refactor(agent/config): Modularize Config and revive Azure support (#6497)

* feat: Refactor config loading and initialization to be modular and decentralized - Refactored the `ConfigBuilder` class to support modular loading and initialization of the configuration from environment variables. - Implemented recursive loading and initialization of nested config objects. - Introduced the `SystemConfiguration` base class to provide common functionality for all system settings. - Added the `from_env` attribute to the `UserConfigurable` decorator to provide environment variable mappings. - Updated the `Config` class and its related classes to inherit from `SystemConfiguration` and use the `UserConfigurable` decorator. - Updated `LoggingConfig` and `TTSConfig` to use the `UserConfigurable` decorator for their fields. - Modified the implementation of the `build_config_from_env` method in `ConfigBuilder` to utilize the new modular and recursive loading and initialization logic. - Updated applicable test cases to reflect the changes in the config loading and initialization logic. This refactor improves the flexibility and maintainability of the configuration loading process by introducing modular and recursive behavior, allowing for easier extension and customization through environment variables. * refactor: Move OpenAI credentials into `OpenAICredentials` sub-config - Move OpenAI API key and other OpenAI credentials from the global config to a new sub-config called OpenAICredentials. - Update the necessary code to use the new OpenAICredentials sub-config instead of the global config when accessing OpenAI credentials. - (Hopefully) unbreak Azure support. - Update azure.yaml.template. - Enable validation of assignment operations on SystemConfiguration and SystemSettings objects. * feat: Update AutoGPT configuration options and setup instructions - Added new configuration options for logging and OpenAI usage to .env.template - Removed deprecated configuration options in config/config.py - Updated setup instructions in Docker and general setup documentation to include information on using Azure's OpenAI services * fix: Fix image generation with Dall-E - Fix issue with image generation with Dall-E API Additional user context: This commit fixes an issue with image generation using the Dall-E API. The code now correctly retrieves the API key from the agent's legacy configuration. * refactor(agent/core): Refactor `autogpt.core.configuration.schema` and update docstrings - Refactor the `schema.py` file in the `autogpt.core.configuration` module. - Added docstring to `SystemConfiguration.from_env()` - Updated docstrings for functions `_get_user_config_values`, `_get_non_default_user_config_values`, `_recursive_init_model`, `_recurse_user_config_fields`, and `_recurse_user_config_values`.
2025-12-19 06:54:22 +01:00 · 2023-12-05 16:28:23 +01:00
parent 03eb921ca6
commit 7b05245286
17 changed files with 666 additions and 401 deletions
--- a/autogpts/autogpt/.env.template
+++ b/autogpts/autogpt/.env.template
@@ -1,5 +1,3 @@
-# For further descriptions of these settings see docs/configuration/options.md or go to docs.agpt.co
-
 ################################################################################
 ### AutoGPT - GENERAL SETTINGS
 ################################################################################
@@ -25,14 +23,6 @@ OPENAI_API_KEY=your-openai-api-key
 ## PROMPT_SETTINGS_FILE - Specifies which Prompt Settings file to use, relative to the AutoGPT root directory. (defaults to prompt_settings.yaml)
 # PROMPT_SETTINGS_FILE=prompt_settings.yaml

-## OPENAI_API_BASE_URL - Custom url for the OpenAI API, useful for connecting to custom backends. No effect if USE_AZURE is true, leave blank to keep the default url
-# the following is an example:
-# OPENAI_API_BASE_URL=http://localhost:443/v1
-
-## OPENAI_FUNCTIONS - Enables OpenAI functions: https://platform.openai.com/docs/guides/gpt/function-calling
-## WARNING: this feature is only supported by OpenAI's newest models. Until these models become the default on 27 June, add a '-0613' suffix to the model of your choosing.
-# OPENAI_FUNCTIONS=False
-
 ## AUTHORISE COMMAND KEY - Key to authorise commands
 # AUTHORISE_COMMAND_KEY=y

@@ -52,6 +42,17 @@ OPENAI_API_KEY=your-openai-api-key
 ## TEMPERATURE - Sets temperature in OpenAI (Default: 0)
 # TEMPERATURE=0

+## OPENAI_API_BASE_URL - Custom url for the OpenAI API, useful for connecting to custom backends. No effect if USE_AZURE is true, leave blank to keep the default url
+# the following is an example:
+# OPENAI_API_BASE_URL=http://localhost:443/v1
+
+# OPENAI_API_TYPE=
+# OPENAI_API_VERSION=
+
+## OPENAI_FUNCTIONS - Enables OpenAI functions: https://platform.openai.com/docs/guides/gpt/function-calling
+## Note: this feature is only supported by OpenAI's newer models.
+# OPENAI_FUNCTIONS=False
+
 ## OPENAI_ORGANIZATION - Your OpenAI Organization key (Default: None)
 # OPENAI_ORGANIZATION=

@@ -90,32 +91,6 @@ OPENAI_API_KEY=your-openai-api-key
 ## SHELL_ALLOWLIST - List of shell commands that ARE allowed to be executed by AutoGPT (Default: None)
 # SHELL_ALLOWLIST=

-################################################################################
-### MEMORY
-################################################################################
-
-### General
-
-## MEMORY_BACKEND - Memory backend type
-# MEMORY_BACKEND=json_file
-
-## MEMORY_INDEX - Value used in the Memory backend for scoping, naming, or indexing (Default: auto-gpt)
-# MEMORY_INDEX=auto-gpt
-
-### Redis
-
-## REDIS_HOST - Redis host (Default: localhost, use "redis" for docker-compose)
-# REDIS_HOST=localhost
-
-## REDIS_PORT - Redis port (Default: 6379)
-# REDIS_PORT=6379
-
-## REDIS_PASSWORD - Redis password (Default: "")
-# REDIS_PASSWORD=
-
-## WIPE_REDIS_ON_START - Wipes data / index on start (Default: True)
-# WIPE_REDIS_ON_START=True
-
 ################################################################################
 ### IMAGE GENERATION PROVIDER
 ################################################################################
@@ -191,13 +166,12 @@ OPENAI_API_KEY=your-openai-api-key
 ################################################################################

 ## TEXT_TO_SPEECH_PROVIDER - Which Text to Speech provider to use (Default: gtts)
+## Options: gtts, streamelements, elevenlabs, macos
 # TEXT_TO_SPEECH_PROVIDER=gtts

-### Only if TEXT_TO_SPEECH_PROVIDER=streamelements
 ## STREAMELEMENTS_VOICE - Voice to use for StreamElements (Default: Brian)
 # STREAMELEMENTS_VOICE=Brian

-### Only if TEXT_TO_SPEECH_PROVIDER=elevenlabs
 ## ELEVENLABS_API_KEY - Eleven Labs API key (Default: None)
 # ELEVENLABS_API_KEY=

@@ -210,3 +184,22 @@ OPENAI_API_KEY=your-openai-api-key

 ## CHAT_MESSAGES_ENABLED - Enable chat messages (Default: False)
 # CHAT_MESSAGES_ENABLED=False
+
+################################################################################
+### LOGGING
+################################################################################
+
+## LOG_LEVEL - Set the minimum level to filter log output by. Setting this to DEBUG implies LOG_FORMAT=debug, unless LOG_FORMAT is set explicitly.
+## Options: DEBUG, INFO, WARNING, ERROR, CRITICAL
+# LOG_LEVEL=INFO
+
+## LOG_FORMAT - The format in which to log messages to the console (and log files).
+## Options: simple, debug, structured_google_cloud
+# LOG_FORMAT=simple
+
+## LOG_FILE_FORMAT - Normally follows the LOG_FORMAT setting, but can be set separately.
+## Note: Log file output is disabled if LOG_FORMAT=structured_google_cloud.
+# LOG_FILE_FORMAT=simple
+
+## PLAIN_OUTPUT - Disables animated typing in the console output.
+# PLAIN_OUTPUT=False
--- a/autogpts/autogpt/autogpt/app/configurator.py
+++ b/autogpts/autogpt/autogpt/app/configurator.py
@@ -3,7 +3,7 @@ from __future__ import annotations

 import logging
 from pathlib import Path
-from typing import Literal, Optional
+from typing import TYPE_CHECKING, Literal, Optional

 import click
 from colorama import Back, Fore, Style
@@ -16,6 +16,9 @@ from autogpt.logs.config import LogFormatName
 from autogpt.logs.helpers import print_attribute, request_user_double_check
 from autogpt.memory.vector import get_supported_memory_backends

+if TYPE_CHECKING:
+    from autogpt.core.resource.model_providers.openai import OpenAICredentials
+
 logger = logging.getLogger(__name__)


@@ -103,7 +106,11 @@ def apply_overrides_to_config(
        config.smart_llm = GPT_3_MODEL
    elif (
        gpt4only
-        and check_model(GPT_4_MODEL, model_type="smart_llm", config=config)
+        and check_model(
+            GPT_4_MODEL,
+            model_type="smart_llm",
+            api_credentials=config.openai_credentials,
+        )
        == GPT_4_MODEL
    ):
        print_attribute("GPT4 Only Mode", "ENABLED")
@@ -111,8 +118,12 @@ def apply_overrides_to_config(
        config.fast_llm = GPT_4_MODEL
        config.smart_llm = GPT_4_MODEL
    else:
-        config.fast_llm = check_model(config.fast_llm, "fast_llm", config=config)
-        config.smart_llm = check_model(config.smart_llm, "smart_llm", config=config)
+        config.fast_llm = check_model(
+            config.fast_llm, "fast_llm", api_credentials=config.openai_credentials
+        )
+        config.smart_llm = check_model(
+            config.smart_llm, "smart_llm", api_credentials=config.openai_credentials
+        )

    if memory_type:
        supported_memory = get_supported_memory_backends()
@@ -187,12 +198,11 @@ def apply_overrides_to_config(
 def check_model(
    model_name: str,
    model_type: Literal["smart_llm", "fast_llm"],
-    config: Config,
+    api_credentials: OpenAICredentials,
 ) -> str:
    """Check if model is available for use. If not, return gpt-3.5-turbo."""
-    openai_credentials = config.get_openai_credentials(model_name)
    api_manager = ApiManager()
-    models = api_manager.get_models(**openai_credentials)
+    models = api_manager.get_models(**api_credentials.get_api_access_kwargs(model_name))

    if any(model_name in m["id"] for m in models):
        return model_name
--- a/autogpts/autogpt/autogpt/app/main.py
+++ b/autogpts/autogpt/autogpt/app/main.py
@@ -14,7 +14,6 @@ from typing import TYPE_CHECKING, Optional

 from colorama import Fore, Style
 from forge.sdk.db import AgentDB
-from pydantic import SecretStr

 if TYPE_CHECKING:
    from autogpt.agents.agent import Agent
@@ -31,7 +30,6 @@ from autogpt.config import (
    ConfigBuilder,
    assert_config_has_openai_api_key,
 )
-from autogpt.core.resource.model_providers import ModelProviderCredentials
 from autogpt.core.resource.model_providers.openai import OpenAIProvider
 from autogpt.core.runner.client_lib.utils import coroutine
 from autogpt.logs.config import configure_chat_plugins, configure_logging
@@ -364,19 +362,11 @@ def _configure_openai_provider(config: Config) -> OpenAIProvider:
    Returns:
        A configured OpenAIProvider object.
    """
-    if config.openai_api_key is None:
+    if config.openai_credentials is None:
        raise RuntimeError("OpenAI key is not configured")

    openai_settings = OpenAIProvider.default_settings.copy(deep=True)
-    openai_settings.credentials = ModelProviderCredentials(
-        api_key=SecretStr(config.openai_api_key),
-        # TODO: support OpenAI Azure credentials
-        api_base=SecretStr(config.openai_api_base) if config.openai_api_base else None,
-        api_type=SecretStr(config.openai_api_type) if config.openai_api_type else None,
-        api_version=SecretStr(config.openai_api_version)
-        if config.openai_api_version
-        else None,
-    )
+    openai_settings.credentials = config.openai_credentials
    return OpenAIProvider(
        settings=openai_settings,
        logger=logging.getLogger("OpenAIProvider"),
--- a/autogpts/autogpt/autogpt/commands/image_gen.py
+++ b/autogpts/autogpt/autogpt/commands/image_gen.py
@@ -147,7 +147,7 @@ def generate_image_with_dalle(
        n=1,
        size=f"{size}x{size}",
        response_format="b64_json",
-        api_key=agent.legacy_config.openai_api_key,
+        api_key=agent.legacy_config.openai_credentials.api_key.get_secret_value(),
    )

    logger.info(f"Image Generated for prompt:{prompt}")
--- a/autogpts/autogpt/autogpt/config/config.py
+++ b/autogpts/autogpt/autogpt/config/config.py
@@ -1,22 +1,26 @@
 """Configuration class to store the state of bools for different scripts access."""
 from __future__ import annotations

-import contextlib
-import logging
 import os
 import re
 from pathlib import Path
-from typing import Any, Dict, Optional, Union
+from typing import Any, Optional, Union

-import yaml
 from auto_gpt_plugin_template import AutoGPTPluginTemplate
 from colorama import Fore
-from pydantic import Field, validator
+from pydantic import Field, SecretStr, validator

 import autogpt
-from autogpt.core.configuration.schema import Configurable, SystemSettings
-from autogpt.core.resource.model_providers.openai import OPEN_AI_CHAT_MODELS
-from autogpt.logs.config import LogFormatName, LoggingConfig
+from autogpt.core.configuration.schema import (
+    Configurable,
+    SystemSettings,
+    UserConfigurable,
+)
+from autogpt.core.resource.model_providers.openai import (
+    OPEN_AI_CHAT_MODELS,
+    OpenAICredentials,
+)
+from autogpt.logs.config import LoggingConfig
 from autogpt.plugins.plugins_config import PluginsConfig
 from autogpt.speech import TTSConfig

@@ -33,6 +37,7 @@ GPT_3_MODEL = "gpt-3.5-turbo"
 class Config(SystemSettings, arbitrary_types_allowed=True):
    name: str = "Auto-GPT configuration"
    description: str = "Default configuration for the Auto-GPT application."
+
    ########################
    # Application Settings #
    ########################
@@ -40,10 +45,12 @@ class Config(SystemSettings, arbitrary_types_allowed=True):
    app_data_dir: Path = project_root / "data"
    skip_news: bool = False
    skip_reprompt: bool = False
-    authorise_key: str = "y"
-    exit_key: str = "n"
+    authorise_key: str = UserConfigurable(default="y", from_env="AUTHORISE_COMMAND_KEY")
+    exit_key: str = UserConfigurable(default="n", from_env="EXIT_KEY")
    noninteractive_mode: bool = False
-    chat_messages_enabled: bool = True
+    chat_messages_enabled: bool = UserConfigurable(
+        default=True, from_env=lambda: os.getenv("CHAT_MESSAGES_ENABLED") == "True"
+    )
    # TTS configuration
    tts_config: TTSConfig = TTSConfig()
    logging: LoggingConfig = LoggingConfig()
@@ -52,15 +59,38 @@ class Config(SystemSettings, arbitrary_types_allowed=True):
    # Agent Control Settings #
    ##########################
    # Paths
-    ai_settings_file: Path = project_root / AI_SETTINGS_FILE
-    prompt_settings_file: Path = project_root / PROMPT_SETTINGS_FILE
+    ai_settings_file: Path = UserConfigurable(
+        default=AI_SETTINGS_FILE,
+        from_env=lambda: Path(f) if (f := os.getenv("AI_SETTINGS_FILE")) else None,
+    )
+    prompt_settings_file: Path = UserConfigurable(
+        default=PROMPT_SETTINGS_FILE,
+        from_env=lambda: Path(f) if (f := os.getenv("PROMPT_SETTINGS_FILE")) else None,
+    )
+
    # Model configuration
-    fast_llm: str = "gpt-3.5-turbo-16k"
-    smart_llm: str = "gpt-4"
-    temperature: float = 0
-    openai_functions: bool = False
-    embedding_model: str = "text-embedding-ada-002"
-    browse_spacy_language_model: str = "en_core_web_sm"
+    fast_llm: str = UserConfigurable(
+        default="gpt-3.5-turbo-16k",
+        from_env=lambda: os.getenv("FAST_LLM"),
+    )
+    smart_llm: str = UserConfigurable(
+        default="gpt-4",
+        from_env=lambda: os.getenv("SMART_LLM"),
+    )
+    temperature: float = UserConfigurable(
+        default=0,
+        from_env=lambda: float(v) if (v := os.getenv("TEMPERATURE")) else None,
+    )
+    openai_functions: bool = UserConfigurable(
+        default=False, from_env=lambda: os.getenv("OPENAI_FUNCTIONS", "False") == "True"
+    )
+    embedding_model: str = UserConfigurable(
+        default="text-embedding-ada-002", from_env="EMBEDDING_MODEL"
+    )
+    browse_spacy_language_model: str = UserConfigurable(
+        default="en_core_web_sm", from_env="BROWSE_SPACY_LANGUAGE_MODEL"
+    )
+
    # Run loop configuration
    continuous_mode: bool = False
    continuous_limit: int = 0
@@ -68,74 +98,138 @@ class Config(SystemSettings, arbitrary_types_allowed=True):
    ##########
    # Memory #
    ##########
-    memory_backend: str = "json_file"
-    memory_index: str = "auto-gpt-memory"
-    redis_host: str = "localhost"
-    redis_port: int = 6379
-    redis_password: str = ""
-    wipe_redis_on_start: bool = True
+    memory_backend: str = UserConfigurable("json_file", from_env="MEMORY_BACKEND")
+    memory_index: str = UserConfigurable("auto-gpt-memory", from_env="MEMORY_INDEX")
+    redis_host: str = UserConfigurable("localhost", from_env="REDIS_HOST")
+    redis_port: int = UserConfigurable(
+        default=6379,
+        from_env=lambda: int(v) if (v := os.getenv("REDIS_PORT")) else None,
+    )
+    redis_password: str = UserConfigurable("", from_env="REDIS_PASSWORD")
+    wipe_redis_on_start: bool = UserConfigurable(
+        default=True,
+        from_env=lambda: os.getenv("WIPE_REDIS_ON_START", "True") == "True",
+    )

    ############
    # Commands #
    ############
    # General
-    disabled_command_categories: list[str] = Field(default_factory=list)
+    disabled_command_categories: list[str] = UserConfigurable(
+        default_factory=list,
+        from_env=lambda: _safe_split(os.getenv("DISABLED_COMMAND_CATEGORIES")),
+    )
+
    # File ops
-    restrict_to_workspace: bool = True
+    restrict_to_workspace: bool = UserConfigurable(
+        default=True,
+        from_env=lambda: os.getenv("RESTRICT_TO_WORKSPACE", "True") == "True",
+    )
    allow_downloads: bool = False
+
    # Shell commands
-    shell_command_control: str = "denylist"
-    execute_local_commands: bool = False
-    shell_denylist: list[str] = Field(default_factory=lambda: ["sudo", "su"])
-    shell_allowlist: list[str] = Field(default_factory=list)
+    shell_command_control: str = UserConfigurable(
+        default="denylist", from_env="SHELL_COMMAND_CONTROL"
+    )
+    execute_local_commands: bool = UserConfigurable(
+        default=False,
+        from_env=lambda: os.getenv("EXECUTE_LOCAL_COMMANDS", "False") == "True",
+    )
+    shell_denylist: list[str] = UserConfigurable(
+        default_factory=lambda: ["sudo", "su"],
+        from_env=lambda: _safe_split(
+            os.getenv("SHELL_DENYLIST", os.getenv("DENY_COMMANDS"))
+        ),
+    )
+    shell_allowlist: list[str] = UserConfigurable(
+        default_factory=list,
+        from_env=lambda: _safe_split(
+            os.getenv("SHELL_ALLOWLIST", os.getenv("ALLOW_COMMANDS"))
+        ),
+    )
+
    # Text to image
-    image_provider: Optional[str] = None
-    huggingface_image_model: str = "CompVis/stable-diffusion-v1-4"
-    sd_webui_url: Optional[str] = "http://localhost:7860"
-    image_size: int = 256
+    image_provider: Optional[str] = UserConfigurable(from_env="IMAGE_PROVIDER")
+    huggingface_image_model: str = UserConfigurable(
+        default="CompVis/stable-diffusion-v1-4", from_env="HUGGINGFACE_IMAGE_MODEL"
+    )
+    sd_webui_url: Optional[str] = UserConfigurable(
+        default="http://localhost:7860", from_env="SD_WEBUI_URL"
+    )
+    image_size: int = UserConfigurable(
+        default=256,
+        from_env=lambda: int(v) if (v := os.getenv("IMAGE_SIZE")) else None,
+    )
+
    # Audio to text
-    audio_to_text_provider: str = "huggingface"
-    huggingface_audio_to_text_model: Optional[str] = None
+    audio_to_text_provider: str = UserConfigurable(
+        default="huggingface", from_env="AUDIO_TO_TEXT_PROVIDER"
+    )
+    huggingface_audio_to_text_model: Optional[str] = UserConfigurable(
+        from_env="HUGGINGFACE_AUDIO_TO_TEXT_MODEL"
+    )
+
    # Web browsing
-    selenium_web_browser: str = "chrome"
-    selenium_headless: bool = True
-    user_agent: str = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36"  # noqa: E501
+    selenium_web_browser: str = UserConfigurable("chrome", from_env="USE_WEB_BROWSER")
+    selenium_headless: bool = UserConfigurable(
+        default=True, from_env=lambda: os.getenv("HEADLESS_BROWSER", "True") == "True"
+    )
+    user_agent: str = UserConfigurable(
+        default="Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36",  # noqa: E501
+        from_env="USER_AGENT",
+    )

    ###################
    # Plugin Settings #
    ###################
-    plugins_dir: str = "plugins"
-    plugins_config_file: Path = project_root / PLUGINS_CONFIG_FILE
+    plugins_dir: str = UserConfigurable("plugins", from_env="PLUGINS_DIR")
+    plugins_config_file: Path = UserConfigurable(
+        default=PLUGINS_CONFIG_FILE,
+        from_env=lambda: Path(f) if (f := os.getenv("PLUGINS_CONFIG_FILE")) else None,
+    )
    plugins_config: PluginsConfig = Field(
        default_factory=lambda: PluginsConfig(plugins={})
    )
    plugins: list[AutoGPTPluginTemplate] = Field(default_factory=list, exclude=True)
-    plugins_allowlist: list[str] = Field(default_factory=list)
-    plugins_denylist: list[str] = Field(default_factory=list)
-    plugins_openai: list[str] = Field(default_factory=list)
+    plugins_allowlist: list[str] = UserConfigurable(
+        default_factory=list,
+        from_env=lambda: _safe_split(os.getenv("ALLOWLISTED_PLUGINS")),
+    )
+    plugins_denylist: list[str] = UserConfigurable(
+        default_factory=list,
+        from_env=lambda: _safe_split(os.getenv("DENYLISTED_PLUGINS")),
+    )
+    plugins_openai: list[str] = UserConfigurable(
+        default_factory=list, from_env=lambda: _safe_split(os.getenv("OPENAI_PLUGINS"))
+    )

    ###############
    # Credentials #
    ###############
    # OpenAI
-    openai_api_key: Optional[str] = None
-    openai_api_type: Optional[str] = None
-    openai_api_base: Optional[str] = None
-    openai_api_version: Optional[str] = None
-    openai_organization: Optional[str] = None
-    use_azure: bool = False
-    azure_config_file: Optional[Path] = project_root / AZURE_CONFIG_FILE
-    azure_model_to_deployment_id_map: Optional[Dict[str, str]] = None
+    openai_credentials: Optional[OpenAICredentials] = None
+    azure_config_file: Optional[Path] = UserConfigurable(
+        default=AZURE_CONFIG_FILE,
+        from_env=lambda: Path(f) if (f := os.getenv("AZURE_CONFIG_FILE")) else None,
+    )
+
    # Github
-    github_api_key: Optional[str] = None
-    github_username: Optional[str] = None
+    github_api_key: Optional[str] = UserConfigurable(from_env="GITHUB_API_KEY")
+    github_username: Optional[str] = UserConfigurable(from_env="GITHUB_USERNAME")
+
    # Google
-    google_api_key: Optional[str] = None
-    google_custom_search_engine_id: Optional[str] = None
+    google_api_key: Optional[str] = UserConfigurable(from_env="GOOGLE_API_KEY")
+    google_custom_search_engine_id: Optional[str] = UserConfigurable(
+        from_env=lambda: os.getenv("GOOGLE_CUSTOM_SEARCH_ENGINE_ID"),
+    )
+
    # Huggingface
-    huggingface_api_token: Optional[str] = None
+    huggingface_api_token: Optional[str] = UserConfigurable(
+        from_env="HUGGINGFACE_API_TOKEN"
+    )
+
    # Stable Diffusion
-    sd_webui_auth: Optional[str] = None
+    sd_webui_auth: Optional[str] = UserConfigurable(from_env="SD_WEBUI_AUTH")

    @validator("plugins", each_item=True)
    def validate_plugins(cls, p: AutoGPTPluginTemplate | Any):
@@ -157,67 +251,6 @@ class Config(SystemSettings, arbitrary_types_allowed=True):
            )
        return v

-    def get_openai_credentials(self, model: str) -> dict[str, str]:
-        credentials = {
-            "api_key": self.openai_api_key,
-            "api_base": self.openai_api_base,
-            "organization": self.openai_organization,
-        }
-        if self.use_azure:
-            azure_credentials = self.get_azure_credentials(model)
-            credentials.update(azure_credentials)
-        return credentials
-
-    def get_azure_credentials(self, model: str) -> dict[str, str]:
-        """Get the kwargs for the Azure API."""
-
-        # Fix --gpt3only and --gpt4only in combination with Azure
-        fast_llm = (
-            self.fast_llm
-            if not (
-                self.fast_llm == self.smart_llm
-                and self.fast_llm.startswith(GPT_4_MODEL)
-            )
-            else f"not_{self.fast_llm}"
-        )
-        smart_llm = (
-            self.smart_llm
-            if not (
-                self.smart_llm == self.fast_llm
-                and self.smart_llm.startswith(GPT_3_MODEL)
-            )
-            else f"not_{self.smart_llm}"
-        )
-
-        deployment_id = {
-            fast_llm: self.azure_model_to_deployment_id_map.get(
-                "fast_llm_deployment_id",
-                self.azure_model_to_deployment_id_map.get(
-                    "fast_llm_model_deployment_id"  # backwards compatibility
-                ),
-            ),
-            smart_llm: self.azure_model_to_deployment_id_map.get(
-                "smart_llm_deployment_id",
-                self.azure_model_to_deployment_id_map.get(
-                    "smart_llm_model_deployment_id"  # backwards compatibility
-                ),
-            ),
-            self.embedding_model: self.azure_model_to_deployment_id_map.get(
-                "embedding_model_deployment_id"
-            ),
-        }.get(model, None)
-
-        kwargs = {
-            "api_type": self.openai_api_type,
-            "api_base": self.openai_api_base,
-            "api_version": self.openai_api_version,
-        }
-        if model == self.embedding_model:
-            kwargs["engine"] = deployment_id
-        else:
-            kwargs["deployment_id"] = deployment_id
-        return kwargs
-

 class ConfigBuilder(Configurable[Config]):
    default_settings = Config()
@@ -225,131 +258,25 @@ class ConfigBuilder(Configurable[Config]):
    @classmethod
    def build_config_from_env(cls, project_root: Path = PROJECT_ROOT) -> Config:
        """Initialize the Config class"""
-        config_dict = {
-            "project_root": project_root,
-            "logging": {
-                "level": logging.getLevelName(os.getenv("LOG_LEVEL", "INFO")),
-                "log_format": LogFormatName(os.getenv("LOG_FORMAT", "simple")),
-                "log_file_format": LogFormatName(
-                    os.getenv("LOG_FILE_FORMAT", os.getenv("LOG_FORMAT", "simple"))
-                ),
-                "plain_console_output": os.getenv("PLAIN_OUTPUT", "False") == "True",
-            },
-            "authorise_key": os.getenv("AUTHORISE_COMMAND_KEY"),
-            "exit_key": os.getenv("EXIT_KEY"),
-            "shell_command_control": os.getenv("SHELL_COMMAND_CONTROL"),
-            "ai_settings_file": project_root
-            / Path(os.getenv("AI_SETTINGS_FILE", AI_SETTINGS_FILE)),
-            "prompt_settings_file": project_root
-            / Path(os.getenv("PROMPT_SETTINGS_FILE", PROMPT_SETTINGS_FILE)),
-            "fast_llm": os.getenv("FAST_LLM", os.getenv("FAST_LLM_MODEL")),
-            "smart_llm": os.getenv("SMART_LLM", os.getenv("SMART_LLM_MODEL")),
-            "embedding_model": os.getenv("EMBEDDING_MODEL"),
-            "browse_spacy_language_model": os.getenv("BROWSE_SPACY_LANGUAGE_MODEL"),
-            "openai_api_key": os.getenv("OPENAI_API_KEY"),
-            "use_azure": os.getenv("USE_AZURE") == "True",
-            "azure_config_file": project_root
-            / Path(os.getenv("AZURE_CONFIG_FILE", AZURE_CONFIG_FILE)),
-            "execute_local_commands": os.getenv("EXECUTE_LOCAL_COMMANDS", "False")
-            == "True",
-            "restrict_to_workspace": os.getenv("RESTRICT_TO_WORKSPACE", "True")
-            == "True",
-            "openai_functions": os.getenv("OPENAI_FUNCTIONS", "False") == "True",
-            "tts_config": {
-                "provider": os.getenv("TEXT_TO_SPEECH_PROVIDER"),
-            },
-            "github_api_key": os.getenv("GITHUB_API_KEY"),
-            "github_username": os.getenv("GITHUB_USERNAME"),
-            "google_api_key": os.getenv("GOOGLE_API_KEY"),
-            "image_provider": os.getenv("IMAGE_PROVIDER"),
-            "huggingface_api_token": os.getenv("HUGGINGFACE_API_TOKEN"),
-            "huggingface_image_model": os.getenv("HUGGINGFACE_IMAGE_MODEL"),
-            "audio_to_text_provider": os.getenv("AUDIO_TO_TEXT_PROVIDER"),
-            "huggingface_audio_to_text_model": os.getenv(
-                "HUGGINGFACE_AUDIO_TO_TEXT_MODEL"
-            ),
-            "sd_webui_url": os.getenv("SD_WEBUI_URL"),
-            "sd_webui_auth": os.getenv("SD_WEBUI_AUTH"),
-            "selenium_web_browser": os.getenv("USE_WEB_BROWSER"),
-            "selenium_headless": os.getenv("HEADLESS_BROWSER", "True") == "True",
-            "user_agent": os.getenv("USER_AGENT"),
-            "memory_backend": os.getenv("MEMORY_BACKEND"),
-            "memory_index": os.getenv("MEMORY_INDEX"),
-            "redis_host": os.getenv("REDIS_HOST"),
-            "redis_password": os.getenv("REDIS_PASSWORD"),
-            "wipe_redis_on_start": os.getenv("WIPE_REDIS_ON_START", "True") == "True",
-            "plugins_dir": os.getenv("PLUGINS_DIR"),
-            "plugins_config_file": project_root
-            / Path(os.getenv("PLUGINS_CONFIG_FILE", PLUGINS_CONFIG_FILE)),
-            "chat_messages_enabled": os.getenv("CHAT_MESSAGES_ENABLED") == "True",
-        }

-        config_dict["disabled_command_categories"] = _safe_split(
-            os.getenv("DISABLED_COMMAND_CATEGORIES")
-        )
+        config = cls.build_agent_configuration()
+        config.project_root = project_root

-        config_dict["shell_denylist"] = _safe_split(
-            os.getenv("SHELL_DENYLIST", os.getenv("DENY_COMMANDS"))
-        )
-        config_dict["shell_allowlist"] = _safe_split(
-            os.getenv("SHELL_ALLOWLIST", os.getenv("ALLOW_COMMANDS"))
-        )
+        # Make relative paths absolute
+        for k in {
+            "ai_settings_file",  # TODO: deprecate or repurpose
+            "prompt_settings_file",  # TODO: deprecate or repurpose
+            "plugins_config_file",  # TODO: move from project root
+            "azure_config_file",  # TODO: move from project root
+        }:
+            setattr(config, k, project_root / getattr(config, k))

-        config_dict["google_custom_search_engine_id"] = os.getenv(
-            "GOOGLE_CUSTOM_SEARCH_ENGINE_ID", os.getenv("CUSTOM_SEARCH_ENGINE_ID")
-        )
-
-        if os.getenv("ELEVENLABS_API_KEY"):
-            config_dict["tts_config"]["elevenlabs"] = {
-                "api_key": os.getenv("ELEVENLABS_API_KEY"),
-                "voice_id": os.getenv("ELEVENLABS_VOICE_ID", ""),
-            }
-        if os.getenv("STREAMELEMENTS_VOICE"):
-            config_dict["tts_config"]["streamelements"] = {
-                "voice": os.getenv("STREAMELEMENTS_VOICE"),
-            }
-
-        if not config_dict["tts_config"]["provider"]:
-            if os.getenv("USE_MAC_OS_TTS"):
-                default_tts_provider = "macos"
-            elif "elevenlabs" in config_dict["tts_config"]:
-                default_tts_provider = "elevenlabs"
-            elif os.getenv("USE_BRIAN_TTS"):
-                default_tts_provider = "streamelements"
-            else:
-                default_tts_provider = "gtts"
-            config_dict["tts_config"]["provider"] = default_tts_provider
-
-        config_dict["plugins_allowlist"] = _safe_split(os.getenv("ALLOWLISTED_PLUGINS"))
-        config_dict["plugins_denylist"] = _safe_split(os.getenv("DENYLISTED_PLUGINS"))
-
-        with contextlib.suppress(TypeError):
-            config_dict["image_size"] = int(os.getenv("IMAGE_SIZE"))
-        with contextlib.suppress(TypeError):
-            config_dict["redis_port"] = int(os.getenv("REDIS_PORT"))
-        with contextlib.suppress(TypeError):
-            config_dict["temperature"] = float(os.getenv("TEMPERATURE"))
-
-        if config_dict["use_azure"]:
-            azure_config = cls.load_azure_config(
-                project_root / config_dict["azure_config_file"]
-            )
-            config_dict.update(azure_config)
-
-        elif os.getenv("OPENAI_API_BASE_URL"):
-            config_dict["openai_api_base"] = os.getenv("OPENAI_API_BASE_URL")
-
-        openai_organization = os.getenv("OPENAI_ORGANIZATION")
-        if openai_organization is not None:
-            config_dict["openai_organization"] = openai_organization
-
-        config_dict_without_none_values = {
-            k: v for k, v in config_dict.items() if v is not None
-        }
-
-        config = cls.build_agent_configuration(config_dict_without_none_values)
-
-        # Set secondary config variables (that depend on other config variables)
+        if (
+            config.openai_credentials
+            and config.openai_credentials.api_type == "azure"
+            and (config_file := config.azure_config_file)
+        ):
+            config.openai_credentials.load_azure_config(config_file)

        config.plugins_config = PluginsConfig.load_config(
            config.plugins_config_file,
@@ -359,36 +286,10 @@ class ConfigBuilder(Configurable[Config]):

        return config

-    @classmethod
-    def load_azure_config(cls, config_file: Path) -> Dict[str, str]:
-        """
-        Loads the configuration parameters for Azure hosting from the specified file
-          path as a yaml file.
-
-        Parameters:
-            config_file (Path): The path to the config yaml file.
-
-        Returns:
-            Dict
-        """
-        with open(config_file) as file:
-            config_params = yaml.load(file, Loader=yaml.FullLoader) or {}
-
-        return {
-            "openai_api_type": config_params.get("azure_api_type", "azure"),
-            "openai_api_base": config_params.get("azure_api_base", ""),
-            "openai_api_version": config_params.get(
-                "azure_api_version", "2023-03-15-preview"
-            ),
-            "azure_model_to_deployment_id_map": config_params.get(
-                "azure_model_map", {}
-            ),
-        }
-

 def assert_config_has_openai_api_key(config: Config) -> None:
    """Check if the OpenAI API key is set in config.py or as an environment variable."""
-    if not config.openai_api_key:
+    if not config.openai_credentials:
        print(
            Fore.RED
            + "Please set your OpenAI API key in .env or as an environment variable."
@@ -402,7 +303,9 @@ def assert_config_has_openai_api_key(config: Config) -> None:
        openai_api_key = openai_api_key.strip()
        if re.search(key_pattern, openai_api_key):
            os.environ["OPENAI_API_KEY"] = openai_api_key
-            config.openai_api_key = openai_api_key
+            config.openai_credentials = OpenAICredentials(
+                api_key=SecretStr(openai_api_key)
+            )
            print(
                Fore.GREEN
                + "OpenAI API key successfully set!\n"
--- a/autogpts/autogpt/autogpt/core/configuration/schema.py
+++ b/autogpts/autogpt/autogpt/core/configuration/schema.py
@@ -1,24 +1,73 @@
 import abc
-import functools
+import os
 import typing
-from typing import Any, Generic, TypeVar
+from typing import Any, Callable, Generic, Optional, Type, TypeVar, get_args

-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, ValidationError
+from pydantic.fields import ModelField, Undefined, UndefinedType
+from pydantic.main import ModelMetaclass
+
+T = TypeVar("T")
+M = TypeVar("M", bound=BaseModel)


-@functools.wraps(Field)
-def UserConfigurable(*args, **kwargs):
-    return Field(*args, **kwargs, user_configurable=True)
+def UserConfigurable(
+    default: T | UndefinedType = Undefined,
+    *args,
+    default_factory: Optional[Callable[[], T]] = None,
+    from_env: Optional[str | Callable[[], T | None]] = None,
+    description: str = "",
+    **kwargs,
+) -> T:
    # TODO: use this to auto-generate docs for the application configuration
+    return Field(
+        default,
+        *args,
+        default_factory=default_factory,
+        from_env=from_env,
+        description=description,
+        **kwargs,
+        user_configurable=True,
+    )


 class SystemConfiguration(BaseModel):
    def get_user_config(self) -> dict[str, Any]:
-        return _get_user_config_fields(self)
+        return _recurse_user_config_values(self)
+
+    @classmethod
+    def from_env(cls):
+        """
+        Initializes the config object from environment variables.
+
+        Environment variables are mapped to UserConfigurable fields using the from_env
+        attribute that can be passed to UserConfigurable.
+        """
+
+        def infer_field_value(field: ModelField):
+            field_info = field.field_info
+            default_value = (
+                field.default
+                if field.default not in (None, Undefined)
+                else (field.default_factory() if field.default_factory else Undefined)
+            )
+            if from_env := field_info.extra.get("from_env"):
+                val_from_env = (
+                    os.getenv(from_env) if type(from_env) is str else from_env()
+                )
+                if val_from_env is not None:
+                    return val_from_env
+            return default_value
+
+        return _recursive_init_model(cls, infer_field_value)

    class Config:
        extra = "forbid"
        use_enum_values = True
+        validate_assignment = True
+
+
+SC = TypeVar("SC", bound=SystemConfiguration)


 class SystemSettings(BaseModel):
@@ -30,6 +79,7 @@ class SystemSettings(BaseModel):
    class Config:
        extra = "forbid"
        use_enum_values = True
+        validate_assignment = True


 S = TypeVar("S", bound=SystemSettings)
@@ -43,55 +93,238 @@ class Configurable(abc.ABC, Generic[S]):

    @classmethod
    def get_user_config(cls) -> dict[str, Any]:
-        return _get_user_config_fields(cls.default_settings)
+        return _recurse_user_config_values(cls.default_settings)

    @classmethod
-    def build_agent_configuration(cls, configuration: dict) -> S:
+    def build_agent_configuration(cls, overrides: dict = {}) -> S:
        """Process the configuration for this object."""

-        defaults = cls.default_settings.dict()
-        final_configuration = deep_update(defaults, configuration)
+        base_config = _update_user_config_from_env(cls.default_settings)
+        final_configuration = deep_update(base_config, overrides)

        return cls.default_settings.__class__.parse_obj(final_configuration)


-def _get_user_config_fields(instance: BaseModel) -> dict[str, Any]:
+def _update_user_config_from_env(instance: BaseModel) -> dict[str, Any]:
    """
-    Get the user config fields of a Pydantic model instance.
+    Update config fields of a Pydantic model instance from environment variables.

-    Args:
+    Precedence:
+    1. Non-default value already on the instance
+    2. Value returned by `from_env()`
+    3. Default value for the field
+
+    Params:
        instance: The Pydantic model instance.

    Returns:
        The user config fields of the instance.
    """
+
+    def infer_field_value(field: ModelField, value):
+        field_info = field.field_info
+        default_value = (
+            field.default
+            if field.default not in (None, Undefined)
+            else (field.default_factory() if field.default_factory else None)
+        )
+        if value == default_value and (from_env := field_info.extra.get("from_env")):
+            val_from_env = os.getenv(from_env) if type(from_env) is str else from_env()
+            if val_from_env is not None:
+                return val_from_env
+        return value
+
+    def init_sub_config(model: Type[SC]) -> SC | None:
+        try:
+            return model.from_env()
+        except ValidationError as e:
+            # Gracefully handle missing fields
+            if all(e["type"] == "value_error.missing" for e in e.errors()):
+                return None
+            raise
+
+    return _recurse_user_config_fields(instance, infer_field_value, init_sub_config)
+
+
+def _recursive_init_model(
+    model: Type[M],
+    infer_field_value: Callable[[ModelField], Any],
+) -> M:
+    """
+    Recursively initialize the user configuration fields of a Pydantic model.
+
+    Parameters:
+        model: The Pydantic model type.
+        infer_field_value: A callback function to infer the value of each field.
+            Parameters:
+                ModelField: The Pydantic ModelField object describing the field.
+
+    Returns:
+        BaseModel: An instance of the model with the initialized configuration.
+    """
+    user_config_fields = {}
+    for name, field in model.__fields__.items():
+        if "user_configurable" in field.field_info.extra:
+            user_config_fields[name] = infer_field_value(field)
+        elif type(field.outer_type_) is ModelMetaclass and issubclass(
+            field.outer_type_, SystemConfiguration
+        ):
+            try:
+                user_config_fields[name] = _recursive_init_model(
+                    model=field.outer_type_,
+                    infer_field_value=infer_field_value,
+                )
+            except ValidationError as e:
+                # Gracefully handle missing fields
+                if all(e["type"] == "value_error.missing" for e in e.errors()):
+                    user_config_fields[name] = None
+                raise
+
+    user_config_fields = remove_none_items(user_config_fields)
+
+    return model.parse_obj(user_config_fields)
+
+
+def _recurse_user_config_fields(
+    model: BaseModel,
+    infer_field_value: Callable[[ModelField, Any], Any],
+    init_sub_config: Optional[
+        Callable[[Type[SystemConfiguration]], SystemConfiguration | None]
+    ] = None,
+) -> dict[str, Any]:
+    """
+    Recursively process the user configuration fields of a Pydantic model instance.
+
+    Params:
+        model: The Pydantic model to iterate over.
+        infer_field_value: A callback function to process each field.
+            Params:
+                ModelField: The Pydantic ModelField object describing the field.
+                Any: The current value of the field.
+        init_sub_config: An optional callback function to initialize a sub-config.
+            Params:
+                Type[SystemConfiguration]: The type of the sub-config to initialize.
+
+    Returns:
+        dict[str, Any]: The processed user configuration fields of the instance.
+    """
    user_config_fields = {}

-    for name, value in instance.__dict__.items():
-        field_info = instance.__fields__[name]
-        if "user_configurable" in field_info.field_info.extra:
-            user_config_fields[name] = value
+    for name, field in model.__fields__.items():
+        value = getattr(model, name)
+
+        # Handle individual field
+        if "user_configurable" in field.field_info.extra:
+            user_config_fields[name] = infer_field_value(field, value)
+
+        # Recurse into nested config object
        elif isinstance(value, SystemConfiguration):
-            user_config_fields[name] = value.get_user_config()
+            user_config_fields[name] = _recurse_user_config_fields(
+                model=value,
+                infer_field_value=infer_field_value,
+                init_sub_config=init_sub_config,
+            )
+
+        # Recurse into optional nested config object
+        elif value is None and init_sub_config:
+            field_type = get_args(field.annotation)[0]  # Optional[T] -> T
+            if type(field_type) is ModelMetaclass and issubclass(
+                field_type, SystemConfiguration
+            ):
+                sub_config = init_sub_config(field_type)
+                if sub_config:
+                    user_config_fields[name] = _recurse_user_config_fields(
+                        model=sub_config,
+                        infer_field_value=infer_field_value,
+                        init_sub_config=init_sub_config,
+                    )
+
        elif isinstance(value, list) and all(
            isinstance(i, SystemConfiguration) for i in value
        ):
-            user_config_fields[name] = [i.get_user_config() for i in value]
+            user_config_fields[name] = [
+                _recurse_user_config_fields(i, infer_field_value, init_sub_config)
+                for i in value
+            ]
        elif isinstance(value, dict) and all(
            isinstance(i, SystemConfiguration) for i in value.values()
        ):
            user_config_fields[name] = {
-                k: v.get_user_config() for k, v in value.items()
+                k: _recurse_user_config_fields(v, infer_field_value, init_sub_config)
+                for k, v in value.items()
            }

    return user_config_fields


+def _recurse_user_config_values(
+    instance: BaseModel,
+    get_field_value: Callable[[ModelField, T], T] = lambda _, v: v,
+) -> dict[str, Any]:
+    """
+    This function recursively traverses the user configuration values in a Pydantic
+    model instance.
+
+    Params:
+        instance: A Pydantic model instance.
+        get_field_value: A callback function to process each field. Parameters:
+            ModelField: The Pydantic ModelField object that describes the field.
+            Any: The current value of the field.
+
+    Returns:
+        A dictionary containing the processed user configuration fields of the instance.
+    """
+    user_config_values = {}
+
+    for name, value in instance.__dict__.items():
+        field = instance.__fields__[name]
+        if "user_configurable" in field.field_info.extra:
+            user_config_values[name] = get_field_value(field, value)
+        elif isinstance(value, SystemConfiguration):
+            user_config_values[name] = _recurse_user_config_values(
+                instance=value, get_field_value=get_field_value
+            )
+        elif isinstance(value, list) and all(
+            isinstance(i, SystemConfiguration) for i in value
+        ):
+            user_config_values[name] = [
+                _recurse_user_config_values(i, get_field_value) for i in value
+            ]
+        elif isinstance(value, dict) and all(
+            isinstance(i, SystemConfiguration) for i in value.values()
+        ):
+            user_config_values[name] = {
+                k: _recurse_user_config_values(v, get_field_value)
+                for k, v in value.items()
+            }
+
+    return user_config_values
+
+
+def _get_non_default_user_config_values(instance: BaseModel) -> dict[str, Any]:
+    """
+    Get the non-default user config fields of a Pydantic model instance.
+
+    Params:
+        instance: The Pydantic model instance.
+
+    Returns:
+        dict[str, Any]: The non-default user config values on the instance.
+    """
+
+    def get_field_value(field: ModelField, value):
+        default = field.default_factory() if field.default_factory else field.default
+        if value != default:
+            return value
+
+    return remove_none_items(_recurse_user_config_values(instance, get_field_value))
+
+
 def deep_update(original_dict: dict, update_dict: dict) -> dict:
    """
    Recursively update a dictionary.

-    Args:
+    Params:
        original_dict (dict): The dictionary to be updated.
        update_dict (dict): The dictionary to update with.

@@ -108,3 +341,11 @@ def deep_update(original_dict: dict, update_dict: dict) -> dict:
        else:
            original_dict[key] = value
    return original_dict
+
+
+def remove_none_items(d):
+    if isinstance(d, dict):
+        return {
+            k: remove_none_items(v) for k, v in d.items() if v not in (None, Undefined)
+        }
+    return d
--- a/autogpts/autogpt/autogpt/core/resource/model_providers/openai.py
+++ b/autogpts/autogpt/autogpt/core/resource/model_providers/openai.py
@@ -2,12 +2,16 @@ import enum
 import functools
 import logging
 import math
+import os
 import time
+from pathlib import Path
 from typing import Callable, Optional, ParamSpec, TypeVar

 import openai
 import tiktoken
+import yaml
 from openai.error import APIError, RateLimitError
+from pydantic import SecretStr

 from autogpt.core.configuration import (
    Configurable,
@@ -167,6 +171,68 @@ class OpenAIConfiguration(SystemConfiguration):
    retries_per_request: int = UserConfigurable()


+class OpenAICredentials(ModelProviderCredentials):
+    """Credentials for OpenAI."""
+
+    api_key: SecretStr = UserConfigurable(from_env="OPENAI_API_KEY")
+    api_base: Optional[SecretStr] = UserConfigurable(
+        default=None, from_env="OPENAI_API_BASE_URL"
+    )
+    organization: Optional[SecretStr] = UserConfigurable(from_env="OPENAI_ORGANIZATION")
+
+    api_type: str = UserConfigurable(
+        default="",
+        from_env=lambda: (
+            "azure"
+            if os.getenv("USE_AZURE") == "True"
+            else os.getenv("OPENAI_API_TYPE")
+        ),
+    )
+    api_version: str = UserConfigurable("", from_env="OPENAI_API_VERSION")
+    azure_model_to_deploy_id_map: Optional[dict[str, str]] = None
+
+    def get_api_access_kwargs(self, model: str = "") -> dict[str, str]:
+        credentials = {k: v for k, v in self.unmasked().items() if type(v) is str}
+        if self.api_type == "azure" and model:
+            azure_credentials = self._get_azure_access_kwargs(model)
+            credentials.update(azure_credentials)
+        return credentials
+
+    def load_azure_config(self, config_file: Path) -> None:
+        with open(config_file) as file:
+            config_params = yaml.load(file, Loader=yaml.FullLoader) or {}
+
+        try:
+            assert (
+                azure_api_base := config_params.get("azure_api_base", "")
+            ) != "", "Azure API base URL not set"
+            assert config_params.get(
+                "azure_model_map", {}
+            ), "Azure model->deployment_id map is empty"
+        except AssertionError as e:
+            raise ValueError(*e.args)
+
+        self.api_base = SecretStr(azure_api_base)
+        self.api_type = config_params.get("azure_api_type", "azure")
+        self.api_version = config_params.get("azure_api_version", "")
+        self.azure_model_to_deploy_id_map = config_params.get("azure_model_map")
+
+    def _get_azure_access_kwargs(self, model: str) -> dict[str, str]:
+        """Get the kwargs for the Azure API."""
+
+        if not self.azure_model_to_deploy_id_map:
+            raise ValueError("Azure model deployment map not configured")
+
+        if model not in self.azure_model_to_deploy_id_map:
+            raise ValueError(f"No Azure deployment ID configured for model '{model}'")
+        deployment_id = self.azure_model_to_deploy_id_map[model]
+
+        if model in OPEN_AI_EMBEDDING_MODELS:
+            return {"engine": deployment_id}
+        else:
+            return {"deployment_id": deployment_id}
+
+
 class OpenAIModelProviderBudget(ModelProviderBudget):
    graceful_shutdown_threshold: float = UserConfigurable()
    warning_threshold: float = UserConfigurable()
@@ -174,7 +240,7 @@ class OpenAIModelProviderBudget(ModelProviderBudget):

 class OpenAISettings(ModelProviderSettings):
    configuration: OpenAIConfiguration
-    credentials: ModelProviderCredentials
+    credentials: Optional[OpenAICredentials]
    budget: OpenAIModelProviderBudget


@@ -187,7 +253,7 @@ class OpenAIProvider(
        configuration=OpenAIConfiguration(
            retries_per_request=10,
        ),
-        credentials=ModelProviderCredentials(),
+        credentials=None,
        budget=OpenAIModelProviderBudget(
            total_budget=math.inf,
            total_cost=0.0,
@@ -207,6 +273,7 @@ class OpenAIProvider(
        settings: OpenAISettings,
        logger: logging.Logger,
    ):
+        assert settings.credentials, "Cannot create OpenAIProvider without credentials"
        self._configuration = settings.configuration
        self._credentials = settings.credentials
        self._budget = settings.budget
@@ -362,7 +429,7 @@ class OpenAIProvider(
        completion_kwargs = {
            "model": model_name,
            **kwargs,
-            **self._credentials.unmasked(),
+            **self._credentials.get_api_access_kwargs(model_name),
        }

        if functions:
--- a/autogpts/autogpt/autogpt/logs/config.py
+++ b/autogpts/autogpt/autogpt/logs/config.py
@@ -3,6 +3,7 @@ from __future__ import annotations

 import enum
 import logging
+import os
 import sys
 from pathlib import Path
 from typing import TYPE_CHECKING, Optional
@@ -14,7 +15,7 @@ if TYPE_CHECKING:
    from autogpt.config import Config
    from autogpt.speech import TTSConfig

-from autogpt.core.configuration import SystemConfiguration
+from autogpt.core.configuration import SystemConfiguration, UserConfigurable
 from autogpt.core.runner.client_lib.logging import BelowLevelFilter

 from .formatters import AutoGptFormatter, StructuredLoggingFormatter
@@ -49,15 +50,29 @@ TEXT_LOG_FORMAT_MAP = {


 class LoggingConfig(SystemConfiguration):
-    level: int = logging.INFO
+    level: int = UserConfigurable(
+        default=logging.INFO,
+        from_env=lambda: logging.getLevelName(os.getenv("LOG_LEVEL", "INFO")),
+    )

    # Console output
-    log_format: LogFormatName = LogFormatName.SIMPLE
-    plain_console_output: bool = False
+    log_format: LogFormatName = UserConfigurable(
+        default=LogFormatName.SIMPLE,
+        from_env=lambda: LogFormatName(os.getenv("LOG_FORMAT", "simple")),
+    )
+    plain_console_output: bool = UserConfigurable(
+        default=False,
+        from_env=lambda: os.getenv("PLAIN_OUTPUT", "False") == "True",
+    )

    # File output
    log_dir: Path = LOG_DIR
-    log_file_format: Optional[LogFormatName] = LogFormatName.SIMPLE
+    log_file_format: Optional[LogFormatName] = UserConfigurable(
+        default=LogFormatName.SIMPLE,
+        from_env=lambda: LogFormatName(
+            os.getenv("LOG_FILE_FORMAT", os.getenv("LOG_FORMAT", "simple"))
+        ),
+    )


 def configure_logging(
--- a/autogpts/autogpt/autogpt/speech/eleven_labs.py
+++ b/autogpts/autogpt/autogpt/speech/eleven_labs.py
@@ -17,8 +17,8 @@ PLACEHOLDERS = {"your-voice-id"}


 class ElevenLabsConfig(SystemConfiguration):
-    api_key: str = UserConfigurable()
-    voice_id: str = UserConfigurable()
+    api_key: str = UserConfigurable(from_env="ELEVENLABS_API_KEY")
+    voice_id: str = UserConfigurable(from_env="ELEVENLABS_VOICE_ID")


 class ElevenLabsSpeech(VoiceBase):
--- a/autogpts/autogpt/autogpt/speech/say.py
+++ b/autogpts/autogpt/autogpt/speech/say.py
@@ -1,6 +1,7 @@
 """ Text to speech module """
 from __future__ import annotations

+import os
 import threading
 from threading import Semaphore
 from typing import Literal, Optional
@@ -20,11 +21,23 @@ _QUEUE_SEMAPHORE = Semaphore(

 class TTSConfig(SystemConfiguration):
    speak_mode: bool = False
-    provider: Literal[
-        "elevenlabs", "gtts", "macos", "streamelements"
-    ] = UserConfigurable(default="gtts")
    elevenlabs: Optional[ElevenLabsConfig] = None
    streamelements: Optional[StreamElementsConfig] = None
+    provider: Literal[
+        "elevenlabs", "gtts", "macos", "streamelements"
+    ] = UserConfigurable(
+        default="gtts",
+        from_env=lambda: os.getenv("TEXT_TO_SPEECH_PROVIDER")
+        or (
+            "macos"
+            if os.getenv("USE_MAC_OS_TTS")
+            else "elevenlabs"
+            if os.getenv("ELEVENLABS_API_KEY")
+            else "streamelements"
+            if os.getenv("USE_BRIAN_TTS")
+            else "gtts"
+        ),
+    )  # type: ignore


 class TextToSpeechProvider:
--- a/autogpts/autogpt/autogpt/speech/stream_elements_speech.py
+++ b/autogpts/autogpt/autogpt/speech/stream_elements_speech.py
@@ -13,7 +13,7 @@ logger = logging.getLogger(__name__)


 class StreamElementsConfig(SystemConfiguration):
-    voice: str = UserConfigurable(default="Brian")
+    voice: str = UserConfigurable(default="Brian", from_env="STREAMELEMENTS_VOICE")


 class StreamElementsSpeech(VoiceBase):
--- a/autogpts/autogpt/azure.yaml.template
+++ b/autogpts/autogpt/azure.yaml.template
@@ -2,6 +2,6 @@ azure_api_type: azure
 azure_api_base: your-base-url-for-azure
 azure_api_version: api-version-for-azure
 azure_model_map:
-    fast_llm_deployment_id: gpt35-deployment-id-for-azure
-    smart_llm_deployment_id: gpt4-deployment-id-for-azure
-    embedding_model_deployment_id: embedding-deployment-id-for-azure
+    gpt-3.5-turbo: gpt35-deployment-id-for-azure
+    gpt-4: gpt4-deployment-id-for-azure
+    text-embedding-ada-002: embedding-deployment-id-for-azure
--- a/autogpts/autogpt/tests/conftest.py
+++ b/autogpts/autogpt/tests/conftest.py
@@ -30,7 +30,9 @@ def tmp_project_root(tmp_path: Path) -> Path:

@pytest.fixture()
 def app_data_dir(tmp_project_root: Path) -> Path:
-    return tmp_project_root / "data"
+    dir = tmp_project_root / "data"
+    dir.mkdir(parents=True, exist_ok=True)
+    return dir


@pytest.fixture()
@@ -71,9 +73,9 @@ def config(
    app_data_dir: Path,
    mocker: MockerFixture,
 ):
-    config = ConfigBuilder.build_config_from_env(project_root=tmp_project_root)
    if not os.environ.get("OPENAI_API_KEY"):
        os.environ["OPENAI_API_KEY"] = "sk-dummy"
+    config = ConfigBuilder.build_config_from_env(project_root=tmp_project_root)

    config.app_data_dir = app_data_dir

--- a/autogpts/autogpt/tests/integration/test_image_gen.py
+++ b/autogpts/autogpt/tests/integration/test_image_gen.py
@@ -99,7 +99,8 @@ def generate_and_validate(
 ):
    """Generate an image and validate the output."""
    agent.legacy_config.image_provider = image_provider
-    agent.legacy_config.huggingface_image_model = hugging_face_image_model
+    if hugging_face_image_model:
+        agent.legacy_config.huggingface_image_model = hugging_face_image_model
    prompt = "astronaut riding a horse"

    image_path = lst(generate_image(prompt, agent, image_size, **kwargs))
--- a/autogpts/autogpt/tests/unit/test_config.py
+++ b/autogpts/autogpt/tests/unit/test_config.py
@@ -8,10 +8,10 @@ from unittest import mock
 from unittest.mock import patch

 import pytest
+from pydantic import SecretStr

 from autogpt.app.configurator import GPT_3_MODEL, GPT_4_MODEL, apply_overrides_to_config
 from autogpt.config import Config, ConfigBuilder
-from autogpt.file_workspace import FileWorkspace


 def test_initial_values(config: Config) -> None:
@@ -107,73 +107,81 @@ def test_smart_and_fast_llms_set_to_gpt4(mock_list_models: Any, config: Config)
    config.smart_llm = smart_llm


-def test_missing_azure_config(workspace: FileWorkspace) -> None:
-    config_file = workspace.get_path("azure_config.yaml")
+def test_missing_azure_config(config: Config) -> None:
+    assert config.openai_credentials is not None
+
+    config_file = config.app_data_dir / "azure_config.yaml"
    with pytest.raises(FileNotFoundError):
-        ConfigBuilder.load_azure_config(config_file)
+        config.openai_credentials.load_azure_config(config_file)

    config_file.write_text("")
-    azure_config = ConfigBuilder.load_azure_config(config_file)
+    with pytest.raises(ValueError):
+        config.openai_credentials.load_azure_config(config_file)

-    assert azure_config["openai_api_type"] == "azure"
-    assert azure_config["openai_api_base"] == ""
-    assert azure_config["openai_api_version"] == "2023-03-15-preview"
-    assert azure_config["azure_model_to_deployment_id_map"] == {}
+    assert config.openai_credentials.api_type != "azure"
+    assert config.openai_credentials.api_version == ""
+    assert config.openai_credentials.azure_model_to_deploy_id_map is None


-def test_azure_config(config: Config, workspace: FileWorkspace) -> None:
-    config_file = workspace.get_path("azure_config.yaml")
-    yaml_content = """
+def test_azure_config(config: Config) -> None:
+    config_file = config.app_data_dir / "azure_config.yaml"
+    config_file.write_text(
+        f"""
 azure_api_type: azure
 azure_api_base: https://dummy.openai.azure.com
 azure_api_version: 2023-06-01-preview
 azure_model_map:
-    fast_llm_deployment_id: FAST-LLM_ID
-    smart_llm_deployment_id: SMART-LLM_ID
-    embedding_model_deployment_id: embedding-deployment-id-for-azure
+    {config.fast_llm}: FAST-LLM_ID
+    {config.smart_llm}: SMART-LLM_ID
+    {config.embedding_model}: embedding-deployment-id-for-azure
 """
-    config_file.write_text(yaml_content)
+    )

    os.environ["USE_AZURE"] = "True"
    os.environ["AZURE_CONFIG_FILE"] = str(config_file)
-    config = ConfigBuilder.build_config_from_env(project_root=workspace.root.parent)
+    config = ConfigBuilder.build_config_from_env(project_root=config.project_root)

-    assert config.openai_api_type == "azure"
-    assert config.openai_api_base == "https://dummy.openai.azure.com"
-    assert config.openai_api_version == "2023-06-01-preview"
-    assert config.azure_model_to_deployment_id_map == {
-        "fast_llm_deployment_id": "FAST-LLM_ID",
-        "smart_llm_deployment_id": "SMART-LLM_ID",
-        "embedding_model_deployment_id": "embedding-deployment-id-for-azure",
+    assert (credentials := config.openai_credentials) is not None
+    assert credentials.api_type == "azure"
+    assert credentials.api_base == SecretStr("https://dummy.openai.azure.com")
+    assert credentials.api_version == "2023-06-01-preview"
+    assert credentials.azure_model_to_deploy_id_map == {
+        config.fast_llm: "FAST-LLM_ID",
+        config.smart_llm: "SMART-LLM_ID",
+        config.embedding_model: "embedding-deployment-id-for-azure",
    }

    fast_llm = config.fast_llm
    smart_llm = config.smart_llm
    assert (
-        config.get_azure_credentials(config.fast_llm)["deployment_id"] == "FAST-LLM_ID"
+        credentials.get_api_access_kwargs(config.fast_llm)["deployment_id"]
+        == "FAST-LLM_ID"
    )
    assert (
-        config.get_azure_credentials(config.smart_llm)["deployment_id"]
+        credentials.get_api_access_kwargs(config.smart_llm)["deployment_id"]
        == "SMART-LLM_ID"
    )

    # Emulate --gpt4only
    config.fast_llm = smart_llm
    assert (
-        config.get_azure_credentials(config.fast_llm)["deployment_id"] == "SMART-LLM_ID"
+        credentials.get_api_access_kwargs(config.fast_llm)["deployment_id"]
+        == "SMART-LLM_ID"
    )
    assert (
-        config.get_azure_credentials(config.smart_llm)["deployment_id"]
+        credentials.get_api_access_kwargs(config.smart_llm)["deployment_id"]
        == "SMART-LLM_ID"
    )

    # Emulate --gpt3only
    config.fast_llm = config.smart_llm = fast_llm
    assert (
-        config.get_azure_credentials(config.fast_llm)["deployment_id"] == "FAST-LLM_ID"
+        credentials.get_api_access_kwargs(config.fast_llm)["deployment_id"]
+        == "FAST-LLM_ID"
    )
    assert (
-        config.get_azure_credentials(config.smart_llm)["deployment_id"] == "FAST-LLM_ID"
+        credentials.get_api_access_kwargs(config.smart_llm)["deployment_id"]
+        == "FAST-LLM_ID"
    )

    del os.environ["USE_AZURE"]
--- a/docs/content/AutoGPT/setup/docker.md
+++ b/docs/content/AutoGPT/setup/docker.md
@@ -144,7 +144,7 @@ found in the [repository].

    **Note:** Azure support has been dropped in `master`, so these instructions will only work with v0.4.7 (or earlier).

-[repository]: https://github.com/Significant-Gravitas/AutoGPT/autogpts/autogpt
+[repository]: https://github.com/Significant-Gravitas/AutoGPT/tree/master/autogpts/autogpt
 [show hidden files/Windows]: https://support.microsoft.com/en-us/windows/view-hidden-files-and-folders-in-windows-97fbc472-c603-9d90-91d0-1166d1d9f4b5
 [show hidden files/macOS]: https://www.pcmag.com/how-to/how-to-access-your-macs-hidden-files
 [openai-python docs]: https://github.com/openai/openai-python#microsoft-azure-endpoints
--- a/docs/content/AutoGPT/setup/index.md
+++ b/docs/content/AutoGPT/setup/index.md
@@ -88,6 +88,28 @@ Once you have cloned or downloaded the project, you can find the AutoGPT Agent i
    ```yaml
    OPENAI_API_KEY=sk-qwertykeys123456
    ```
+
+    !!! info "Using a GPT Azure-instance"
+        If you want to use GPT on an Azure instance, set `USE_AZURE` to `True` and
+        make an Azure configuration file.
+
+        Rename `azure.yaml.template` to `azure.yaml` and provide the relevant
+        `azure_api_base`, `azure_api_version` and deployment IDs for the models that you
+        want to use.
+
+        E.g. if you want to use `gpt-3.5-turbo-16k` and `gpt-4-0314`:
+
+        ```yaml
+        # Please specify all of these values as double-quoted strings
+        # Replace string in angled brackets (<>) to your own deployment Name
+        azure_model_map:
+            gpt-3.5-turbo-16k: "<auto-gpt-deployment>"
+            ...
+        ```
+
+        Details can be found in the [openai-python docs], and in the [Azure OpenAI docs] for the embedding model.
+        If you're on Windows you may need to install an [MSVC library](https://learn.microsoft.com/en-us/cpp/windows/latest-supported-vc-redist?view=msvc-170).
+
 6. Enter any other API keys or tokens for services you would like to use.

    !!! note