minimum=True)
 
     async def run(
         self,
@@ -121,6 +121,7 @@ async def run(
             few_shots=[{"role": "user", "content": self.question}, {"role": "assistant", "content": self.answer}],
             new_user_content=user_content,
             max_tokens=self.chatgpt_token_limit - response_token_limit,
+            fallback_to_default=True,
         )
 
         chat_completion = await self.openai_client.chat.completions.create(

Diff for: app/backend/load_azd_env.py

@@ -0,0 +1,23 @@
+import json
+import logging
+import subprocess
+
+from dotenv import load_dotenv
+
+logger = logging.getLogger("ragapp")
+
+
+def load_azd_env():
+    """Get path to current azd env file and load file using python-dotenv"""
+    result = subprocess.run("azd env list -o json", shell=True, capture_output=True, text=True)
+    if result.returncode != 0:
+        raise Exception("Error loading azd env")
+    env_json = json.loads(result.stdout)
+    env_file_path = None
+    for entry in env_json:
+        if entry["IsDefault"]:
+            env_file_path = entry["DotEnvPath"]
+    if not env_file_path:
+        raise Exception("No default azd env file found")
+    logger.info(f"Loading azd env from {env_file_path}")
+    load_dotenv(env_file_path, override=True)

Diff for: app/backend/main.py

@@ -1,3 +1,12 @@
+import os
+
 from app import create_app
+from load_azd_env import load_azd_env
+
+# WEBSITE_HOSTNAME is always set by App Service, RUNNING_IN_PRODUCTION is set in main.bicep
+RUNNING_ON_AZURE = os.getenv("WEBSITE_HOSTNAME") is not None or os.getenv("RUNNING_IN_PRODUCTION") is not None
+
+if not RUNNING_ON_AZURE:
+    load_azd_env()
 
 app = create_app()