diff --git a/.env.template b/.env.template index 22bf8d74..6565dfdc 100644 --- a/.env.template +++ b/.env.template @@ -11,6 +11,9 @@ BROWSE_SUMMARY_MAX_TOKEN=300 # USER_AGENT="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" # AI_SETTINGS_FILE - Specifies which AI Settings file to use (defaults to ai_settings.yaml) AI_SETTINGS_FILE=ai_settings.yaml +# USE_WEB_BROWSER - Sets the web-browser drivers to use with selenium (defaults to chrome). +# Note: set this to either 'chrome', 'firefox', or 'safari' depending on your current browser +# USE_WEB_BROWSER=chrome ################################################################################ ### LLM PROVIDER @@ -21,20 +24,11 @@ AI_SETTINGS_FILE=ai_settings.yaml # TEMPERATURE - Sets temperature in OpenAI (Default: 1) # USE_AZURE - Use Azure OpenAI or not (Default: False) OPENAI_API_KEY=your-openai-api-key -TEMPERATURE=1 +TEMPERATURE=0 USE_AZURE=False ### AZURE -# OPENAI_AZURE_API_BASE - OpenAI API base URL for Azure (Example: https://my-azure-openai-url.com) -# OPENAI_AZURE_API_VERSION - OpenAI API version for Azure (Example: v1) -# OPENAI_AZURE_DEPLOYMENT_ID - OpenAI deployment ID for Azure (Example: my-deployment-id) -# OPENAI_AZURE_CHAT_DEPLOYMENT_ID - OpenAI deployment ID for Azure Chat (Example: my-deployment-id-for-azure-chat) -# OPENAI_AZURE_EMBEDDINGS_DEPLOYMENT_ID - OpenAI deployment ID for Embedding (Example: my-deployment-id-for-azure-embeddigs) -OPENAI_AZURE_API_BASE=your-base-url-for-azure -OPENAI_AZURE_API_VERSION=api-version-for-azure -OPENAI_AZURE_DEPLOYMENT_ID=deployment-id-for-azure -OPENAI_AZURE_CHAT_DEPLOYMENT_ID=deployment-id-for-azure-chat -OPENAI_AZURE_EMBEDDINGS_DEPLOYMENT_ID=deployment-id-for-azure-embeddigs +# cleanup azure env as already moved to `azure.yaml.template` ################################################################################ ### LLM MODELS @@ -48,7 +42,7 @@ FAST_LLM_MODEL=gpt-3.5-turbo ### LLM MODEL SETTINGS # FAST_TOKEN_LIMIT - Fast token limit for OpenAI (Default: 4000) # SMART_TOKEN_LIMIT - Smart token limit for OpenAI (Default: 8000) -# When using --gpt3onlythis needs to be set to 4000. +# When using --gpt3only this needs to be set to 4000. FAST_TOKEN_LIMIT=4000 SMART_TOKEN_LIMIT=8000 @@ -77,6 +71,13 @@ REDIS_PASSWORD= WIPE_REDIS_ON_START=False MEMORY_INDEX=auto-gpt +### MILVUS +# MILVUS_ADDR - Milvus remote address (e.g. localhost:19530) +# MILVUS_COLLECTION - Milvus collection, +# change it if you want to start a new memory and retain the old memory. +MILVUS_ADDR=your-milvus-cluster-host-port +MILVUS_COLLECTION=autogpt + ################################################################################ ### IMAGE GENERATION PROVIDER ################################################################################ @@ -92,6 +93,16 @@ IMAGE_PROVIDER=dalle # HUGGINGFACE_API_TOKEN - HuggingFace API token (Example: my-huggingface-api-token) HUGGINGFACE_API_TOKEN=your-huggingface-api-token +################################################################################ +### GIT Provider for repository actions +################################################################################ + +### GITHUB +# GITHUB_API_KEY - Github API key / PAT (Example: github_pat_123) +# GITHUB_USERNAME - Github username +GITHUB_API_KEY=github_pat_123 +GITHUB_USERNAME=your-github-username + ################################################################################ ### SEARCH PROVIDER ################################################################################ diff --git a/.gitignore b/.gitignore index 9874d425..3209297c 100644 --- a/.gitignore +++ b/.gitignore @@ -19,6 +19,8 @@ auto-gpt.json log.txt log-ingestion.txt logs +*.log +*.mp3 # Byte-compiled / optimized / DLL files __pycache__/ @@ -151,4 +153,7 @@ dmypy.json # Pyre type checker .pyre/ llama-* -vicuna-* \ No newline at end of file +vicuna-* + +# mac +.DS_Store diff --git a/Dockerfile b/Dockerfile index 3ae1ac12..309b857c 100644 --- a/Dockerfile +++ b/Dockerfile @@ -1,6 +1,10 @@ # Use an official Python base image from the Docker Hub FROM python:3.11-slim +# Install git +RUN apt-get -y update +RUN apt-get -y install git + # Set environment variables ENV PIP_NO_CACHE_DIR=yes \ PYTHONUNBUFFERED=1 \ @@ -13,11 +17,11 @@ RUN chown appuser:appuser /home/appuser USER appuser # Copy the requirements.txt file and install the requirements -COPY --chown=appuser:appuser requirements.txt . -RUN pip install --no-cache-dir --user -r requirements.txt +COPY --chown=appuser:appuser requirements-docker.txt . +RUN pip install --no-cache-dir --user -r requirements-docker.txt # Copy the application files -COPY --chown=appuser:appuser autogpt/ . +COPY --chown=appuser:appuser autogpt/ ./autogpt # Set the entrypoint ENTRYPOINT ["python", "-m", "autogpt"] diff --git a/README.md b/README.md index 7da3172a..5e2ac0e7 100644 --- a/README.md +++ b/README.md @@ -1,8 +1,10 @@ # Auto-GPT: An Autonomous GPT-4 Experiment ### πŸ”΄ πŸ”΄ πŸ”΄ Urgent: USE `stable` not `master` πŸ”΄ πŸ”΄ πŸ”΄ -We've improved our workflow. `master` will often be **in a broken state**. -Download the latest `stable` release here: https://github.com/Torantulino/Auto-GPT/releases/latest -This information SUPERCEDES any following information. Takes precedence. Do this not that. + +> This information takes precedence over any other information provided and should be followed accordingly. + +Our workflow has been improved, but please note that `master` branch may often be in a **broken** state. +Please download the latest `stable` release from here: https://github.com/Torantulino/Auto-GPT/releases/latest. ![GitHub Repo stars](https://img.shields.io/github/stars/Torantulino/auto-gpt?style=social) [![Twitter Follow](https://img.shields.io/twitter/follow/siggravitas?style=social)](https://twitter.com/SigGravitas) @@ -35,22 +37,27 @@ Your support is greatly appreciated ## Table of Contents - [Auto-GPT: An Autonomous GPT-4 Experiment](#auto-gpt-an-autonomous-gpt-4-experiment) - - [Demo (30/03/2023):](#demo-30032023) + - [πŸ”΄ πŸ”΄ πŸ”΄ Urgent: USE `stable` not `master` πŸ”΄ πŸ”΄ πŸ”΄](#----urgent-use-stable-not-master----) + - [Demo (30/03/2023):](#demo-30032023) - [Table of Contents](#table-of-contents) - [πŸš€ Features](#-features) - [πŸ“‹ Requirements](#-requirements) - [πŸ’Ύ Installation](#-installation) - [πŸ”§ Usage](#-usage) - [Logs](#logs) + - [Docker](#docker) + - [Command Line Arguments](#command-line-arguments) - [πŸ—£οΈ Speech Mode](#️-speech-mode) - [πŸ” Google API Keys Configuration](#-google-api-keys-configuration) - [Setting up environment variables](#setting-up-environment-variables) - - [Redis Setup](#redis-setup) - - [🌲 Pinecone API Key Setup](#-pinecone-api-key-setup) + - [Memory Backend Setup](#memory-backend-setup) + - [Redis Setup](#redis-setup) + - [🌲 Pinecone API Key Setup](#-pinecone-api-key-setup) + - [Milvus Setup](#milvus-setup) - [Setting up environment variables](#setting-up-environment-variables-1) - [Setting Your Cache Type](#setting-your-cache-type) - [View Memory Usage](#view-memory-usage) - - [🧠 Memory pre-seeding](#memory-pre-seeding) + - [🧠 Memory pre-seeding](#-memory-pre-seeding) - [πŸ’€ Continuous Mode ⚠️](#-continuous-mode-️) - [GPT3.5 ONLY Mode](#gpt35-only-mode) - [πŸ–Ό Image Generation](#-image-generation) @@ -75,69 +82,76 @@ Your support is greatly appreciated - [Python 3.8 or later](https://www.tutorialspoint.com/how-to-install-python-in-windows) - [OpenAI API key](https://platform.openai.com/account/api-keys) - Optional: -- [PINECONE API key](https://www.pinecone.io/) (If you want Pinecone backed memory) +- Memory backend + - [PINECONE API key](https://www.pinecone.io/) (If you want Pinecone backed memory) + - [Milvus](https://milvus.io/) (If you want Milvus as memory backend) - ElevenLabs Key (If you want the AI to speak) ## πŸ’Ύ Installation To install Auto-GPT, follow these steps: -1. Make sure you have all the **requirements** above, if not, install/get them. +1. Make sure you have all the **requirements** listed above, if not, install/get them -_The following commands should be executed in a CMD, Bash or Powershell window. To do this, go to a folder on your computer, click in the folder path at the top and type CMD, then press enter._ +_To execute the following commands, open a CMD, Bash, or Powershell window by navigating to a folder on your computer and typing `CMD` in the folder path at the top, then press enter._ -2. Clone the repository: - For this step you need Git installed, but you can just download the zip file instead by clicking the button at the top of this page ☝️ +2. Clone the repository: For this step, you need Git installed. Alternatively, you can download the zip file by clicking the button at the top of this page ☝️ -``` +```bash git clone https://github.com/Torantulino/Auto-GPT.git ``` -3. Navigate to the project directory: - _(Type this into your CMD window, you're aiming to navigate the CMD window to the repository you just downloaded)_ +3. Navigate to the directory where the repository was downloaded -``` -cd 'Auto-GPT' +```bash +cd Auto-GPT ``` -4. Install the required dependencies: - _(Again, type this into your CMD window)_ +4. Install the required dependencies -``` +```bash pip install -r requirements.txt ``` 5. Rename `.env.template` to `.env` and fill in your `OPENAI_API_KEY`. If you plan to use Speech Mode, fill in your `ELEVEN_LABS_API_KEY` as well. - - Obtain your OpenAI API key from: https://platform.openai.com/account/api-keys. + - See [OpenAI API Keys Configuration](#openai-api-keys-configuration) to obtain your OpenAI API key. - Obtain your ElevenLabs API key from: https://elevenlabs.io. You can view your xi-api-key using the "Profile" tab on the website. - - If you want to use GPT on an Azure instance, set `USE_AZURE` to `True` and then: - - Rename `azure.yaml.template` to `azure.yaml` and provide the relevant `azure_api_base`, `azure_api_version` and all of the deployment ids for the relevant models in the `azure_model_map` section: - - `fast_llm_model_deployment_id` - your gpt-3.5-turbo or gpt-4 deployment id - - `smart_llm_model_deployment_id` - your gpt-4 deployment id - - `embedding_model_deployment_id` - your text-embedding-ada-002 v2 deployment id - - Please specify all of these values as double quoted strings - - details can be found here: https://pypi.org/project/openai/ in the `Microsoft Azure Endpoints` section and here: https://learn.microsoft.com/en-us/azure/cognitive-services/openai/tutorials/embeddings?tabs=command-line for the embedding model. + - If you want to use GPT on an Azure instance, set `USE_AZURE` to `True` and then follow these steps: + - Rename `azure.yaml.template` to `azure.yaml` and provide the relevant `azure_api_base`, `azure_api_version` and all the deployment IDs for the relevant models in the `azure_model_map` section: + - `fast_llm_model_deployment_id` - your gpt-3.5-turbo or gpt-4 deployment ID + - `smart_llm_model_deployment_id` - your gpt-4 deployment ID + - `embedding_model_deployment_id` - your text-embedding-ada-002 v2 deployment ID + - Please specify all of these values as double-quoted strings + > Replace string in angled brackets (<>) to your own ID + ```yaml + azure_model_map: + fast_llm_model_deployment_id: "" + ... + ``` + - Details can be found here: https://pypi.org/project/openai/ in the `Microsoft Azure Endpoints` section and here: https://learn.microsoft.com/en-us/azure/cognitive-services/openai/tutorials/embeddings?tabs=command-line for the embedding model. ## πŸ”§ Usage -1. Run the `autogpt` Python module in your terminal: - _(Type this into your CMD window)_ +1. Run `autogpt` Python module in your terminal ``` python -m autogpt ``` -2. After each of action, enter 'y' to authorise command, 'y -N' to run N continuous commands, 'n' to exit program, or enter additional feedback for the AI. +2. After each action, choose from options to authorize command(s), +exit the program, or provide feedback to the AI. + 1. Authorize a single command, enter `y` + 2. Authorize a series of _N_ continuous commands, enter `y -N` + 3. Exit the program, enter `n` ### Logs -You will find activity and error logs in the folder `./output/logs` +Activity and error logs are located in the `./output/logs` -To output debug logs: +To print out debug logs: ``` python -m autogpt --debug @@ -159,20 +173,40 @@ docker run -it --env-file=./.env -v $PWD/auto_gpt_workspace:/app/auto_gpt_worksp ### Command Line Arguments Here are some common arguments you can use when running Auto-GPT: > Replace anything in angled brackets (<>) to a value you want to specify -* `python scripts/main.py --help` to see a list of all available command line arguments. -* `python scripts/main.py --ai-settings ` to run Auto-GPT with a different AI Settings file. -* `python scripts/main.py --use-memory ` to specify one of 3 memory backends: `local`, `redis`, `pinecone` or 'no_memory'. +* View all available command line arguments +```bash +python scripts/main.py --help +``` +* Run Auto-GPT with a different AI Settings file +```bash +python scripts/main.py --ai-settings +``` +* Specify one of 3 memory backends: `local`, `redis`, `pinecone` or `no_memory` +```bash +python scripts/main.py --use-memory +``` > **NOTE**: There are shorthands for some of these flags, for example `-m` for `--use-memory`. Use `python scripts/main.py --help` for more information ## πŸ—£οΈ Speech Mode -Use this to use TTS for Auto-GPT +Use this to use TTS _(Text-to-Speech)_ for Auto-GPT -``` +```bash python -m autogpt --speak ``` +## OpenAI API Keys Configuration + +Obtain your OpenAI API key from: https://platform.openai.com/account/api-keys. + +To use OpenAI API key for Auto-GPT, you NEED to have billing set up (AKA paid account). + +You can set up paid account at https://platform.openai.com/account/billing/overview. + +![For OpenAI API key to work, set up paid account at OpenAI API > Billing](./docs/imgs/openai-api-key-billing-paid-account.png) + + ## πŸ” Google API Keys Configuration This section is optional, use the official google api if you are having issues with error 429 when running a google search. @@ -195,58 +229,50 @@ _Remember that your free daily custom search quota allows only up to 100 searche For Windows Users: -``` +```bash setx GOOGLE_API_KEY "YOUR_GOOGLE_API_KEY" setx CUSTOM_SEARCH_ENGINE_ID "YOUR_CUSTOM_SEARCH_ENGINE_ID" - ``` For macOS and Linux users: -``` +```bash export GOOGLE_API_KEY="YOUR_GOOGLE_API_KEY" export CUSTOM_SEARCH_ENGINE_ID="YOUR_CUSTOM_SEARCH_ENGINE_ID" - ``` ## Redis Setup - -Install docker desktop. - -Run: - -``` +> _**CAUTION**_ \ +This is not intended to be publicly accessible and lacks security measures. Therefore, avoid exposing Redis to the internet without a password or at all +1. Install docker desktop +```bash docker run -d --name redis-stack-server -p 6379:6379 redis/redis-stack-server:latest ``` +> See https://hub.docker.com/r/redis/redis-stack-server for setting a password and additional configuration. -See https://hub.docker.com/r/redis/redis-stack-server for setting a password and additional configuration. - -Set the following environment variables: - -``` +2. Set the following environment variables +> Replace **PASSWORD** in angled brackets (<>) +```bash MEMORY_BACKEND=redis REDIS_HOST=localhost REDIS_PORT=6379 -REDIS_PASSWORD= +REDIS_PASSWORD= ``` - -Note that this is not intended to be run facing the internet and is not secure, do not expose redis to the internet without a password or at all really. - You can optionally set -``` +```bash WIPE_REDIS_ON_START=False ``` -To persist memory stored in Redis. +To persist memory stored in Redis You can specify the memory index for redis using the following: -``` -MEMORY_INDEX=whatever +```bash +MEMORY_INDEX= ``` -## 🌲 Pinecone API Key Setup +### 🌲 Pinecone API Key Setup Pinecone enables the storage of vast amounts of vector-based memory, allowing for only relevant memories to be loaded for the agent at any given time. @@ -254,34 +280,46 @@ Pinecone enables the storage of vast amounts of vector-based memory, allowing fo 2. Choose the `Starter` plan to avoid being charged. 3. Find your API key and region under the default project in the left sidebar. +### Milvus Setup + +[Milvus](https://milvus.io/) is a open-source, high scalable vector database to storage huge amount of vector-based memory and provide fast relevant search. + +- setup milvus database, keep your pymilvus version and milvus version same to avoid compatible issues. + - setup by open source [Install Milvus](https://milvus.io/docs/install_standalone-operator.md) + - or setup by [Zilliz Cloud](https://zilliz.com/cloud) +- set `MILVUS_ADDR` in `.env` to your milvus address `host:ip`. +- set `MEMORY_BACKEND` in `.env` to `milvus` to enable milvus as backend. +- optional + - set `MILVUS_COLLECTION` in `.env` to change milvus collection name as you want, `autogpt` is the default name. + ### Setting up environment variables In the `.env` file set: - `PINECONE_API_KEY` -- `PINECONE_ENV` (something like: us-east4-gcp) +- `PINECONE_ENV` (example: _"us-east4-gcp"_) - `MEMORY_BACKEND=pinecone` Alternatively, you can set them from the command line (advanced): For Windows Users: -``` -setx PINECONE_API_KEY "YOUR_PINECONE_API_KEY" -setx PINECONE_ENV "Your pinecone region" # something like: us-east4-gcp +```bash +setx PINECONE_API_KEY "" +setx PINECONE_ENV "" # e.g: "us-east4-gcp" setx MEMORY_BACKEND "pinecone" ``` For macOS and Linux users: -``` -export PINECONE_API_KEY="YOUR_PINECONE_API_KEY" -export PINECONE_ENV="Your pinecone region" # something like: us-east4-gcp +```bash +export PINECONE_API_KEY="" +export PINECONE_ENV="" # e.g: "us-east4-gcp" export MEMORY_BACKEND="pinecone" ``` ## Setting Your Cache Type -By default Auto-GPT is going to use LocalCache instead of redis or Pinecone. +By default, Auto-GPT is going to use LocalCache instead of redis or Pinecone. To switch to either, change the `MEMORY_BACKEND` env variable to the value that you want: @@ -296,7 +334,6 @@ To switch to either, change the `MEMORY_BACKEND` env variable to the value that ## 🧠 Memory pre-seeding -``` # python autogpt/data_ingestion.py -h usage: data_ingestion.py [-h] (--file FILE | --dir DIR) [--init] [--overlap OVERLAP] [--max_length MAX_LENGTH] @@ -322,35 +359,34 @@ To pre-seed the memory, the content of each document is split into chunks of a s This technique is particularly useful when working with large amounts of data or when there is specific information that the AI needs to be able to access quickly. By pre-seeding the memory, the AI can retrieve and use this information more efficiently, saving time, API call and improving the accuracy of its responses. -You could for example download the documentation of an API, a Github repository, etc. and ingest it into memory before running Auto-GPT. +You could for example download the documentation of an API, a GitHub repository, etc. and ingest it into memory before running Auto-GPT. -⚠️ If you use Redis as your memory, make sure to run Auto-GPT with the WIPE_REDIS_ON_START set to False in your .env file. +⚠️ If you use Redis as your memory, make sure to run Auto-GPT with the `WIPE_REDIS_ON_START` set to `False` in your `.env` file. -⚠️For other memory backend, we currently forcefully wipe the memory when starting Auto-GPT. To ingest data with those memory backend, you can call the data_ingestion.py script anytime during an Auto-GPT run. +⚠️For other memory backend, we currently forcefully wipe the memory when starting Auto-GPT. To ingest data with those memory backend, you can call the `data_ingestion.py` script anytime during an Auto-GPT run. Memories will be available to the AI immediately as they are ingested, even if ingested while Auto-GPT is running. -In the example above, the script initializes the memory, ingests all files within the seed_data directory into memory with an overlap between chunks of 200 and a maximum length of each chunk of 4000. -Note that you can also use the --file argument to ingest a single file into memory and that the script will only ingest files within the auto_gpt_workspace directory. +In the example above, the script initializes the memory, ingests all files within the `/seed_data` directory into memory with an overlap between chunks of 200 and a maximum length of each chunk of 4000. +Note that you can also use the `--file` argument to ingest a single file into memory and that the script will only ingest files within the `/auto_gpt_workspace` directory. -You can adjust the max_length and overlap parameters to fine-tune the way the docuents are presented to the AI when it "recall" that memory: +You can adjust the `max_length` and overlap parameters to fine-tune the way the docuents are presented to the AI when it "recall" that memory: - Adjusting the overlap value allows the AI to access more contextual information from each chunk when recalling information, but will result in more chunks being created and therefore increase memory backend usage and OpenAI API requests. -- Reducing the max_length value will create more chunks, which can save prompt tokens by allowing for more message history in the context, but will also increase the number of chunks. -- Increasing the max_length value will provide the AI with more contextual information from each chunk, reducing the number of chunks created and saving on OpenAI API requests. However, this may also use more prompt tokens and decrease the overall context available to the AI. +- Reducing the `max_length` value will create more chunks, which can save prompt tokens by allowing for more message history in the context, but will also increase the number of chunks. +- Increasing the `max_length` value will provide the AI with more contextual information from each chunk, reducing the number of chunks created and saving on OpenAI API requests. However, this may also use more prompt tokens and decrease the overall context available to the AI. ## πŸ’€ Continuous Mode ⚠️ -Run the AI **without** user authorisation, 100% automated. -Continuous mode is not recommended. -It is potentially dangerous and may cause your AI to run forever or carry out actions you would not usually authorise. +Run the AI **without** user authorization, 100% automated. +Continuous mode is NOT recommended. +It is potentially dangerous and may cause your AI to run forever or carry out actions you would not usually authorize. Use at your own risk. 1. Run the `autogpt` python module in your terminal: -``` +```bash python -m autogpt --speak --continuous - ``` 2. To exit the program, press Ctrl + C @@ -359,7 +395,7 @@ python -m autogpt --speak --continuous If you don't have access to the GPT4 api, this mode will allow you to use Auto-GPT! -``` +```bash python -m autogpt --speak --gpt3only ``` @@ -367,15 +403,20 @@ It is recommended to use a virtual machine for tasks that require high security ## πŸ–Ό Image Generation -By default, Auto-GPT uses DALL-e for image generation. To use Stable Diffusion, a [HuggingFace API Token](https://huggingface.co/settings/tokens) is required. +By default, Auto-GPT uses DALL-e for image generation. To use Stable Diffusion, a [Hugging Face API Token](https://huggingface.co/settings/tokens) is required. Once you have a token, set these variables in your `.env`: -``` +```bash IMAGE_PROVIDER=sd HUGGINGFACE_API_TOKEN="YOUR_HUGGINGFACE_API_TOKEN" ``` +## Selenium +```bash +sudo Xvfb :10 -ac -screen 0 1024x768x24 & DISPLAY=:10 +``` + ## ⚠️ Limitations This experiment aims to showcase the potential of GPT-4 but comes with some limitations: @@ -416,13 +457,13 @@ We look forward to connecting with you and hearing your thoughts, ideas, and exp To run tests, run the following command: -``` +```bash python -m unittest discover tests ``` To run tests and see coverage, run the following command: -``` +```bash coverage run -m unittest discover tests ``` @@ -432,7 +473,7 @@ This project uses [flake8](https://flake8.pycqa.org/en/latest/) for linting. We To run the linter, run the following command: -``` +```bash flake8 autogpt/ tests/ # Or, if you want to run flake8 with the same configuration as the CI: diff --git a/autogpt/__main__.py b/autogpt/__main__.py index e7ac6810..29ccddbf 100644 --- a/autogpt/__main__.py +++ b/autogpt/__main__.py @@ -1,371 +1,21 @@ -import argparse -import json +"""Main script for the autogpt package.""" import logging -import traceback +from colorama import Fore +from autogpt.agent.agent import Agent +from autogpt.args import parse_arguments -from colorama import Fore, Style +from autogpt.config import Config, check_openai_api_key +from autogpt.logs import logger +from autogpt.memory import get_memory -from autogpt import chat -from autogpt import commands as cmd -from autogpt import speak, utils -from autogpt.ai_config import AIConfig -from autogpt.config import Config -from autogpt.json_parser import fix_and_parse_json -from autogpt.logger import logger -from autogpt.memory import get_memory, get_supported_memory_backends -from autogpt.spinner import Spinner +from autogpt.prompt import construct_prompt -cfg = Config() -config = None +# Load environment variables from .env file -def check_openai_api_key(): - """Check if the OpenAI API key is set in config.py or as an environment variable.""" - if not cfg.openai_api_key: - print( - Fore.RED - + "Please set your OpenAI API key in .env or as an environment variable." - ) - print("You can get your key from https://beta.openai.com/account/api-keys") - exit(1) - - -def attempt_to_fix_json_by_finding_outermost_brackets(json_string): - if cfg.speak_mode and cfg.debug_mode: - speak.say_text( - "I have received an invalid JSON response from the OpenAI API. " - "Trying to fix it now." - ) - logger.typewriter_log("Attempting to fix JSON by finding outermost brackets\n") - - try: - # Use regex to search for JSON objects - import regex - - json_pattern = regex.compile(r"\{(?:[^{}]|(?R))*\}") - json_match = json_pattern.search(json_string) - - if json_match: - # Extract the valid JSON object from the string - json_string = json_match.group(0) - logger.typewriter_log( - title="Apparently json was fixed.", title_color=Fore.GREEN - ) - if cfg.speak_mode and cfg.debug_mode: - speak.say_text("Apparently json was fixed.") - else: - raise ValueError("No valid JSON object found") - - except (json.JSONDecodeError, ValueError) as e: - if cfg.debug_mode: - logger.error("Error: Invalid JSON: %s\n", json_string) - if cfg.speak_mode: - speak.say_text("Didn't work. I will have to ignore this response then.") - logger.error("Error: Invalid JSON, setting it to empty JSON now.\n") - json_string = {} - - return json_string - - -def print_assistant_thoughts(assistant_reply): - """Prints the assistant's thoughts to the console""" - global ai_name - global cfg - try: - try: - # Parse and print Assistant response - assistant_reply_json = fix_and_parse_json(assistant_reply) - except json.JSONDecodeError as e: - logger.error("Error: Invalid JSON in assistant thoughts\n", assistant_reply) - assistant_reply_json = attempt_to_fix_json_by_finding_outermost_brackets( - assistant_reply - ) - assistant_reply_json = fix_and_parse_json(assistant_reply_json) - - # Check if assistant_reply_json is a string and attempt to parse it into a - # JSON object - if isinstance(assistant_reply_json, str): - try: - assistant_reply_json = json.loads(assistant_reply_json) - except json.JSONDecodeError as e: - logger.error("Error: Invalid JSON\n", assistant_reply) - assistant_reply_json = ( - attempt_to_fix_json_by_finding_outermost_brackets( - assistant_reply_json - ) - ) - - assistant_thoughts_reasoning = None - assistant_thoughts_plan = None - assistant_thoughts_speak = None - assistant_thoughts_criticism = None - assistant_thoughts = assistant_reply_json.get("thoughts", {}) - assistant_thoughts_text = assistant_thoughts.get("text") - - if assistant_thoughts: - assistant_thoughts_reasoning = assistant_thoughts.get("reasoning") - assistant_thoughts_plan = assistant_thoughts.get("plan") - assistant_thoughts_criticism = assistant_thoughts.get("criticism") - assistant_thoughts_speak = assistant_thoughts.get("speak") - - logger.typewriter_log( - f"{ai_name.upper()} THOUGHTS:", Fore.YELLOW, f"{assistant_thoughts_text}" - ) - logger.typewriter_log( - "REASONING:", Fore.YELLOW, f"{assistant_thoughts_reasoning}" - ) - - if assistant_thoughts_plan: - logger.typewriter_log("PLAN:", Fore.YELLOW, "") - # If it's a list, join it into a string - if isinstance(assistant_thoughts_plan, list): - assistant_thoughts_plan = "\n".join(assistant_thoughts_plan) - elif isinstance(assistant_thoughts_plan, dict): - assistant_thoughts_plan = str(assistant_thoughts_plan) - - # Split the input_string using the newline character and dashes - lines = assistant_thoughts_plan.split("\n") - for line in lines: - line = line.lstrip("- ") - logger.typewriter_log("- ", Fore.GREEN, line.strip()) - - logger.typewriter_log( - "CRITICISM:", Fore.YELLOW, f"{assistant_thoughts_criticism}" - ) - # Speak the assistant's thoughts - if cfg.speak_mode and assistant_thoughts_speak: - speak.say_text(assistant_thoughts_speak) - - return assistant_reply_json - except json.decoder.JSONDecodeError: - call_stack = traceback.format_exc() - logger.error("Error: Invalid JSON\n", assistant_reply) - logger.error("Traceback: \n", call_stack) - if cfg.speak_mode: - speak.say_text( - "I have received an invalid JSON response from the OpenAI API." - " I cannot ignore this response." - ) - - # All other errors, return "Error: + error message" - except Exception: - call_stack = traceback.format_exc() - logger.error("Error: \n", call_stack) - - -def construct_prompt(): - """Construct the prompt for the AI to respond to""" - config: AIConfig = AIConfig.load(cfg.ai_settings_file) - if cfg.skip_reprompt and config.ai_name: - logger.typewriter_log("Name :", Fore.GREEN, config.ai_name) - logger.typewriter_log("Role :", Fore.GREEN, config.ai_role) - logger.typewriter_log("Goals:", Fore.GREEN, f"{config.ai_goals}") - elif config.ai_name: - logger.typewriter_log( - "Welcome back! ", - Fore.GREEN, - f"Would you like me to return to being {config.ai_name}?", - speak_text=True, - ) - should_continue = utils.clean_input( - f"""Continue with the last settings? -Name: {config.ai_name} -Role: {config.ai_role} -Goals: {config.ai_goals} -Continue (y/n): """ - ) - if should_continue.lower() == "n": - config = AIConfig() - - if not config.ai_name: - config = prompt_user() - config.save() - - # Get rid of this global: - global ai_name - ai_name = config.ai_name - - return config.construct_full_prompt() - - -def prompt_user(): - """Prompt the user for input""" - ai_name = "" - # Construct the prompt - logger.typewriter_log( - "Welcome to Auto-GPT! ", - Fore.GREEN, - "Enter the name of your AI and its role below. Entering nothing will load" - " defaults.", - speak_text=True, - ) - - # Get AI Name from User - logger.typewriter_log( - "Name your AI: ", Fore.GREEN, "For example, 'Entrepreneur-GPT'" - ) - ai_name = utils.clean_input("AI Name: ") - if ai_name == "": - ai_name = "Entrepreneur-GPT" - - logger.typewriter_log( - f"{ai_name} here!", Fore.LIGHTBLUE_EX, "I am at your service.", speak_text=True - ) - - # Get AI Role from User - logger.typewriter_log( - "Describe your AI's role: ", - Fore.GREEN, - "For example, 'an AI designed to autonomously develop and run businesses with" - " the sole goal of increasing your net worth.'", - ) - ai_role = utils.clean_input(f"{ai_name} is: ") - if ai_role == "": - ai_role = "an AI designed to autonomously develop and run businesses with the" - " sole goal of increasing your net worth." - - # Enter up to 5 goals for the AI - logger.typewriter_log( - "Enter up to 5 goals for your AI: ", - Fore.GREEN, - "For example: \nIncrease net worth, Grow Twitter Account, Develop and manage" - " multiple businesses autonomously'", - ) - print("Enter nothing to load defaults, enter nothing when finished.", flush=True) - ai_goals = [] - for i in range(5): - ai_goal = utils.clean_input(f"{Fore.LIGHTBLUE_EX}Goal{Style.RESET_ALL} {i+1}: ") - if ai_goal == "": - break - ai_goals.append(ai_goal) - if len(ai_goals) == 0: - ai_goals = [ - "Increase net worth", - "Grow Twitter Account", - "Develop and manage multiple businesses autonomously", - ] - - config = AIConfig(ai_name, ai_role, ai_goals) - return config - - -def parse_arguments(): - """Parses the arguments passed to the script""" - global cfg - cfg.set_debug_mode(False) - cfg.set_continuous_mode(False) - cfg.set_speak_mode(False) - - parser = argparse.ArgumentParser(description="Process arguments.") - parser.add_argument( - "--continuous", "-c", action="store_true", help="Enable Continuous Mode" - ) - parser.add_argument( - "--continuous-limit", - "-l", - type=int, - dest="continuous_limit", - help="Defines the number of times to run in continuous mode", - ) - parser.add_argument("--speak", action="store_true", help="Enable Speak Mode") - parser.add_argument("--debug", action="store_true", help="Enable Debug Mode") - parser.add_argument( - "--gpt3only", action="store_true", help="Enable GPT3.5 Only Mode" - ) - parser.add_argument("--gpt4only", action="store_true", help="Enable GPT4 Only Mode") - parser.add_argument( - "--use-memory", - "-m", - dest="memory_type", - help="Defines which Memory backend to use", - ) - parser.add_argument( - "--skip-reprompt", - "-y", - dest="skip_reprompt", - action="store_true", - help="Skips the re-prompting messages at the beginning of the script", - ) - parser.add_argument( - "--ai-settings", - "-C", - dest="ai_settings_file", - help="Specifies which ai_settings.yaml file to use, will also automatically" - " skip the re-prompt.", - ) - args = parser.parse_args() - - if args.debug: - logger.typewriter_log("Debug Mode: ", Fore.GREEN, "ENABLED") - cfg.set_debug_mode(True) - - if args.continuous: - logger.typewriter_log("Continuous Mode: ", Fore.RED, "ENABLED") - logger.typewriter_log( - "WARNING: ", - Fore.RED, - "Continuous mode is not recommended. It is potentially dangerous and may" - " cause your AI to run forever or carry out actions you would not usually" - " authorise. Use at your own risk.", - ) - cfg.set_continuous_mode(True) - - if args.continuous_limit: - logger.typewriter_log( - "Continuous Limit: ", Fore.GREEN, f"{args.continuous_limit}" - ) - cfg.set_continuous_limit(args.continuous_limit) - - # Check if continuous limit is used without continuous mode - if args.continuous_limit and not args.continuous: - parser.error("--continuous-limit can only be used with --continuous") - - if args.speak: - logger.typewriter_log("Speak Mode: ", Fore.GREEN, "ENABLED") - cfg.set_speak_mode(True) - - if args.gpt3only: - logger.typewriter_log("GPT3.5 Only Mode: ", Fore.GREEN, "ENABLED") - cfg.set_smart_llm_model(cfg.fast_llm_model) - - if args.gpt4only: - logger.typewriter_log("GPT4 Only Mode: ", Fore.GREEN, "ENABLED") - cfg.set_fast_llm_model(cfg.smart_llm_model) - - if args.memory_type: - supported_memory = get_supported_memory_backends() - chosen = args.memory_type - if not chosen in supported_memory: - logger.typewriter_log( - "ONLY THE FOLLOWING MEMORY BACKENDS ARE SUPPORTED: ", - Fore.RED, - f"{supported_memory}", - ) - logger.typewriter_log("Defaulting to: ", Fore.YELLOW, cfg.memory_backend) - else: - cfg.memory_backend = chosen - - if args.skip_reprompt: - logger.typewriter_log("Skip Re-prompt: ", Fore.GREEN, "ENABLED") - cfg.skip_reprompt = True - - if args.ai_settings_file: - file = args.ai_settings_file - - # Validate file - (validated, message) = utils.validate_yaml_file(file) - if not validated: - logger.typewriter_log("FAILED FILE VALIDATION", Fore.RED, message) - logger.double_check() - exit(1) - - logger.typewriter_log("Using AI Settings File:", Fore.GREEN, file) - cfg.ai_settings_file = file - cfg.skip_reprompt = True - - -def main(): - global ai_name, memory +def main() -> None: + """Main function for the script""" + cfg = Config() # TODO: fill in llm values here check_openai_api_key() parse_arguments() @@ -384,7 +34,10 @@ def main(): # Initialize memory and make sure it is empty. # this is particularly important for indexing and referencing pinecone memory memory = get_memory(cfg, init=True) - print(f"Using memory of type: {memory.__class__.__name__}") + logger.typewriter_log( + f"Using memory of type:", Fore.GREEN, f"{memory.__class__.__name__}" + ) + logger.typewriter_log(f"Using Browser:", Fore.GREEN, cfg.selenium_web_browser) agent = Agent( ai_name=ai_name, memory=memory, @@ -396,177 +49,5 @@ def main(): agent.start_interaction_loop() -class Agent: - """Agent class for interacting with Auto-GPT. - - Attributes: - ai_name: The name of the agent. - memory: The memory object to use. - full_message_history: The full message history. - next_action_count: The number of actions to execute. - prompt: The prompt to use. - user_input: The user input. - - """ - - def __init__( - self, - ai_name, - memory, - full_message_history, - next_action_count, - prompt, - user_input, - ): - self.ai_name = ai_name - self.memory = memory - self.full_message_history = full_message_history - self.next_action_count = next_action_count - self.prompt = prompt - self.user_input = user_input - - def start_interaction_loop(self): - # Interaction Loop - loop_count = 0 - command_name = None - arguments = None - while True: - # Discontinue if continuous limit is reached - loop_count += 1 - if ( - cfg.continuous_mode - and cfg.continuous_limit > 0 - and loop_count > cfg.continuous_limit - ): - logger.typewriter_log( - "Continuous Limit Reached: ", Fore.YELLOW, f"{cfg.continuous_limit}" - ) - break - - # Send message to AI, get response - with Spinner("Thinking... "): - assistant_reply = chat.chat_with_ai( - self.prompt, - self.user_input, - self.full_message_history, - self.memory, - cfg.fast_token_limit, - ) # TODO: This hardcodes the model to use GPT3.5. Make this an argument - - # Print Assistant thoughts - print_assistant_thoughts(assistant_reply) - - # Get command name and arguments - try: - command_name, arguments = cmd.get_command( - attempt_to_fix_json_by_finding_outermost_brackets(assistant_reply) - ) - if cfg.speak_mode: - speak.say_text(f"I want to execute {command_name}") - except Exception as e: - logger.error("Error: \n", str(e)) - - if not cfg.continuous_mode and self.next_action_count == 0: - ### GET USER AUTHORIZATION TO EXECUTE COMMAND ### - # Get key press: Prompt the user to press enter to continue or escape - # to exit - self.user_input = "" - logger.typewriter_log( - "NEXT ACTION: ", - Fore.CYAN, - f"COMMAND = {Fore.CYAN}{command_name}{Style.RESET_ALL}" - f" ARGUMENTS = {Fore.CYAN}{arguments}{Style.RESET_ALL}", - ) - print( - "Enter 'y' to authorise command, 'y -N' to run N continuous" - " commands, 'n' to exit program, or enter feedback for" - f" {self.ai_name}...", - flush=True, - ) - while True: - console_input = utils.clean_input( - Fore.MAGENTA + "Input:" + Style.RESET_ALL - ) - if console_input.lower().rstrip() == "y": - self.user_input = "GENERATE NEXT COMMAND JSON" - break - elif console_input.lower().startswith("y -"): - try: - self.next_action_count = abs( - int(console_input.split(" ")[1]) - ) - self.user_input = "GENERATE NEXT COMMAND JSON" - except ValueError: - print( - "Invalid input format. Please enter 'y -n' where n" - " is the number of continuous tasks." - ) - continue - break - elif console_input.lower() == "n": - self.user_input = "EXIT" - break - else: - self.user_input = console_input - command_name = "human_feedback" - break - - if self.user_input == "GENERATE NEXT COMMAND JSON": - logger.typewriter_log( - "-=-=-=-=-=-=-= COMMAND AUTHORISED BY USER -=-=-=-=-=-=-=", - Fore.MAGENTA, - "", - ) - elif self.user_input == "EXIT": - print("Exiting...", flush=True) - break - else: - # Print command - logger.typewriter_log( - "NEXT ACTION: ", - Fore.CYAN, - f"COMMAND = {Fore.CYAN}{command_name}{Style.RESET_ALL}" - f" ARGUMENTS = {Fore.CYAN}{arguments}{Style.RESET_ALL}", - ) - - # Execute command - if command_name is not None and command_name.lower().startswith("error"): - result = ( - f"Command {command_name} threw the following error: {arguments}" - ) - elif command_name == "human_feedback": - result = f"Human feedback: {self.user_input}" - else: - result = ( - f"Command {command_name} " - f"returned: {cmd.execute_command(command_name, arguments)}" - ) - if self.next_action_count > 0: - self.next_action_count -= 1 - - memory_to_add = ( - f"Assistant Reply: {assistant_reply} " - f"\nResult: {result} " - f"\nHuman Feedback: {self.user_input} " - ) - - self.memory.add(memory_to_add) - - # Check if there's a result from the command append it to the message - # history - if result is not None: - self.full_message_history.append( - chat.create_chat_message("system", result) - ) - logger.typewriter_log("SYSTEM: ", Fore.YELLOW, result) - else: - self.full_message_history.append( - chat.create_chat_message("system", "Unable to execute command") - ) - logger.typewriter_log( - "SYSTEM: ", Fore.YELLOW, "Unable to execute command" - ) - - if __name__ == "__main__": main() diff --git a/autogpt/agent/__init__.py b/autogpt/agent/__init__.py new file mode 100644 index 00000000..e928af22 --- /dev/null +++ b/autogpt/agent/__init__.py @@ -0,0 +1,4 @@ +from autogpt.agent.agent import Agent +from autogpt.agent.agent_manager import AgentManager + +__all__ = ["Agent", "AgentManager"] diff --git a/autogpt/agent.py b/autogpt/agent/agent.py similarity index 56% rename from autogpt/agent.py rename to autogpt/agent/agent.py index 56c13d17..301d3f02 100644 --- a/autogpt/agent.py +++ b/autogpt/agent/agent.py @@ -1,15 +1,13 @@ -import json -import regex -import traceback - from colorama import Fore, Style +from autogpt.app import execute_command, get_command from autogpt.chat import chat_with_ai, create_chat_message -import autogpt.commands as cmd from autogpt.config import Config -from autogpt.json_parser import fix_and_parse_json -from autogpt.logger import logger -from autogpt.speak import say_text +from autogpt.json_fixes.bracket_termination import ( + attempt_to_fix_json_by_finding_outermost_brackets, +) +from autogpt.logs import logger, print_assistant_thoughts +from autogpt.speech import say_text from autogpt.spinner import Spinner from autogpt.utils import clean_input @@ -77,7 +75,7 @@ class Agent: # Get command name and arguments try: - command_name, arguments = cmd.get_command( + command_name, arguments = get_command( attempt_to_fix_json_by_finding_outermost_brackets(assistant_reply) ) if cfg.speak_mode: @@ -158,7 +156,7 @@ class Agent: else: result = ( f"Command {command_name} returned: " - f"{cmd.execute_command(command_name, arguments)}" + f"{execute_command(command_name, arguments)}" ) if self.next_action_count > 0: self.next_action_count -= 1 @@ -183,122 +181,3 @@ class Agent: logger.typewriter_log( "SYSTEM: ", Fore.YELLOW, "Unable to execute command" ) - - -def attempt_to_fix_json_by_finding_outermost_brackets(json_string): - cfg = Config() - if cfg.speak_mode and cfg.debug_mode: - say_text( - "I have received an invalid JSON response from the OpenAI API. " - "Trying to fix it now." - ) - logger.typewriter_log("Attempting to fix JSON by finding outermost brackets\n") - - try: - json_pattern = regex.compile(r"\{(?:[^{}]|(?R))*\}") - json_match = json_pattern.search(json_string) - - if json_match: - # Extract the valid JSON object from the string - json_string = json_match.group(0) - logger.typewriter_log( - title="Apparently json was fixed.", title_color=Fore.GREEN - ) - if cfg.speak_mode and cfg.debug_mode: - say_text("Apparently json was fixed.") - else: - raise ValueError("No valid JSON object found") - - except (json.JSONDecodeError, ValueError): - if cfg.speak_mode: - say_text("Didn't work. I will have to ignore this response then.") - logger.error("Error: Invalid JSON, setting it to empty JSON now.\n") - json_string = {} - - return json_string - - -def print_assistant_thoughts(ai_name, assistant_reply): - """Prints the assistant's thoughts to the console""" - cfg = Config() - try: - try: - # Parse and print Assistant response - assistant_reply_json = fix_and_parse_json(assistant_reply) - except json.JSONDecodeError: - logger.error("Error: Invalid JSON in assistant thoughts\n", assistant_reply) - assistant_reply_json = attempt_to_fix_json_by_finding_outermost_brackets( - assistant_reply - ) - if isinstance(assistant_reply_json, str): - assistant_reply_json = fix_and_parse_json(assistant_reply_json) - - # Check if assistant_reply_json is a string and attempt to parse - # it into a JSON object - if isinstance(assistant_reply_json, str): - try: - assistant_reply_json = json.loads(assistant_reply_json) - except json.JSONDecodeError: - logger.error("Error: Invalid JSON\n", assistant_reply) - assistant_reply_json = ( - attempt_to_fix_json_by_finding_outermost_brackets( - assistant_reply_json - ) - ) - - assistant_thoughts_reasoning = None - assistant_thoughts_plan = None - assistant_thoughts_speak = None - assistant_thoughts_criticism = None - if not isinstance(assistant_reply_json, dict): - assistant_reply_json = {} - assistant_thoughts = assistant_reply_json.get("thoughts", {}) - assistant_thoughts_text = assistant_thoughts.get("text") - - if assistant_thoughts: - assistant_thoughts_reasoning = assistant_thoughts.get("reasoning") - assistant_thoughts_plan = assistant_thoughts.get("plan") - assistant_thoughts_criticism = assistant_thoughts.get("criticism") - assistant_thoughts_speak = assistant_thoughts.get("speak") - - logger.typewriter_log( - f"{ai_name.upper()} THOUGHTS:", Fore.YELLOW, f"{assistant_thoughts_text}" - ) - logger.typewriter_log( - "REASONING:", Fore.YELLOW, f"{assistant_thoughts_reasoning}" - ) - - if assistant_thoughts_plan: - logger.typewriter_log("PLAN:", Fore.YELLOW, "") - # If it's a list, join it into a string - if isinstance(assistant_thoughts_plan, list): - assistant_thoughts_plan = "\n".join(assistant_thoughts_plan) - elif isinstance(assistant_thoughts_plan, dict): - assistant_thoughts_plan = str(assistant_thoughts_plan) - - # Split the input_string using the newline character and dashes - lines = assistant_thoughts_plan.split("\n") - for line in lines: - line = line.lstrip("- ") - logger.typewriter_log("- ", Fore.GREEN, line.strip()) - - logger.typewriter_log( - "CRITICISM:", Fore.YELLOW, f"{assistant_thoughts_criticism}" - ) - # Speak the assistant's thoughts - if cfg.speak_mode and assistant_thoughts_speak: - say_text(assistant_thoughts_speak) - - return assistant_reply_json - except json.decoder.JSONDecodeError: - logger.error("Error: Invalid JSON\n", assistant_reply) - if cfg.speak_mode: - say_text( - "I have received an invalid JSON response from the OpenAI API." - " I cannot ignore this response." - ) - - # All other errors, return "Error: + error message" - except Exception: - call_stack = traceback.format_exc() - logger.error("Error: \n", call_stack) diff --git a/autogpt/agent/agent_manager.py b/autogpt/agent/agent_manager.py new file mode 100644 index 00000000..3467f8bf --- /dev/null +++ b/autogpt/agent/agent_manager.py @@ -0,0 +1,100 @@ +"""Agent manager for managing GPT agents""" +from typing import List, Tuple, Union +from autogpt.llm_utils import create_chat_completion +from autogpt.config.config import Singleton + + +class AgentManager(metaclass=Singleton): + """Agent manager for managing GPT agents""" + + def __init__(self): + self.next_key = 0 + self.agents = {} # key, (task, full_message_history, model) + + # Create new GPT agent + # TODO: Centralise use of create_chat_completion() to globally enforce token limit + + def create_agent(self, task: str, prompt: str, model: str) -> Tuple[int, str]: + """Create a new agent and return its key + + Args: + task: The task to perform + prompt: The prompt to use + model: The model to use + + Returns: + The key of the new agent + """ + messages = [ + {"role": "user", "content": prompt}, + ] + + # Start GPT instance + agent_reply = create_chat_completion( + model=model, + messages=messages, + ) + + # Update full message history + messages.append({"role": "assistant", "content": agent_reply}) + + key = self.next_key + # This is done instead of len(agents) to make keys unique even if agents + # are deleted + self.next_key += 1 + + self.agents[key] = (task, messages, model) + + return key, agent_reply + + def message_agent(self, key: Union[str, int], message: str) -> str: + """Send a message to an agent and return its response + + Args: + key: The key of the agent to message + message: The message to send to the agent + + Returns: + The agent's response + """ + task, messages, model = self.agents[int(key)] + + # Add user message to message history before sending to agent + messages.append({"role": "user", "content": message}) + + # Start GPT instance + agent_reply = create_chat_completion( + model=model, + messages=messages, + ) + + # Update full message history + messages.append({"role": "assistant", "content": agent_reply}) + + return agent_reply + + def list_agents(self) -> List[Tuple[Union[str, int], str]]: + """Return a list of all agents + + Returns: + A list of tuples of the form (key, task) + """ + + # Return a list of agent keys and their tasks + return [(key, task) for key, (task, _, _) in self.agents.items()] + + def delete_agent(self, key: Union[str, int]) -> bool: + """Delete an agent from the agent manager + + Args: + key: The key of the agent to delete + + Returns: + True if successful, False otherwise + """ + + try: + del self.agents[int(key)] + return True + except KeyError: + return False diff --git a/autogpt/agent_manager.py b/autogpt/agent_manager.py deleted file mode 100644 index bbf156ee..00000000 --- a/autogpt/agent_manager.py +++ /dev/null @@ -1,75 +0,0 @@ -from autogpt.llm_utils import create_chat_completion - -next_key = 0 -agents = {} # key, (task, full_message_history, model) - -# Create new GPT agent -# TODO: Centralise use of create_chat_completion() to globally enforce token limit - - -def create_agent(task, prompt, model): - """Create a new agent and return its key""" - global next_key - global agents - - messages = [ - {"role": "user", "content": prompt}, - ] - - # Start GPT instance - agent_reply = create_chat_completion( - model=model, - messages=messages, - ) - - # Update full message history - messages.append({"role": "assistant", "content": agent_reply}) - - key = next_key - # This is done instead of len(agents) to make keys unique even if agents - # are deleted - next_key += 1 - - agents[key] = (task, messages, model) - - return key, agent_reply - - -def message_agent(key, message): - """Send a message to an agent and return its response""" - global agents - - task, messages, model = agents[int(key)] - - # Add user message to message history before sending to agent - messages.append({"role": "user", "content": message}) - - # Start GPT instance - agent_reply = create_chat_completion( - model=model, - messages=messages, - ) - - # Update full message history - messages.append({"role": "assistant", "content": agent_reply}) - - return agent_reply - - -def list_agents(): - """Return a list of all agents""" - global agents - - # Return a list of agent keys and their tasks - return [(key, task) for key, (task, _, _) in agents.items()] - - -def delete_agent(key): - """Delete an agent and return True if successful, False otherwise""" - global agents - - try: - del agents[int(key)] - return True - except KeyError: - return False diff --git a/autogpt/ai_functions.py b/autogpt/ai_functions.py deleted file mode 100644 index 84f82685..00000000 --- a/autogpt/ai_functions.py +++ /dev/null @@ -1,77 +0,0 @@ -import json -from typing import List - -from autogpt.call_ai_function import call_ai_function -from autogpt.config import Config - -cfg = Config() - - -def evaluate_code(code: str) -> List[str]: - """ - A function that takes in a string and returns a response from create chat - completion api call. - - Parameters: - code (str): Code to be evaluated. - Returns: - A result string from create chat completion. A list of suggestions to - improve the code. - """ - - function_string = "def analyze_code(code: str) -> List[str]:" - args = [code] - description_string = ( - "Analyzes the given code and returns a list of suggestions" " for improvements." - ) - - return call_ai_function(function_string, args, description_string) - - -def improve_code(suggestions: List[str], code: str) -> str: - """ - A function that takes in code and suggestions and returns a response from create - chat completion api call. - - Parameters: - suggestions (List): A list of suggestions around what needs to be improved. - code (str): Code to be improved. - Returns: - A result string from create chat completion. Improved code in response. - """ - - function_string = ( - "def generate_improved_code(suggestions: List[str], code: str) -> str:" - ) - args = [json.dumps(suggestions), code] - description_string = ( - "Improves the provided code based on the suggestions" - " provided, making no other changes." - ) - - return call_ai_function(function_string, args, description_string) - - -def write_tests(code: str, focus: List[str]) -> str: - """ - A function that takes in code and focus topics and returns a response from create - chat completion api call. - - Parameters: - focus (List): A list of suggestions around what needs to be improved. - code (str): Code for test cases to be generated against. - Returns: - A result string from create chat completion. Test cases for the submitted code - in response. - """ - - function_string = ( - "def create_test_cases(code: str, focus: Optional[str] = None) -> str:" - ) - args = [code, json.dumps(focus)] - description_string = ( - "Generates test cases for the existing code, focusing on" - " specific areas if required." - ) - - return call_ai_function(function_string, args, description_string) diff --git a/autogpt/commands.py b/autogpt/app.py similarity index 55% rename from autogpt/commands.py rename to autogpt/app.py index 352b3340..e84241c5 100644 --- a/autogpt/commands.py +++ b/autogpt/app.py @@ -1,29 +1,43 @@ +""" Command and Control """ import json -import datetime -import autogpt.agent_manager as agents +from typing import List, NoReturn, Union +from autogpt.agent.agent_manager import AgentManager +from autogpt.commands.evaluate_code import evaluate_code +from autogpt.commands.google_search import google_official_search, google_search +from autogpt.commands.improve_code import improve_code +from autogpt.commands.write_tests import write_tests from autogpt.config import Config -from autogpt.json_parser import fix_and_parse_json -from autogpt.image_gen import generate_image -from duckduckgo_search import ddg -from autogpt.ai_functions import evaluate_code, improve_code, write_tests -from autogpt.browse import scrape_links, scrape_text, summarize_text -from autogpt.execute_code import execute_python_file, execute_shell -from autogpt.file_operations import ( +from autogpt.commands.image_gen import generate_image +from autogpt.commands.web_requests import scrape_links, scrape_text +from autogpt.commands.execute_code import execute_python_file, execute_shell +from autogpt.commands.file_operations import ( append_to_file, delete_file, read_file, search_files, write_to_file, ) +from autogpt.json_fixes.parsing import fix_and_parse_json from autogpt.memory import get_memory -from autogpt.speak import say_text -from autogpt.web import browse_website +from autogpt.processing.text import summarize_text +from autogpt.speech import say_text +from autogpt.commands.web_selenium import browse_website +from autogpt.commands.git_operations import clone_repository -cfg = Config() +CFG = Config() +AGENT_MANAGER = AgentManager() -def is_valid_int(value) -> bool: +def is_valid_int(value: str) -> bool: + """Check if the value is a valid integer + + Args: + value (str): The value to check + + Returns: + bool: True if the value is a valid integer, False otherwise + """ try: int(value) return True @@ -31,8 +45,20 @@ def is_valid_int(value) -> bool: return False -def get_command(response): - """Parse the response and return the command name and arguments""" +def get_command(response: str): + """Parse the response and return the command name and arguments + + Args: + response (str): The response from the user + + Returns: + tuple: The command name and arguments + + Raises: + json.decoder.JSONDecodeError: If the response is not valid JSON + + Exception: If any other error occurs + """ try: response_json = fix_and_parse_json(response) @@ -62,20 +88,45 @@ def get_command(response): return "Error:", str(e) -def execute_command(command_name, arguments): - """Execute the command and return the result""" - memory = get_memory(cfg) +def map_command_synonyms(command_name: str): + """Takes the original command name given by the AI, and checks if the + string matches a list of common/known hallucinations + """ + synonyms = [ + ("write_file", "write_to_file"), + ("create_file", "write_to_file"), + ("search", "google"), + ] + for seen_command, actual_command_name in synonyms: + if command_name == seen_command: + return actual_command_name + return command_name + + +def execute_command(command_name: str, arguments): + """Execute the command and return the result + + Args: + command_name (str): The name of the command to execute + arguments (dict): The arguments for the command + + Returns: + str: The result of the command""" + memory = get_memory(CFG) try: + command_name = map_command_synonyms(command_name) if command_name == "google": # Check if the Google API key is set and use the official search method # If the API key is not set or has only whitespaces, use the unofficial # search method - key = cfg.google_api_key + key = CFG.google_api_key if key and key.strip() and key != "your-google-api-key": - return google_official_search(arguments["input"]) + google_result = google_official_search(arguments["input"]) else: - return google_search(arguments["input"]) + google_result = google_search(arguments["input"]) + safe_message = google_result.encode("utf-8", "ignore") + return str(safe_message) elif command_name == "memory_add": return memory.add(arguments["string"]) elif command_name == "start_agent": @@ -92,6 +143,10 @@ def execute_command(command_name, arguments): return get_text_summary(arguments["url"], arguments["question"]) elif command_name == "get_hyperlinks": return get_hyperlinks(arguments["url"]) + elif command_name == "clone_repository": + return clone_repository( + arguments["repository_url"], arguments["clone_path"] + ) elif command_name == "read_file": return read_file(arguments["file"]) elif command_name == "write_to_file": @@ -116,7 +171,7 @@ def execute_command(command_name, arguments): elif command_name == "execute_python_file": # Add this command return execute_python_file(arguments["file"]) elif command_name == "execute_shell": - if cfg.execute_local_commands: + if CFG.execute_local_commands: return execute_shell(arguments["command_line"]) else: return ( @@ -136,96 +191,55 @@ def execute_command(command_name, arguments): " list for available commands and only respond in the specified JSON" " format." ) - # All errors, return "Error: + error message" except Exception as e: - return "Error: " + str(e) + return f"Error: {str(e)}" -def get_datetime(): - """Return the current date and time""" - return "Current date and time: " + datetime.datetime.now().strftime( - "%Y-%m-%d %H:%M:%S" - ) +def get_text_summary(url: str, question: str) -> str: + """Return the results of a google search + Args: + url (str): The url to scrape + question (str): The question to summarize the text for -def google_search(query, num_results=8): - """Return the results of a google search""" - search_results = [] - if not query: - return json.dumps(search_results) - - for j in ddg(query, max_results=num_results): - search_results.append(j) - - return json.dumps(search_results, ensure_ascii=False, indent=4) - - -def google_official_search(query, num_results=8): - """Return the results of a google search using the official Google API""" - import json - - from googleapiclient.discovery import build - from googleapiclient.errors import HttpError - - try: - # Get the Google API key and Custom Search Engine ID from the config file - api_key = cfg.google_api_key - custom_search_engine_id = cfg.custom_search_engine_id - - # Initialize the Custom Search API service - service = build("customsearch", "v1", developerKey=api_key) - - # Send the search query and retrieve the results - result = ( - service.cse() - .list(q=query, cx=custom_search_engine_id, num=num_results) - .execute() - ) - - # Extract the search result items from the response - search_results = result.get("items", []) - - # Create a list of only the URLs from the search results - search_results_links = [item["link"] for item in search_results] - - except HttpError as e: - # Handle errors in the API call - error_details = json.loads(e.content.decode()) - - # Check if the error is related to an invalid or missing API key - if error_details.get("error", {}).get( - "code" - ) == 403 and "invalid API key" in error_details.get("error", {}).get( - "message", "" - ): - return "Error: The provided Google API key is invalid or missing." - else: - return f"Error: {e}" - - # Return the list of search result URLs - return search_results_links - - -def get_text_summary(url, question): - """Return the results of a google search""" + Returns: + str: The summary of the text + """ text = scrape_text(url) summary = summarize_text(url, text, question) - return """ "Result" : """ + summary + return f""" "Result" : {summary}""" -def get_hyperlinks(url): - """Return the results of a google search""" +def get_hyperlinks(url: str) -> Union[str, List[str]]: + """Return the results of a google search + + Args: + url (str): The url to scrape + + Returns: + str or list: The hyperlinks on the page + """ return scrape_links(url) -def shutdown(): +def shutdown() -> NoReturn: """Shut down the program""" print("Shutting down...") quit() -def start_agent(name, task, prompt, model=cfg.fast_llm_model): - """Start an agent with a given name, task, and prompt""" +def start_agent(name: str, task: str, prompt: str, model=CFG.fast_llm_model) -> str: + """Start an agent with a given name, task, and prompt + + Args: + name (str): The name of the agent + task (str): The task of the agent + prompt (str): The prompt for the agent + model (str): The model to use for the agent + + Returns: + str: The response of the agent + """ # Remove underscores from name voice_name = name.replace("_", " ") @@ -233,42 +247,52 @@ def start_agent(name, task, prompt, model=cfg.fast_llm_model): agent_intro = f"{voice_name} here, Reporting for duty!" # Create agent - if cfg.speak_mode: + if CFG.speak_mode: say_text(agent_intro, 1) - key, ack = agents.create_agent(task, first_message, model) + key, ack = AGENT_MANAGER.create_agent(task, first_message, model) - if cfg.speak_mode: + if CFG.speak_mode: say_text(f"Hello {voice_name}. Your task is as follows. {task}.") # Assign task (prompt), get response - agent_response = agents.message_agent(key, prompt) + agent_response = AGENT_MANAGER.message_agent(key, prompt) return f"Agent {name} created with key {key}. First response: {agent_response}" -def message_agent(key, message): +def message_agent(key: str, message: str) -> str: """Message an agent with a given key and message""" # Check if the key is a valid integer if is_valid_int(key): - agent_response = agents.message_agent(int(key), message) - # Check if the key is a valid string - elif isinstance(key, str): - agent_response = agents.message_agent(key, message) + agent_response = AGENT_MANAGER.message_agent(int(key), message) else: - return "Invalid key, must be an integer or a string." + return "Invalid key, must be an integer." # Speak response - if cfg.speak_mode: + if CFG.speak_mode: say_text(agent_response, 1) return agent_response def list_agents(): - """List all agents""" - return list_agents() + """List all agents + + Returns: + str: A list of all agents + """ + return "List of agents:\n" + "\n".join( + [str(x[0]) + ": " + x[1] for x in AGENT_MANAGER.list_agents()] + ) -def delete_agent(key): - """Delete an agent with a given key""" - result = agents.delete_agent(key) +def delete_agent(key: str) -> str: + """Delete an agent with a given key + + Args: + key (str): The key of the agent to delete + + Returns: + str: A message indicating whether the agent was deleted or not + """ + result = AGENT_MANAGER.delete_agent(key) return f"Agent {key} deleted." if result else f"Agent {key} does not exist." diff --git a/autogpt/args.py b/autogpt/args.py new file mode 100644 index 00000000..eca32334 --- /dev/null +++ b/autogpt/args.py @@ -0,0 +1,137 @@ +"""This module contains the argument parsing logic for the script.""" +import argparse + +from colorama import Fore +from autogpt import utils +from autogpt.config import Config +from autogpt.logs import logger +from autogpt.memory import get_supported_memory_backends + +CFG = Config() + + +def parse_arguments() -> None: + """Parses the arguments passed to the script + + Returns: + None + """ + CFG.set_debug_mode(False) + CFG.set_continuous_mode(False) + CFG.set_speak_mode(False) + + parser = argparse.ArgumentParser(description="Process arguments.") + parser.add_argument( + "--continuous", "-c", action="store_true", help="Enable Continuous Mode" + ) + parser.add_argument( + "--continuous-limit", + "-l", + type=int, + dest="continuous_limit", + help="Defines the number of times to run in continuous mode", + ) + parser.add_argument("--speak", action="store_true", help="Enable Speak Mode") + parser.add_argument("--debug", action="store_true", help="Enable Debug Mode") + parser.add_argument( + "--gpt3only", action="store_true", help="Enable GPT3.5 Only Mode" + ) + parser.add_argument("--gpt4only", action="store_true", help="Enable GPT4 Only Mode") + parser.add_argument( + "--use-memory", + "-m", + dest="memory_type", + help="Defines which Memory backend to use", + ) + parser.add_argument( + "--skip-reprompt", + "-y", + dest="skip_reprompt", + action="store_true", + help="Skips the re-prompting messages at the beginning of the script", + ) + parser.add_argument( + "--use-browser", + "-b", + dest="browser_name", + help="Specifies which web-browser to use when using selenium to scrape the web.", + ) + parser.add_argument( + "--ai-settings", + "-C", + dest="ai_settings_file", + help="Specifies which ai_settings.yaml file to use, will also automatically" + " skip the re-prompt.", + ) + args = parser.parse_args() + + if args.debug: + logger.typewriter_log("Debug Mode: ", Fore.GREEN, "ENABLED") + CFG.set_debug_mode(True) + + if args.continuous: + logger.typewriter_log("Continuous Mode: ", Fore.RED, "ENABLED") + logger.typewriter_log( + "WARNING: ", + Fore.RED, + "Continuous mode is not recommended. It is potentially dangerous and may" + " cause your AI to run forever or carry out actions you would not usually" + " authorise. Use at your own risk.", + ) + CFG.set_continuous_mode(True) + + if args.continuous_limit: + logger.typewriter_log( + "Continuous Limit: ", Fore.GREEN, f"{args.continuous_limit}" + ) + CFG.set_continuous_limit(args.continuous_limit) + + # Check if continuous limit is used without continuous mode + if args.continuous_limit and not args.continuous: + parser.error("--continuous-limit can only be used with --continuous") + + if args.speak: + logger.typewriter_log("Speak Mode: ", Fore.GREEN, "ENABLED") + CFG.set_speak_mode(True) + + if args.gpt3only: + logger.typewriter_log("GPT3.5 Only Mode: ", Fore.GREEN, "ENABLED") + CFG.set_smart_llm_model(CFG.fast_llm_model) + + if args.gpt4only: + logger.typewriter_log("GPT4 Only Mode: ", Fore.GREEN, "ENABLED") + CFG.set_fast_llm_model(CFG.smart_llm_model) + + if args.memory_type: + supported_memory = get_supported_memory_backends() + chosen = args.memory_type + if chosen not in supported_memory: + logger.typewriter_log( + "ONLY THE FOLLOWING MEMORY BACKENDS ARE SUPPORTED: ", + Fore.RED, + f"{supported_memory}", + ) + logger.typewriter_log("Defaulting to: ", Fore.YELLOW, CFG.memory_backend) + else: + CFG.memory_backend = chosen + + if args.skip_reprompt: + logger.typewriter_log("Skip Re-prompt: ", Fore.GREEN, "ENABLED") + CFG.skip_reprompt = True + + if args.ai_settings_file: + file = args.ai_settings_file + + # Validate file + (validated, message) = utils.validate_yaml_file(file) + if not validated: + logger.typewriter_log("FAILED FILE VALIDATION", Fore.RED, message) + logger.double_check() + exit(1) + + logger.typewriter_log("Using AI Settings File:", Fore.GREEN, file) + CFG.ai_settings_file = file + CFG.skip_reprompt = True + + if args.browser_name: + CFG.selenium_web_browser = args.browser_name diff --git a/autogpt/browse.py b/autogpt/browse.py deleted file mode 100644 index 60d33744..00000000 --- a/autogpt/browse.py +++ /dev/null @@ -1,198 +0,0 @@ -from urllib.parse import urljoin, urlparse - -import requests -from bs4 import BeautifulSoup - -from autogpt.config import Config -from autogpt.llm_utils import create_chat_completion -from autogpt.memory import get_memory - -cfg = Config() -memory = get_memory(cfg) - -session = requests.Session() -session.headers.update({"User-Agent": cfg.user_agent}) - - -# Function to check if the URL is valid -def is_valid_url(url): - try: - result = urlparse(url) - return all([result.scheme, result.netloc]) - except ValueError: - return False - - -# Function to sanitize the URL -def sanitize_url(url): - return urljoin(url, urlparse(url).path) - - -# Define and check for local file address prefixes -def check_local_file_access(url): - local_prefixes = [ - "file:///", - "file://localhost", - "http://localhost", - "https://localhost", - ] - return any(url.startswith(prefix) for prefix in local_prefixes) - - -def get_response(url, timeout=10): - try: - # Restrict access to local files - if check_local_file_access(url): - raise ValueError("Access to local files is restricted") - - # Most basic check if the URL is valid: - if not url.startswith("http://") and not url.startswith("https://"): - raise ValueError("Invalid URL format") - - sanitized_url = sanitize_url(url) - - response = session.get(sanitized_url, timeout=timeout) - - # Check if the response contains an HTTP error - if response.status_code >= 400: - return None, "Error: HTTP " + str(response.status_code) + " error" - - return response, None - except ValueError as ve: - # Handle invalid URL format - return None, "Error: " + str(ve) - - except requests.exceptions.RequestException as re: - # Handle exceptions related to the HTTP request - # (e.g., connection errors, timeouts, etc.) - return None, "Error: " + str(re) - - -def scrape_text(url): - """Scrape text from a webpage""" - response, error_message = get_response(url) - if error_message: - return error_message - if not response: - return "Error: Could not get response" - - soup = BeautifulSoup(response.text, "html.parser") - - for script in soup(["script", "style"]): - script.extract() - - text = soup.get_text() - lines = (line.strip() for line in text.splitlines()) - chunks = (phrase.strip() for line in lines for phrase in line.split(" ")) - text = "\n".join(chunk for chunk in chunks if chunk) - - return text - - -def extract_hyperlinks(soup): - """Extract hyperlinks from a BeautifulSoup object""" - hyperlinks = [] - for link in soup.find_all("a", href=True): - hyperlinks.append((link.text, link["href"])) - return hyperlinks - - -def format_hyperlinks(hyperlinks): - """Format hyperlinks into a list of strings""" - formatted_links = [] - for link_text, link_url in hyperlinks: - formatted_links.append(f"{link_text} ({link_url})") - return formatted_links - - -def scrape_links(url): - """Scrape links from a webpage""" - response, error_message = get_response(url) - if error_message: - return error_message - if not response: - return "Error: Could not get response" - soup = BeautifulSoup(response.text, "html.parser") - - for script in soup(["script", "style"]): - script.extract() - - hyperlinks = extract_hyperlinks(soup) - - return format_hyperlinks(hyperlinks) - - -def split_text(text, max_length=cfg.browse_chunk_max_length): - """Split text into chunks of a maximum length""" - paragraphs = text.split("\n") - current_length = 0 - current_chunk = [] - - for paragraph in paragraphs: - if current_length + len(paragraph) + 1 <= max_length: - current_chunk.append(paragraph) - current_length += len(paragraph) + 1 - else: - yield "\n".join(current_chunk) - current_chunk = [paragraph] - current_length = len(paragraph) + 1 - - if current_chunk: - yield "\n".join(current_chunk) - - -def create_message(chunk, question): - """Create a message for the user to summarize a chunk of text""" - return { - "role": "user", - "content": f'"""{chunk}""" Using the above text, please answer the following' - f' question: "{question}" -- if the question cannot be answered using the' - " text, please summarize the text.", - } - - -def summarize_text(url, text, question): - """Summarize text using the LLM model""" - if not text: - return "Error: No text to summarize" - - text_length = len(text) - print(f"Text length: {text_length} characters") - - summaries = [] - chunks = list(split_text(text)) - - for i, chunk in enumerate(chunks): - print(f"Adding chunk {i + 1} / {len(chunks)} to memory") - - memory_to_add = f"Source: {url}\n" f"Raw content part#{i + 1}: {chunk}" - - memory.add(memory_to_add) - - print(f"Summarizing chunk {i + 1} / {len(chunks)}") - messages = [create_message(chunk, question)] - - summary = create_chat_completion( - model=cfg.fast_llm_model, - messages=messages, - max_tokens=cfg.browse_summary_max_token, - ) - summaries.append(summary) - print(f"Added chunk {i + 1} summary to memory") - - memory_to_add = f"Source: {url}\n" f"Content summary part#{i + 1}: {summary}" - - memory.add(memory_to_add) - - print(f"Summarized {len(chunks)} chunks.") - - combined_summary = "\n".join(summaries) - messages = [create_message(combined_summary, question)] - - final_summary = create_chat_completion( - model=cfg.fast_llm_model, - messages=messages, - max_tokens=cfg.browse_summary_max_token, - ) - - return final_summary diff --git a/autogpt/call_ai_function.py b/autogpt/call_ai_function.py deleted file mode 100644 index bbbd8e35..00000000 --- a/autogpt/call_ai_function.py +++ /dev/null @@ -1,26 +0,0 @@ -from autogpt.config import Config -from autogpt.llm_utils import create_chat_completion - -cfg = Config() - - -# This is a magic function that can do anything with no-code. See -# https://github.com/Torantulino/AI-Functions for more info. -def call_ai_function(function, args, description, model=None) -> str: - """Call an AI function""" - if model is None: - model = cfg.smart_llm_model - # For each arg, if any are None, convert to "None": - args = [str(arg) if arg is not None else "None" for arg in args] - # parse args to comma separated string - args = ", ".join(args) - messages = [ - { - "role": "system", - "content": f"You are now the following python function: ```# {description}" - f"\n{function}```\n\nOnly respond with your `return` value.", - }, - {"role": "user", "content": args}, - ] - - return create_chat_completion(model=model, messages=messages, temperature=0) diff --git a/autogpt/chat.py b/autogpt/chat.py index 12797d9f..b0886967 100644 --- a/autogpt/chat.py +++ b/autogpt/chat.py @@ -5,7 +5,7 @@ from openai.error import RateLimitError from autogpt import token_counter from autogpt.config import Config from autogpt.llm_utils import create_chat_completion -from autogpt.logger import logger +from autogpt.logs import logger cfg = Config() diff --git a/autogpt/commands/__init__.py b/autogpt/commands/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/autogpt/commands/evaluate_code.py b/autogpt/commands/evaluate_code.py new file mode 100644 index 00000000..a36952e5 --- /dev/null +++ b/autogpt/commands/evaluate_code.py @@ -0,0 +1,25 @@ +"""Code evaluation module.""" +from typing import List + +from autogpt.llm_utils import call_ai_function + + +def evaluate_code(code: str) -> List[str]: + """ + A function that takes in a string and returns a response from create chat + completion api call. + + Parameters: + code (str): Code to be evaluated. + Returns: + A result string from create chat completion. A list of suggestions to + improve the code. + """ + + function_string = "def analyze_code(code: str) -> List[str]:" + args = [code] + description_string = ( + "Analyzes the given code and returns a list of suggestions" " for improvements." + ) + + return call_ai_function(function_string, args, description_string) diff --git a/autogpt/commands/execute_code.py b/autogpt/commands/execute_code.py new file mode 100644 index 00000000..86d6c177 --- /dev/null +++ b/autogpt/commands/execute_code.py @@ -0,0 +1,125 @@ +"""Execute code in a Docker container""" +import os +from pathlib import Path +import subprocess + +import docker +from docker.errors import ImageNotFound + +WORKING_DIRECTORY = Path(__file__).parent.parent / "auto_gpt_workspace" + + +def execute_python_file(file: str): + """Execute a Python file in a Docker container and return the output + + Args: + file (str): The name of the file to execute + + Returns: + str: The output of the file + """ + + print(f"Executing file '{file}' in workspace '{WORKING_DIRECTORY}'") + + if not file.endswith(".py"): + return "Error: Invalid file type. Only .py files are allowed." + + file_path = os.path.join(WORKING_DIRECTORY, file) + + if not os.path.isfile(file_path): + return f"Error: File '{file}' does not exist." + + if we_are_running_in_a_docker_container(): + result = subprocess.run( + f"python {file_path}", capture_output=True, encoding="utf8", shell=True + ) + if result.returncode == 0: + return result.stdout + else: + return f"Error: {result.stderr}" + + try: + client = docker.from_env() + + image_name = "python:3.10" + try: + client.images.get(image_name) + print(f"Image '{image_name}' found locally") + except ImageNotFound: + print(f"Image '{image_name}' not found locally, pulling from Docker Hub") + # Use the low-level API to stream the pull response + low_level_client = docker.APIClient() + for line in low_level_client.pull(image_name, stream=True, decode=True): + # Print the status and progress, if available + status = line.get("status") + progress = line.get("progress") + if status and progress: + print(f"{status}: {progress}") + elif status: + print(status) + + # You can replace 'python:3.8' with the desired Python image/version + # You can find available Python images on Docker Hub: + # https://hub.docker.com/_/python + container = client.containers.run( + image_name, + f"python {file}", + volumes={ + os.path.abspath(WORKING_DIRECTORY): { + "bind": "/workspace", + "mode": "ro", + } + }, + working_dir="/workspace", + stderr=True, + stdout=True, + detach=True, + ) + + container.wait() + logs = container.logs().decode("utf-8") + container.remove() + + # print(f"Execution complete. Output: {output}") + # print(f"Logs: {logs}") + + return logs + + except Exception as e: + return f"Error: {str(e)}" + + +def execute_shell(command_line: str) -> str: + """Execute a shell command and return the output + + Args: + command_line (str): The command line to execute + + Returns: + str: The output of the command + """ + current_dir = os.getcwd() + # Change dir into workspace if necessary + if str(WORKING_DIRECTORY) not in current_dir: + work_dir = os.path.join(os.getcwd(), WORKING_DIRECTORY) + os.chdir(work_dir) + + print(f"Executing command '{command_line}' in working directory '{os.getcwd()}'") + + result = subprocess.run(command_line, capture_output=True, shell=True) + output = f"STDOUT:\n{result.stdout}\nSTDERR:\n{result.stderr}" + + # Change back to whatever the prior working dir was + + os.chdir(current_dir) + + return output + + +def we_are_running_in_a_docker_container() -> bool: + """Check if we are running in a Docker container + + Returns: + bool: True if we are running in a Docker container, False otherwise + """ + return os.path.exists("/.dockerenv") diff --git a/autogpt/file_operations.py b/autogpt/commands/file_operations.py similarity index 57% rename from autogpt/file_operations.py rename to autogpt/commands/file_operations.py index 1c8f9290..47a04dce 100644 --- a/autogpt/file_operations.py +++ b/autogpt/commands/file_operations.py @@ -1,16 +1,29 @@ +"""File operations for AutoGPT""" import os import os.path +from pathlib import Path +from typing import Generator, List # Set a dedicated folder for file I/O -working_directory = "auto_gpt_workspace" +WORKING_DIRECTORY = Path(__file__).parent.parent / "auto_gpt_workspace" # Create the directory if it doesn't exist -if not os.path.exists(working_directory): - os.makedirs(working_directory) +if not os.path.exists(WORKING_DIRECTORY): + os.makedirs(WORKING_DIRECTORY) + +WORKING_DIRECTORY = str(WORKING_DIRECTORY) -def safe_join(base, *paths): - """Join one or more path components intelligently.""" +def safe_join(base: str, *paths) -> str: + """Join one or more path components intelligently. + + Args: + base (str): The base path + *paths (str): The paths to join to the base path + + Returns: + str: The joined path + """ new_path = os.path.join(base, *paths) norm_new_path = os.path.normpath(new_path) @@ -20,12 +33,14 @@ def safe_join(base, *paths): return norm_new_path -def split_file(content, max_length=4000, overlap=0): +def split_file( + content: str, max_length: int = 4000, overlap: int = 0 +) -> Generator[str, None, None]: """ Split text into chunks of a specified maximum length with a specified overlap between chunks. - :param text: The input text to be split into chunks + :param content: The input text to be split into chunks :param max_length: The maximum length of each chunk, default is 4000 (about 1k token) :param overlap: The number of overlapping characters between chunks, @@ -45,10 +60,17 @@ def split_file(content, max_length=4000, overlap=0): start += max_length - overlap -def read_file(filename) -> str: - """Read a file and return the contents""" +def read_file(filename: str) -> str: + """Read a file and return the contents + + Args: + filename (str): The name of the file to read + + Returns: + str: The contents of the file + """ try: - filepath = safe_join(working_directory, filename) + filepath = safe_join(WORKING_DIRECTORY, filename) with open(filepath, "r", encoding="utf-8") as f: content = f.read() return content @@ -56,7 +78,9 @@ def read_file(filename) -> str: return f"Error: {str(e)}" -def ingest_file(filename, memory, max_length=4000, overlap=200): +def ingest_file( + filename: str, memory, max_length: int = 4000, overlap: int = 200 +) -> None: """ Ingest a file by reading its content, splitting it into chunks with a specified maximum length and overlap, and adding the chunks to the memory storage. @@ -88,10 +112,18 @@ def ingest_file(filename, memory, max_length=4000, overlap=200): print(f"Error while ingesting file '{filename}': {str(e)}") -def write_to_file(filename, text): - """Write text to a file""" +def write_to_file(filename: str, text: str) -> str: + """Write text to a file + + Args: + filename (str): The name of the file to write to + text (str): The text to write to the file + + Returns: + str: A message indicating success or failure + """ try: - filepath = safe_join(working_directory, filename) + filepath = safe_join(WORKING_DIRECTORY, filename) directory = os.path.dirname(filepath) if not os.path.exists(directory): os.makedirs(directory) @@ -99,43 +131,66 @@ def write_to_file(filename, text): f.write(text) return "File written to successfully." except Exception as e: - return "Error: " + str(e) + return f"Error: {str(e)}" -def append_to_file(filename, text): - """Append text to a file""" +def append_to_file(filename: str, text: str) -> str: + """Append text to a file + + Args: + filename (str): The name of the file to append to + text (str): The text to append to the file + + Returns: + str: A message indicating success or failure + """ try: - filepath = safe_join(working_directory, filename) + filepath = safe_join(WORKING_DIRECTORY, filename) with open(filepath, "a") as f: f.write(text) return "Text appended successfully." except Exception as e: - return "Error: " + str(e) + return f"Error: {str(e)}" -def delete_file(filename): - """Delete a file""" +def delete_file(filename: str) -> str: + """Delete a file + + Args: + filename (str): The name of the file to delete + + Returns: + str: A message indicating success or failure + """ try: - filepath = safe_join(working_directory, filename) + filepath = safe_join(WORKING_DIRECTORY, filename) os.remove(filepath) return "File deleted successfully." except Exception as e: - return "Error: " + str(e) + return f"Error: {str(e)}" -def search_files(directory): +def search_files(directory: str) -> List[str]: + """Search for files in a directory + + Args: + directory (str): The directory to search in + + Returns: + List[str]: A list of files found in the directory + """ found_files = [] - if directory == "" or directory == "/": - search_directory = working_directory + if directory in {"", "/"}: + search_directory = WORKING_DIRECTORY else: - search_directory = safe_join(working_directory, directory) + search_directory = safe_join(WORKING_DIRECTORY, directory) for root, _, files in os.walk(search_directory): for file in files: if file.startswith("."): continue - relative_path = os.path.relpath(os.path.join(root, file), working_directory) + relative_path = os.path.relpath(os.path.join(root, file), WORKING_DIRECTORY) found_files.append(relative_path) return found_files diff --git a/autogpt/commands/git_operations.py b/autogpt/commands/git_operations.py new file mode 100644 index 00000000..3483474b --- /dev/null +++ b/autogpt/commands/git_operations.py @@ -0,0 +1,20 @@ +"""Git operations for autogpt""" +import git +from autogpt.config import Config + +CFG = Config() + + +def clone_repository(repo_url: str, clone_path: str) -> str: + """Clone a github repository locally + + Args: + repo_url (str): The URL of the repository to clone + clone_path (str): The path to clone the repository to + + Returns: + str: The result of the clone operation""" + split_url = repo_url.split("//") + auth_repo_url = f"//{CFG.github_username}:{CFG.github_api_key}@".join(split_url) + git.Repo.clone_from(auth_repo_url, clone_path) + return f"""Cloned {repo_url} to {clone_path}""" diff --git a/autogpt/commands/google_search.py b/autogpt/commands/google_search.py new file mode 100644 index 00000000..6deb9b50 --- /dev/null +++ b/autogpt/commands/google_search.py @@ -0,0 +1,86 @@ +"""Google search command for Autogpt.""" +import json +from typing import List, Union + +from duckduckgo_search import ddg + +from autogpt.config import Config + +CFG = Config() + + +def google_search(query: str, num_results: int = 8) -> str: + """Return the results of a google search + + Args: + query (str): The search query. + num_results (int): The number of results to return. + + Returns: + str: The results of the search. + """ + search_results = [] + if not query: + return json.dumps(search_results) + + results = ddg(query, max_results=num_results) + if not results: + return json.dumps(search_results) + + for j in results: + search_results.append(j) + + return json.dumps(search_results, ensure_ascii=False, indent=4) + + +def google_official_search(query: str, num_results: int = 8) -> Union[str, List[str]]: + """Return the results of a google search using the official Google API + + Args: + query (str): The search query. + num_results (int): The number of results to return. + + Returns: + str: The results of the search. + """ + + from googleapiclient.discovery import build + from googleapiclient.errors import HttpError + + try: + # Get the Google API key and Custom Search Engine ID from the config file + api_key = CFG.google_api_key + custom_search_engine_id = CFG.custom_search_engine_id + + # Initialize the Custom Search API service + service = build("customsearch", "v1", developerKey=api_key) + + # Send the search query and retrieve the results + result = ( + service.cse() + .list(q=query, cx=custom_search_engine_id, num=num_results) + .execute() + ) + + # Extract the search result items from the response + search_results = result.get("items", []) + + # Create a list of only the URLs from the search results + search_results_links = [item["link"] for item in search_results] + + except HttpError as e: + # Handle errors in the API call + error_details = json.loads(e.content.decode()) + + # Check if the error is related to an invalid or missing API key + if error_details.get("error", {}).get( + "code" + ) == 403 and "invalid API key" in error_details.get("error", {}).get( + "message", "" + ): + return "Error: The provided Google API key is invalid or missing." + else: + return f"Error: {e}" + + # Return the list of search result URLs + return search_results_links diff --git a/autogpt/commands/image_gen.py b/autogpt/commands/image_gen.py new file mode 100644 index 00000000..39e08845 --- /dev/null +++ b/autogpt/commands/image_gen.py @@ -0,0 +1,99 @@ +""" Image Generation Module for AutoGPT.""" +import io +import os.path +import uuid +from base64 import b64decode + +import openai +import requests +from PIL import Image +from pathlib import Path +from autogpt.config import Config + +CFG = Config() + +WORKING_DIRECTORY = Path(__file__).parent.parent / "auto_gpt_workspace" + + +def generate_image(prompt: str) -> str: + """Generate an image from a prompt. + + Args: + prompt (str): The prompt to use + + Returns: + str: The filename of the image + """ + filename = f"{str(uuid.uuid4())}.jpg" + + # DALL-E + if CFG.image_provider == "dalle": + return generate_image_with_dalle(prompt, filename) + elif CFG.image_provider == "sd": + return generate_image_with_hf(prompt, filename) + else: + return "No Image Provider Set" + + +def generate_image_with_hf(prompt: str, filename: str) -> str: + """Generate an image with HuggingFace's API. + + Args: + prompt (str): The prompt to use + filename (str): The filename to save the image to + + Returns: + str: The filename of the image + """ + API_URL = ( + "https://api-inference.huggingface.co/models/CompVis/stable-diffusion-v1-4" + ) + if CFG.huggingface_api_token is None: + raise ValueError( + "You need to set your Hugging Face API token in the config file." + ) + headers = {"Authorization": f"Bearer {CFG.huggingface_api_token}"} + + response = requests.post( + API_URL, + headers=headers, + json={ + "inputs": prompt, + }, + ) + + image = Image.open(io.BytesIO(response.content)) + print(f"Image Generated for prompt:{prompt}") + + image.save(os.path.join(WORKING_DIRECTORY, filename)) + + return f"Saved to disk:{filename}" + + +def generate_image_with_dalle(prompt: str, filename: str) -> str: + """Generate an image with DALL-E. + + Args: + prompt (str): The prompt to use + filename (str): The filename to save the image to + + Returns: + str: The filename of the image + """ + openai.api_key = CFG.openai_api_key + + response = openai.Image.create( + prompt=prompt, + n=1, + size="256x256", + response_format="b64_json", + ) + + print(f"Image Generated for prompt:{prompt}") + + image_data = b64decode(response["data"][0]["b64_json"]) + + with open(f"{WORKING_DIRECTORY}/{filename}", mode="wb") as png: + png.write(image_data) + + return f"Saved to disk:{filename}" diff --git a/autogpt/commands/improve_code.py b/autogpt/commands/improve_code.py new file mode 100644 index 00000000..05fe89e9 --- /dev/null +++ b/autogpt/commands/improve_code.py @@ -0,0 +1,28 @@ +import json +from typing import List + +from autogpt.llm_utils import call_ai_function + + +def improve_code(suggestions: List[str], code: str) -> str: + """ + A function that takes in code and suggestions and returns a response from create + chat completion api call. + + Parameters: + suggestions (List): A list of suggestions around what needs to be improved. + code (str): Code to be improved. + Returns: + A result string from create chat completion. Improved code in response. + """ + + function_string = ( + "def generate_improved_code(suggestions: List[str], code: str) -> str:" + ) + args = [json.dumps(suggestions), code] + description_string = ( + "Improves the provided code based on the suggestions" + " provided, making no other changes." + ) + + return call_ai_function(function_string, args, description_string) diff --git a/autogpt/commands/times.py b/autogpt/commands/times.py new file mode 100644 index 00000000..3c9b8a4f --- /dev/null +++ b/autogpt/commands/times.py @@ -0,0 +1,10 @@ +from datetime import datetime + + +def get_datetime() -> str: + """Return the current date and time + + Returns: + str: The current date and time + """ + return "Current date and time: " + datetime.now().strftime("%Y-%m-%d %H:%M:%S") diff --git a/autogpt/commands/web_requests.py b/autogpt/commands/web_requests.py new file mode 100644 index 00000000..230b1ff0 --- /dev/null +++ b/autogpt/commands/web_requests.py @@ -0,0 +1,198 @@ +"""Browse a webpage and summarize it using the LLM model""" +from typing import List, Tuple, Union +from urllib.parse import urljoin, urlparse + +import requests +from requests import Response +from bs4 import BeautifulSoup + +from autogpt.config import Config +from autogpt.memory import get_memory + +CFG = Config() +memory = get_memory(CFG) + +session = requests.Session() +session.headers.update({"User-Agent": CFG.user_agent}) + + +def is_valid_url(url: str) -> bool: + """Check if the URL is valid + + Args: + url (str): The URL to check + + Returns: + bool: True if the URL is valid, False otherwise + """ + try: + result = urlparse(url) + return all([result.scheme, result.netloc]) + except ValueError: + return False + + +def sanitize_url(url: str) -> str: + """Sanitize the URL + + Args: + url (str): The URL to sanitize + + Returns: + str: The sanitized URL + """ + return urljoin(url, urlparse(url).path) + + +def check_local_file_access(url: str) -> bool: + """Check if the URL is a local file + + Args: + url (str): The URL to check + + Returns: + bool: True if the URL is a local file, False otherwise + """ + local_prefixes = [ + "file:///", + "file://localhost", + "http://localhost", + "https://localhost", + ] + return any(url.startswith(prefix) for prefix in local_prefixes) + + +def get_response( + url: str, timeout: int = 10 +) -> Union[Tuple[None, str], Tuple[Response, None]]: + """Get the response from a URL + + Args: + url (str): The URL to get the response from + timeout (int): The timeout for the HTTP request + + Returns: + Tuple[None, str] | Tuple[Response, None]: The response and error message + + Raises: + ValueError: If the URL is invalid + requests.exceptions.RequestException: If the HTTP request fails + """ + try: + # Restrict access to local files + if check_local_file_access(url): + raise ValueError("Access to local files is restricted") + + # Most basic check if the URL is valid: + if not url.startswith("http://") and not url.startswith("https://"): + raise ValueError("Invalid URL format") + + sanitized_url = sanitize_url(url) + + response = session.get(sanitized_url, timeout=timeout) + + # Check if the response contains an HTTP error + if response.status_code >= 400: + return None, f"Error: HTTP {str(response.status_code)} error" + + return response, None + except ValueError as ve: + # Handle invalid URL format + return None, f"Error: {str(ve)}" + + except requests.exceptions.RequestException as re: + # Handle exceptions related to the HTTP request + # (e.g., connection errors, timeouts, etc.) + return None, f"Error: {str(re)}" + + +def scrape_text(url: str) -> str: + """Scrape text from a webpage + + Args: + url (str): The URL to scrape text from + + Returns: + str: The scraped text + """ + response, error_message = get_response(url) + if error_message: + return error_message + if not response: + return "Error: Could not get response" + + soup = BeautifulSoup(response.text, "html.parser") + + for script in soup(["script", "style"]): + script.extract() + + text = soup.get_text() + lines = (line.strip() for line in text.splitlines()) + chunks = (phrase.strip() for line in lines for phrase in line.split(" ")) + text = "\n".join(chunk for chunk in chunks if chunk) + + return text + + +def extract_hyperlinks(soup: BeautifulSoup) -> List[Tuple[str, str]]: + """Extract hyperlinks from a BeautifulSoup object + + Args: + soup (BeautifulSoup): The BeautifulSoup object + + Returns: + List[Tuple[str, str]]: The extracted hyperlinks + """ + hyperlinks = [] + for link in soup.find_all("a", href=True): + hyperlinks.append((link.text, link["href"])) + return hyperlinks + + +def format_hyperlinks(hyperlinks: List[Tuple[str, str]]) -> List[str]: + """Format hyperlinks into a list of strings + + Args: + hyperlinks (List[Tuple[str, str]]): The hyperlinks to format + + Returns: + List[str]: The formatted hyperlinks + """ + formatted_links = [] + for link_text, link_url in hyperlinks: + formatted_links.append(f"{link_text} ({link_url})") + return formatted_links + + +def scrape_links(url: str) -> Union[str, List[str]]: + """Scrape links from a webpage + + Args: + url (str): The URL to scrape links from + + Returns: + Union[str, List[str]]: The scraped links + """ + response, error_message = get_response(url) + if error_message: + return error_message + if not response: + return "Error: Could not get response" + soup = BeautifulSoup(response.text, "html.parser") + + for script in soup(["script", "style"]): + script.extract() + + hyperlinks = extract_hyperlinks(soup) + + return format_hyperlinks(hyperlinks) + + +def create_message(chunk, question): + """Create a message for the user to summarize a chunk of text""" + return { + "role": "user", + "content": f'"""{chunk}""" Using the above text, answer the following' + f' question: "{question}" -- if the question cannot be answered using the' + " text, summarize the text.", + } diff --git a/autogpt/commands/web_selenium.py b/autogpt/commands/web_selenium.py new file mode 100644 index 00000000..d2ef1a7e --- /dev/null +++ b/autogpt/commands/web_selenium.py @@ -0,0 +1,164 @@ +"""Selenium web scraping module.""" +from selenium import webdriver +import autogpt.processing.text as summary +from bs4 import BeautifulSoup +from selenium.webdriver.remote.webdriver import WebDriver +from selenium.webdriver.common.by import By +from selenium.webdriver.support.wait import WebDriverWait +from selenium.webdriver.support import expected_conditions as EC +from webdriver_manager.chrome import ChromeDriverManager +from webdriver_manager.firefox import GeckoDriverManager +from selenium.webdriver.chrome.options import Options as ChromeOptions +from selenium.webdriver.firefox.options import Options as FirefoxOptions +from selenium.webdriver.safari.options import Options as SafariOptions +import logging +from pathlib import Path +from autogpt.config import Config +from typing import List, Tuple, Union + +FILE_DIR = Path(__file__).parent.parent +CFG = Config() + + +def browse_website(url: str, question: str) -> Tuple[str, WebDriver]: + """Browse a website and return the answer and links to the user + + Args: + url (str): The url of the website to browse + question (str): The question asked by the user + + Returns: + Tuple[str, WebDriver]: The answer and links to the user and the webdriver + """ + driver, text = scrape_text_with_selenium(url) + add_header(driver) + summary_text = summary.summarize_text(url, text, question, driver) + links = scrape_links_with_selenium(driver) + + # Limit links to 5 + if len(links) > 5: + links = links[:5] + close_browser(driver) + return f"Answer gathered from website: {summary_text} \n \n Links: {links}", driver + + +def scrape_text_with_selenium(url: str) -> Tuple[WebDriver, str]: + """Scrape text from a website using selenium + + Args: + url (str): The url of the website to scrape + + Returns: + Tuple[WebDriver, str]: The webdriver and the text scraped from the website + """ + logging.getLogger("selenium").setLevel(logging.CRITICAL) + + options_available = { + "chrome": ChromeOptions, + "safari": SafariOptions, + "firefox": FirefoxOptions, + } + + options = options_available[CFG.selenium_web_browser]() + options.add_argument( + "user-agent=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/112.0.5615.49 Safari/537.36" + ) + + if CFG.selenium_web_browser == "firefox": + driver = webdriver.Firefox( + executable_path=GeckoDriverManager().install(), options=options + ) + elif CFG.selenium_web_browser == "safari": + # Requires a bit more setup on the users end + # See https://developer.apple.com/documentation/webkit/testing_with_webdriver_in_safari + driver = webdriver.Safari(options=options) + else: + driver = webdriver.Chrome( + executable_path=ChromeDriverManager().install(), options=options + ) + driver.get(url) + + WebDriverWait(driver, 10).until( + EC.presence_of_element_located((By.TAG_NAME, "body")) + ) + + # Get the HTML content directly from the browser's DOM + page_source = driver.execute_script("return document.body.outerHTML;") + soup = BeautifulSoup(page_source, "html.parser") + + for script in soup(["script", "style"]): + script.extract() + + text = soup.get_text() + lines = (line.strip() for line in text.splitlines()) + chunks = (phrase.strip() for line in lines for phrase in line.split(" ")) + text = "\n".join(chunk for chunk in chunks if chunk) + return driver, text + + +def scrape_links_with_selenium(driver: WebDriver) -> List[str]: + """Scrape links from a website using selenium + + Args: + driver (WebDriver): The webdriver to use to scrape the links + + Returns: + List[str]: The links scraped from the website + """ + page_source = driver.page_source + soup = BeautifulSoup(page_source, "html.parser") + + for script in soup(["script", "style"]): + script.extract() + + hyperlinks = extract_hyperlinks(soup) + + return format_hyperlinks(hyperlinks) + + +def close_browser(driver: WebDriver) -> None: + """Close the browser + + Args: + driver (WebDriver): The webdriver to close + + Returns: + None + """ + driver.quit() + + +def extract_hyperlinks(soup: BeautifulSoup) -> List[Tuple[str, str]]: + """Extract hyperlinks from a BeautifulSoup object + + Args: + soup (BeautifulSoup): The BeautifulSoup object to extract the hyperlinks from + + Returns: + List[Tuple[str, str]]: The hyperlinks extracted from the BeautifulSoup object + """ + return [(link.text, link["href"]) for link in soup.find_all("a", href=True)] + + +def format_hyperlinks(hyperlinks: List[Tuple[str, str]]) -> List[str]: + """Format hyperlinks to be displayed to the user + + Args: + hyperlinks (List[Tuple[str, str]]): The hyperlinks to format + + Returns: + List[str]: The formatted hyperlinks + """ + return [f"{link_text} ({link_url})" for link_text, link_url in hyperlinks] + + +def add_header(driver: WebDriver) -> None: + """Add a header to the website + + Args: + driver (WebDriver): The webdriver to use to add the header + + Returns: + None + """ + driver.execute_script(open(f"{FILE_DIR}/js/overlay.js", "r").read()) diff --git a/autogpt/commands/write_tests.py b/autogpt/commands/write_tests.py new file mode 100644 index 00000000..f1d6c9b2 --- /dev/null +++ b/autogpt/commands/write_tests.py @@ -0,0 +1,29 @@ +"""A module that contains a function to generate test cases for the submitted code.""" +import json +from typing import List +from autogpt.llm_utils import call_ai_function + + +def write_tests(code: str, focus: List[str]) -> str: + """ + A function that takes in code and focus topics and returns a response from create + chat completion api call. + + Parameters: + focus (List): A list of suggestions around what needs to be improved. + code (str): Code for test cases to be generated against. + Returns: + A result string from create chat completion. Test cases for the submitted code + in response. + """ + + function_string = ( + "def create_test_cases(code: str, focus: Optional[str] = None) -> str:" + ) + args = [code, json.dumps(focus)] + description_string = ( + "Generates test cases for the existing code, focusing on" + " specific areas if required." + ) + + return call_ai_function(function_string, args, description_string) diff --git a/autogpt/config/__init__.py b/autogpt/config/__init__.py new file mode 100644 index 00000000..ceb5566c --- /dev/null +++ b/autogpt/config/__init__.py @@ -0,0 +1,14 @@ +""" +This module contains the configuration classes for AutoGPT. +""" +from autogpt.config.ai_config import AIConfig +from autogpt.config.config import check_openai_api_key, Config +from autogpt.config.singleton import AbstractSingleton, Singleton + +__all__ = [ + "check_openai_api_key", + "AbstractSingleton", + "AIConfig", + "Config", + "Singleton", +] diff --git a/autogpt/ai_config.py b/autogpt/config/ai_config.py similarity index 87% rename from autogpt/ai_config.py rename to autogpt/config/ai_config.py index 0957b886..c72b088b 100644 --- a/autogpt/ai_config.py +++ b/autogpt/config/ai_config.py @@ -1,9 +1,11 @@ +# sourcery skip: do-not-use-staticmethod +""" +A module that contains the AIConfig class object that contains the configuration +""" import os -from typing import Type +from typing import List, Optional, Type import yaml -from autogpt.prompt import get_prompt - class AIConfig: """ @@ -16,7 +18,7 @@ class AIConfig: """ def __init__( - self, ai_name: str = "", ai_role: str = "", ai_goals: list = [] + self, ai_name: str = "", ai_role: str = "", ai_goals: Optional[List] = None ) -> None: """ Initialize a class instance @@ -28,7 +30,8 @@ class AIConfig: Returns: None """ - + if ai_goals is None: + ai_goals = [] self.ai_name = ai_name self.ai_role = ai_role self.ai_goals = ai_goals @@ -36,15 +39,14 @@ class AIConfig: # Soon this will go in a folder where it remembers more stuff about the run(s) SAVE_FILE = os.path.join(os.path.dirname(__file__), "..", "ai_settings.yaml") - @classmethod - def load(cls: "Type[AIConfig]", config_file: str = SAVE_FILE) -> "Type[AIConfig]": + @staticmethod + def load(config_file: str = SAVE_FILE) -> "AIConfig": """ Returns class object with parameters (ai_name, ai_role, ai_goals) loaded from yaml file if yaml file exists, else returns class with no parameters. Parameters: - cls (class object): An AIConfig Class object. config_file (int): The path to the config yaml file. DEFAULT: "../ai_settings.yaml" @@ -62,7 +64,7 @@ class AIConfig: ai_role = config_params.get("ai_role", "") ai_goals = config_params.get("ai_goals", []) # type: Type[AIConfig] - return cls(ai_name, ai_role, ai_goals) + return AIConfig(ai_name, ai_role, ai_goals) def save(self, config_file: str = SAVE_FILE) -> None: """ @@ -103,6 +105,8 @@ class AIConfig: "" ) + from autogpt.prompt import get_prompt + # Construct full prompt full_prompt = ( f"You are {self.ai_name}, {self.ai_role}\n{prompt_start}\n\nGOALS:\n\n" diff --git a/autogpt/config.py b/autogpt/config/config.py similarity index 69% rename from autogpt/config.py rename to autogpt/config/config.py index 26132a5a..c47b19af 100644 --- a/autogpt/config.py +++ b/autogpt/config/config.py @@ -1,30 +1,15 @@ -import abc +"""Configuration class to store the state of bools for different scripts access.""" import os +from colorama import Fore + +from autogpt.config.singleton import Singleton import openai import yaml + from dotenv import load_dotenv -# Load environment variables from .env file -load_dotenv() - - -class Singleton(abc.ABCMeta, type): - """ - Singleton metaclass for ensuring only one instance of a class. - """ - - _instances = {} - - def __call__(cls, *args, **kwargs): - """Call method for the singleton metaclass.""" - if cls not in cls._instances: - cls._instances[cls] = super(Singleton, cls).__call__(*args, **kwargs) - return cls._instances[cls] - - -class AbstractSingleton(abc.ABC, metaclass=Singleton): - pass +load_dotenv(verbose=True) class Config(metaclass=Singleton): @@ -32,7 +17,7 @@ class Config(metaclass=Singleton): Configuration class to store the state of bools for different scripts access. """ - def __init__(self): + def __init__(self) -> None: """Initialize the Config class""" self.debug_mode = False self.continuous_mode = False @@ -40,6 +25,7 @@ class Config(metaclass=Singleton): self.speak_mode = False self.skip_reprompt = False + self.selenium_web_browser = os.getenv("USE_WEB_BROWSER", "chrome") self.ai_settings_file = os.getenv("AI_SETTINGS_FILE", "ai_settings.yaml") self.fast_llm_model = os.getenv("FAST_LLM_MODEL", "gpt-3.5-turbo") self.smart_llm_model = os.getenv("SMART_LLM_MODEL", "gpt-4") @@ -71,20 +57,29 @@ class Config(metaclass=Singleton): self.use_brian_tts = False self.use_brian_tts = os.getenv("USE_BRIAN_TTS") + self.github_api_key = os.getenv("GITHUB_API_KEY") + self.github_username = os.getenv("GITHUB_USERNAME") + self.google_api_key = os.getenv("GOOGLE_API_KEY") self.custom_search_engine_id = os.getenv("CUSTOM_SEARCH_ENGINE_ID") self.pinecone_api_key = os.getenv("PINECONE_API_KEY") self.pinecone_region = os.getenv("PINECONE_ENV") + # milvus configuration, e.g., localhost:19530. + self.milvus_addr = os.getenv("MILVUS_ADDR", "localhost:19530") + self.milvus_collection = os.getenv("MILVUS_COLLECTION", "autogpt") + self.image_provider = os.getenv("IMAGE_PROVIDER") self.huggingface_api_token = os.getenv("HUGGINGFACE_API_TOKEN") # User agent headers to use when browsing web - # Some websites might just completely deny request with an error code if no user agent was found. + # Some websites might just completely deny request with an error code if + # no user agent was found. self.user_agent = os.getenv( "USER_AGENT", - "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", + "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", ) self.redis_host = os.getenv("REDIS_HOST", "localhost") self.redis_port = os.getenv("REDIS_PORT", "6379") @@ -108,15 +103,17 @@ class Config(metaclass=Singleton): The matching deployment id if found, otherwise an empty string. """ if model == self.fast_llm_model: - return self.azure_model_to_deployment_id_map["fast_llm_model_deployment_id"] + return self.azure_model_to_deployment_id_map[ + "fast_llm_model_deployment_id" + ] # type: ignore elif model == self.smart_llm_model: return self.azure_model_to_deployment_id_map[ "smart_llm_model_deployment_id" - ] + ] # type: ignore elif model == "text-embedding-ada-002": return self.azure_model_to_deployment_id_map[ "embedding_model_deployment_id" - ] + ] # type: ignore else: return "" @@ -124,7 +121,8 @@ class Config(metaclass=Singleton): def load_azure_config(self, config_file: str = AZURE_CONFIG_FILE) -> None: """ - Loads the configuration parameters for Azure hosting from the specified file path as a yaml file. + Loads the configuration parameters for Azure hosting from the specified file + path as a yaml file. Parameters: config_file(str): The path to the config yaml file. DEFAULT: "../azure.yaml" @@ -137,85 +135,93 @@ class Config(metaclass=Singleton): config_params = yaml.load(file, Loader=yaml.FullLoader) except FileNotFoundError: config_params = {} - self.openai_api_type = os.getenv( - "OPENAI_API_TYPE", config_params.get("azure_api_type", "azure") - ) - self.openai_api_base = os.getenv( - "OPENAI_AZURE_API_BASE", config_params.get("azure_api_base", "") - ) - self.openai_api_version = os.getenv( - "OPENAI_AZURE_API_VERSION", config_params.get("azure_api_version", "") + self.openai_api_type = config_params.get("azure_api_type") or "azure" + self.openai_api_base = config_params.get("azure_api_base") or "" + self.openai_api_version = ( + config_params.get("azure_api_version") or "2023-03-15-preview" ) self.azure_model_to_deployment_id_map = config_params.get("azure_model_map", []) - def set_continuous_mode(self, value: bool): + def set_continuous_mode(self, value: bool) -> None: """Set the continuous mode value.""" self.continuous_mode = value - def set_continuous_limit(self, value: int): + def set_continuous_limit(self, value: int) -> None: """Set the continuous limit value.""" self.continuous_limit = value - def set_speak_mode(self, value: bool): + def set_speak_mode(self, value: bool) -> None: """Set the speak mode value.""" self.speak_mode = value - def set_fast_llm_model(self, value: str): + def set_fast_llm_model(self, value: str) -> None: """Set the fast LLM model value.""" self.fast_llm_model = value - def set_smart_llm_model(self, value: str): + def set_smart_llm_model(self, value: str) -> None: """Set the smart LLM model value.""" self.smart_llm_model = value - def set_fast_token_limit(self, value: int): + def set_fast_token_limit(self, value: int) -> None: """Set the fast token limit value.""" self.fast_token_limit = value - def set_smart_token_limit(self, value: int): + def set_smart_token_limit(self, value: int) -> None: """Set the smart token limit value.""" self.smart_token_limit = value - def set_browse_chunk_max_length(self, value: int): + def set_browse_chunk_max_length(self, value: int) -> None: """Set the browse_website command chunk max length value.""" self.browse_chunk_max_length = value - def set_browse_summary_max_token(self, value: int): + def set_browse_summary_max_token(self, value: int) -> None: """Set the browse_website command summary max token value.""" self.browse_summary_max_token = value - def set_openai_api_key(self, value: str): + def set_openai_api_key(self, value: str) -> None: """Set the OpenAI API key value.""" self.openai_api_key = value - def set_elevenlabs_api_key(self, value: str): + def set_elevenlabs_api_key(self, value: str) -> None: """Set the ElevenLabs API key value.""" self.elevenlabs_api_key = value - def set_elevenlabs_voice_1_id(self, value: str): + def set_elevenlabs_voice_1_id(self, value: str) -> None: """Set the ElevenLabs Voice 1 ID value.""" self.elevenlabs_voice_1_id = value - def set_elevenlabs_voice_2_id(self, value: str): + def set_elevenlabs_voice_2_id(self, value: str) -> None: """Set the ElevenLabs Voice 2 ID value.""" self.elevenlabs_voice_2_id = value - def set_google_api_key(self, value: str): + def set_google_api_key(self, value: str) -> None: """Set the Google API key value.""" self.google_api_key = value - def set_custom_search_engine_id(self, value: str): + def set_custom_search_engine_id(self, value: str) -> None: """Set the custom search engine id value.""" self.custom_search_engine_id = value - def set_pinecone_api_key(self, value: str): + def set_pinecone_api_key(self, value: str) -> None: """Set the Pinecone API key value.""" self.pinecone_api_key = value - def set_pinecone_region(self, value: str): + def set_pinecone_region(self, value: str) -> None: """Set the Pinecone region value.""" self.pinecone_region = value - def set_debug_mode(self, value: bool): + def set_debug_mode(self, value: bool) -> None: """Set the debug mode value.""" self.debug_mode = value + + +def check_openai_api_key() -> None: + """Check if the OpenAI API key is set in config.py or as an environment variable.""" + cfg = Config() + if not cfg.openai_api_key: + print( + Fore.RED + + "Please set your OpenAI API key in .env or as an environment variable." + ) + print("You can get your key from https://beta.openai.com/account/api-keys") + exit(1) diff --git a/autogpt/config/singleton.py b/autogpt/config/singleton.py new file mode 100644 index 00000000..55b2aeea --- /dev/null +++ b/autogpt/config/singleton.py @@ -0,0 +1,24 @@ +"""The singleton metaclass for ensuring only one instance of a class.""" +import abc + + +class Singleton(abc.ABCMeta, type): + """ + Singleton metaclass for ensuring only one instance of a class. + """ + + _instances = {} + + def __call__(cls, *args, **kwargs): + """Call method for the singleton metaclass.""" + if cls not in cls._instances: + cls._instances[cls] = super(Singleton, cls).__call__(*args, **kwargs) + return cls._instances[cls] + + +class AbstractSingleton(abc.ABC, metaclass=Singleton): + """ + Abstract singleton class for ensuring only one instance of a class. + """ + + pass diff --git a/autogpt/data_ingestion.py b/autogpt/data_ingestion.py index 9df2b9e6..01bafc2a 100644 --- a/autogpt/data_ingestion.py +++ b/autogpt/data_ingestion.py @@ -2,7 +2,7 @@ import argparse import logging from autogpt.config import Config -from autogpt.file_operations import ingest_file, search_files +from autogpt.commands.file_operations import ingest_file, search_files from autogpt.memory import get_memory cfg = Config() @@ -87,7 +87,8 @@ def main() -> None: print(f"Error while ingesting directory '{args.dir}': {str(e)}") else: print( - "Please provide either a file path (--file) or a directory name (--dir) inside the auto_gpt_workspace directory as input." + "Please provide either a file path (--file) or a directory name (--dir)" + " inside the auto_gpt_workspace directory as input." ) diff --git a/autogpt/execute_code.py b/autogpt/execute_code.py deleted file mode 100644 index c5f85a04..00000000 --- a/autogpt/execute_code.py +++ /dev/null @@ -1,105 +0,0 @@ -import os -import subprocess - -import docker -from docker.errors import ImageNotFound - -WORKSPACE_FOLDER = "auto_gpt_workspace" - - -def execute_python_file(file): - """Execute a Python file in a Docker container and return the output""" - - print(f"Executing file '{file}' in workspace '{WORKSPACE_FOLDER}'") - - if not file.endswith(".py"): - return "Error: Invalid file type. Only .py files are allowed." - - file_path = os.path.join(WORKSPACE_FOLDER, file) - - if not os.path.isfile(file_path): - return f"Error: File '{file}' does not exist." - - if we_are_running_in_a_docker_container(): - result = subprocess.run( - f"python {file_path}", capture_output=True, encoding="utf8", shell=True - ) - if result.returncode == 0: - return result.stdout - else: - return f"Error: {result.stderr}" - else: - try: - client = docker.from_env() - - image_name = "python:3.10" - try: - client.images.get(image_name) - print(f"Image '{image_name}' found locally") - except ImageNotFound: - print( - f"Image '{image_name}' not found locally, pulling from Docker Hub" - ) - # Use the low-level API to stream the pull response - low_level_client = docker.APIClient() - for line in low_level_client.pull(image_name, stream=True, decode=True): - # Print the status and progress, if available - status = line.get("status") - progress = line.get("progress") - if status and progress: - print(f"{status}: {progress}") - elif status: - print(status) - - # You can replace 'python:3.8' with the desired Python image/version - # You can find available Python images on Docker Hub: - # https://hub.docker.com/_/python - container = client.containers.run( - image_name, - f"python {file}", - volumes={ - os.path.abspath(WORKSPACE_FOLDER): { - "bind": "/workspace", - "mode": "ro", - } - }, - working_dir="/workspace", - stderr=True, - stdout=True, - detach=True, - ) - - container.wait() - logs = container.logs().decode("utf-8") - container.remove() - - # print(f"Execution complete. Output: {output}") - # print(f"Logs: {logs}") - - return logs - - except Exception as e: - return f"Error: {str(e)}" - - -def execute_shell(command_line): - current_dir = os.getcwd() - - if WORKSPACE_FOLDER not in current_dir: # Change dir into workspace if necessary - work_dir = os.path.join(os.getcwd(), WORKSPACE_FOLDER) - os.chdir(work_dir) - - print(f"Executing command '{command_line}' in working directory '{os.getcwd()}'") - - result = subprocess.run(command_line, capture_output=True, shell=True) - output = f"STDOUT:\n{result.stdout}\nSTDERR:\n{result.stderr}" - - # Change back to whatever the prior working dir was - - os.chdir(current_dir) - - return output - - -def we_are_running_in_a_docker_container(): - os.path.exists("/.dockerenv") diff --git a/autogpt/image_gen.py b/autogpt/image_gen.py deleted file mode 100644 index 9d5694e4..00000000 --- a/autogpt/image_gen.py +++ /dev/null @@ -1,67 +0,0 @@ -import io -import os.path -import uuid -from base64 import b64decode - -import openai -import requests -from PIL import Image - -from autogpt.config import Config - -cfg = Config() - -working_directory = "auto_gpt_workspace" - - -def generate_image(prompt): - filename = str(uuid.uuid4()) + ".jpg" - - # DALL-E - if cfg.image_provider == "dalle": - openai.api_key = cfg.openai_api_key - - response = openai.Image.create( - prompt=prompt, - n=1, - size="256x256", - response_format="b64_json", - ) - - print("Image Generated for prompt:" + prompt) - - image_data = b64decode(response["data"][0]["b64_json"]) - - with open(working_directory + "/" + filename, mode="wb") as png: - png.write(image_data) - - return "Saved to disk:" + filename - - # STABLE DIFFUSION - elif cfg.image_provider == "sd": - API_URL = ( - "https://api-inference.huggingface.co/models/CompVis/stable-diffusion-v1-4" - ) - if cfg.huggingface_api_token is None: - raise ValueError( - "You need to set your Hugging Face API token in the config file." - ) - headers = {"Authorization": "Bearer " + cfg.huggingface_api_token} - - response = requests.post( - API_URL, - headers=headers, - json={ - "inputs": prompt, - }, - ) - - image = Image.open(io.BytesIO(response.content)) - print("Image Generated for prompt:" + prompt) - - image.save(os.path.join(working_directory, filename)) - - return "Saved to disk:" + filename - - else: - return "No Image Provider Set" diff --git a/autogpt/json_fixes/__init__.py b/autogpt/json_fixes/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/autogpt/json_fixes/auto_fix.py b/autogpt/json_fixes/auto_fix.py new file mode 100644 index 00000000..56664ba9 --- /dev/null +++ b/autogpt/json_fixes/auto_fix.py @@ -0,0 +1,52 @@ +"""This module contains the function to fix JSON strings using GPT-3.""" +import json + +from autogpt.llm_utils import call_ai_function +from autogpt.logs import logger +from autogpt.config import Config + +CFG = Config() + + +def fix_json(json_string: str, schema: str) -> str: + """Fix the given JSON string to make it parseable and fully compliant with + the provided schema. + + Args: + json_string (str): The JSON string to fix. + schema (str): The schema to use to fix the JSON. + Returns: + str: The fixed JSON string. + """ + # Try to fix the JSON using GPT: + function_string = "def fix_json(json_string: str, schema:str=None) -> str:" + args = [f"'''{json_string}'''", f"'''{schema}'''"] + description_string = ( + "Fixes the provided JSON string to make it parseable" + " and fully compliant with the provided schema.\n If an object or" + " field specified in the schema isn't contained within the correct" + " JSON, it is omitted.\n This function is brilliant at guessing" + " when the format is incorrect." + ) + + # If it doesn't already start with a "`", add one: + if not json_string.startswith("`"): + json_string = "```json\n" + json_string + "\n```" + result_string = call_ai_function( + function_string, args, description_string, model=CFG.fast_llm_model + ) + logger.debug("------------ JSON FIX ATTEMPT ---------------") + logger.debug(f"Original JSON: {json_string}") + logger.debug("-----------") + logger.debug(f"Fixed JSON: {result_string}") + logger.debug("----------- END OF FIX ATTEMPT ----------------") + + try: + json.loads(result_string) # just check the validity + return result_string + except json.JSONDecodeError: # noqa: E722 + # Get the call stack: + # import traceback + # call_stack = traceback.format_exc() + # print(f"Failed to fix JSON: '{json_string}' "+call_stack) + return "failed" diff --git a/autogpt/json_fixes/bracket_termination.py b/autogpt/json_fixes/bracket_termination.py new file mode 100644 index 00000000..13c2ccfd --- /dev/null +++ b/autogpt/json_fixes/bracket_termination.py @@ -0,0 +1,73 @@ +"""Fix JSON brackets.""" +import contextlib +import json +from typing import Optional +import regex +from colorama import Fore + +from autogpt.logs import logger +from autogpt.config import Config +from autogpt.speech import say_text + +CFG = Config() + + +def attempt_to_fix_json_by_finding_outermost_brackets(json_string: str): + if CFG.speak_mode and CFG.debug_mode: + say_text( + "I have received an invalid JSON response from the OpenAI API. " + "Trying to fix it now." + ) + logger.typewriter_log("Attempting to fix JSON by finding outermost brackets\n") + + try: + json_pattern = regex.compile(r"\{(?:[^{}]|(?R))*\}") + json_match = json_pattern.search(json_string) + + if json_match: + # Extract the valid JSON object from the string + json_string = json_match.group(0) + logger.typewriter_log( + title="Apparently json was fixed.", title_color=Fore.GREEN + ) + if CFG.speak_mode and CFG.debug_mode: + say_text("Apparently json was fixed.") + else: + raise ValueError("No valid JSON object found") + + except (json.JSONDecodeError, ValueError): + if CFG.debug_mode: + logger.error("Error: Invalid JSON: %s\n", json_string) + if CFG.speak_mode: + say_text("Didn't work. I will have to ignore this response then.") + logger.error("Error: Invalid JSON, setting it to empty JSON now.\n") + json_string = {} + + return json_string + + +def balance_braces(json_string: str) -> Optional[str]: + """ + Balance the braces in a JSON string. + + Args: + json_string (str): The JSON string. + + Returns: + str: The JSON string with braces balanced. + """ + + open_braces_count = json_string.count("{") + close_braces_count = json_string.count("}") + + while open_braces_count > close_braces_count: + json_string += "}" + close_braces_count += 1 + + while close_braces_count > open_braces_count: + json_string = json_string.rstrip("}") + close_braces_count -= 1 + + with contextlib.suppress(json.JSONDecodeError): + json.loads(json_string) + return json_string diff --git a/autogpt/json_fixes/escaping.py b/autogpt/json_fixes/escaping.py new file mode 100644 index 00000000..68eb1714 --- /dev/null +++ b/autogpt/json_fixes/escaping.py @@ -0,0 +1,33 @@ +""" Fix invalid escape sequences in JSON strings. """ +import json + +from autogpt.config import Config +from autogpt.json_fixes.utilities import extract_char_position + +CFG = Config() + + +def fix_invalid_escape(json_to_load: str, error_message: str) -> str: + """Fix invalid escape sequences in JSON strings. + + Args: + json_to_load (str): The JSON string. + error_message (str): The error message from the JSONDecodeError + exception. + + Returns: + str: The JSON string with invalid escape sequences fixed. + """ + while error_message.startswith("Invalid \\escape"): + bad_escape_location = extract_char_position(error_message) + json_to_load = ( + json_to_load[:bad_escape_location] + json_to_load[bad_escape_location + 1 :] + ) + try: + json.loads(json_to_load) + return json_to_load + except json.JSONDecodeError as e: + if CFG.debug_mode: + print("json loads error - fix invalid escape", e) + error_message = str(e) + return json_to_load diff --git a/autogpt/json_fixes/missing_quotes.py b/autogpt/json_fixes/missing_quotes.py new file mode 100644 index 00000000..552a1517 --- /dev/null +++ b/autogpt/json_fixes/missing_quotes.py @@ -0,0 +1,27 @@ +"""Fix quotes in a JSON string.""" +import json +import re + + +def add_quotes_to_property_names(json_string: str) -> str: + """ + Add quotes to property names in a JSON string. + + Args: + json_string (str): The JSON string. + + Returns: + str: The JSON string with quotes added to property names. + """ + + def replace_func(match: re.Match) -> str: + return f'"{match[1]}":' + + property_name_pattern = re.compile(r"(\w+):") + corrected_json_string = property_name_pattern.sub(replace_func, json_string) + + try: + json.loads(corrected_json_string) + return corrected_json_string + except json.JSONDecodeError as e: + raise e diff --git a/autogpt/json_fixes/parsing.py b/autogpt/json_fixes/parsing.py new file mode 100644 index 00000000..26d06793 --- /dev/null +++ b/autogpt/json_fixes/parsing.py @@ -0,0 +1,143 @@ +"""Fix and parse JSON strings.""" + +import contextlib +import json +from typing import Any, Dict, Union + +from autogpt.config import Config +from autogpt.json_fixes.auto_fix import fix_json +from autogpt.json_fixes.bracket_termination import balance_braces +from autogpt.json_fixes.escaping import fix_invalid_escape +from autogpt.json_fixes.missing_quotes import add_quotes_to_property_names +from autogpt.logs import logger + +CFG = Config() + + +JSON_SCHEMA = """ +{ + "command": { + "name": "command name", + "args": { + "arg name": "value" + } + }, + "thoughts": + { + "text": "thought", + "reasoning": "reasoning", + "plan": "- short bulleted\n- list that conveys\n- long-term plan", + "criticism": "constructive self-criticism", + "speak": "thoughts summary to say to user" + } +} +""" + + +def correct_json(json_to_load: str) -> str: + """ + Correct common JSON errors. + + Args: + json_to_load (str): The JSON string. + """ + + try: + if CFG.debug_mode: + print("json", json_to_load) + json.loads(json_to_load) + return json_to_load + except json.JSONDecodeError as e: + if CFG.debug_mode: + print("json loads error", e) + error_message = str(e) + if error_message.startswith("Invalid \\escape"): + json_to_load = fix_invalid_escape(json_to_load, error_message) + if error_message.startswith( + "Expecting property name enclosed in double quotes" + ): + json_to_load = add_quotes_to_property_names(json_to_load) + try: + json.loads(json_to_load) + return json_to_load + except json.JSONDecodeError as e: + if CFG.debug_mode: + print("json loads error - add quotes", e) + error_message = str(e) + if balanced_str := balance_braces(json_to_load): + return balanced_str + return json_to_load + + +def fix_and_parse_json( + json_to_load: str, try_to_fix_with_gpt: bool = True +) -> Union[str, Dict[Any, Any]]: + """Fix and parse JSON string + + Args: + json_to_load (str): The JSON string. + try_to_fix_with_gpt (bool, optional): Try to fix the JSON with GPT. + Defaults to True. + + Returns: + Union[str, Dict[Any, Any]]: The parsed JSON. + """ + + with contextlib.suppress(json.JSONDecodeError): + json_to_load = json_to_load.replace("\t", "") + return json.loads(json_to_load) + + with contextlib.suppress(json.JSONDecodeError): + json_to_load = correct_json(json_to_load) + return json.loads(json_to_load) + # Let's do something manually: + # sometimes GPT responds with something BEFORE the braces: + # "I'm sorry, I don't understand. Please try again." + # {"text": "I'm sorry, I don't understand. Please try again.", + # "confidence": 0.0} + # So let's try to find the first brace and then parse the rest + # of the string + try: + brace_index = json_to_load.index("{") + maybe_fixed_json = json_to_load[brace_index:] + last_brace_index = maybe_fixed_json.rindex("}") + maybe_fixed_json = maybe_fixed_json[: last_brace_index + 1] + return json.loads(maybe_fixed_json) + except (json.JSONDecodeError, ValueError) as e: + return try_ai_fix(try_to_fix_with_gpt, e, json_to_load) + + +def try_ai_fix( + try_to_fix_with_gpt: bool, exception: Exception, json_to_load: str +) -> Union[str, Dict[Any, Any]]: + """Try to fix the JSON with the AI + + Args: + try_to_fix_with_gpt (bool): Whether to try to fix the JSON with the AI. + exception (Exception): The exception that was raised. + json_to_load (str): The JSON string to load. + + Raises: + exception: If try_to_fix_with_gpt is False. + + Returns: + Union[str, Dict[Any, Any]]: The JSON string or dictionary. + """ + if not try_to_fix_with_gpt: + raise exception + + logger.warn( + "Warning: Failed to parse AI output, attempting to fix." + "\n If you see this warning frequently, it's likely that" + " your prompt is confusing the AI. Try changing it up" + " slightly." + ) + # Now try to fix this up using the ai_functions + ai_fixed_json = fix_json(json_to_load, JSON_SCHEMA) + + if ai_fixed_json != "failed": + return json.loads(ai_fixed_json) + # This allows the AI to react to the error message, + # which usually results in it correcting its ways. + logger.error("Failed to fix AI output, telling the AI.") + return json_to_load diff --git a/autogpt/json_fixes/utilities.py b/autogpt/json_fixes/utilities.py new file mode 100644 index 00000000..0852b18a --- /dev/null +++ b/autogpt/json_fixes/utilities.py @@ -0,0 +1,20 @@ +"""Utilities for the json_fixes package.""" +import re + + +def extract_char_position(error_message: str) -> int: + """Extract the character position from the JSONDecodeError message. + + Args: + error_message (str): The error message from the JSONDecodeError + exception. + + Returns: + int: The character position. + """ + + char_pattern = re.compile(r"\(char (\d+)\)") + if match := char_pattern.search(error_message): + return int(match[1]) + else: + raise ValueError("Character position not found in the error message.") diff --git a/autogpt/json_parser.py b/autogpt/json_parser.py deleted file mode 100644 index b32146da..00000000 --- a/autogpt/json_parser.py +++ /dev/null @@ -1,113 +0,0 @@ -import json -from typing import Any, Dict, Union - -from autogpt.call_ai_function import call_ai_function -from autogpt.config import Config -from autogpt.json_utils import correct_json -from autogpt.logger import logger - -cfg = Config() - -JSON_SCHEMA = """ -{ - "command": { - "name": "command name", - "args": { - "arg name": "value" - } - }, - "thoughts": - { - "text": "thought", - "reasoning": "reasoning", - "plan": "- short bulleted\n- list that conveys\n- long-term plan", - "criticism": "constructive self-criticism", - "speak": "thoughts summary to say to user" - } -} -""" - - -def fix_and_parse_json( - json_str: str, try_to_fix_with_gpt: bool = True -) -> Union[str, Dict[Any, Any]]: - """Fix and parse JSON string""" - try: - json_str = json_str.replace("\t", "") - return json.loads(json_str) - except json.JSONDecodeError as _: # noqa: F841 - try: - json_str = correct_json(json_str) - return json.loads(json_str) - except json.JSONDecodeError as _: # noqa: F841 - pass - # Let's do something manually: - # sometimes GPT responds with something BEFORE the braces: - # "I'm sorry, I don't understand. Please try again." - # {"text": "I'm sorry, I don't understand. Please try again.", - # "confidence": 0.0} - # So let's try to find the first brace and then parse the rest - # of the string - try: - brace_index = json_str.index("{") - json_str = json_str[brace_index:] - last_brace_index = json_str.rindex("}") - json_str = json_str[: last_brace_index + 1] - return json.loads(json_str) - # Can throw a ValueError if there is no "{" or "}" in the json_str - except (json.JSONDecodeError, ValueError) as e: # noqa: F841 - if try_to_fix_with_gpt: - logger.warn( - "Warning: Failed to parse AI output, attempting to fix." - "\n If you see this warning frequently, it's likely that" - " your prompt is confusing the AI. Try changing it up" - " slightly." - ) - # Now try to fix this up using the ai_functions - ai_fixed_json = fix_json(json_str, JSON_SCHEMA) - - if ai_fixed_json != "failed": - return json.loads(ai_fixed_json) - else: - # This allows the AI to react to the error message, - # which usually results in it correcting its ways. - logger.error("Failed to fix AI output, telling the AI.") - return json_str - else: - raise e - - -def fix_json(json_str: str, schema: str) -> str: - """Fix the given JSON string to make it parseable and fully compliant with the provided schema.""" - # Try to fix the JSON using GPT: - function_string = "def fix_json(json_str: str, schema:str=None) -> str:" - args = [f"'''{json_str}'''", f"'''{schema}'''"] - description_string = ( - "Fixes the provided JSON string to make it parseable" - " and fully compliant with the provided schema.\n If an object or" - " field specified in the schema isn't contained within the correct" - " JSON, it is omitted.\n This function is brilliant at guessing" - " when the format is incorrect." - ) - - # If it doesn't already start with a "`", add one: - if not json_str.startswith("`"): - json_str = "```json\n" + json_str + "\n```" - result_string = call_ai_function( - function_string, args, description_string, model=cfg.fast_llm_model - ) - logger.debug("------------ JSON FIX ATTEMPT ---------------") - logger.debug(f"Original JSON: {json_str}") - logger.debug("-----------") - logger.debug(f"Fixed JSON: {result_string}") - logger.debug("----------- END OF FIX ATTEMPT ----------------") - - try: - json.loads(result_string) # just check the validity - return result_string - except: # noqa: E722 - # Get the call stack: - # import traceback - # call_stack = traceback.format_exc() - # print(f"Failed to fix JSON: '{json_str}' "+call_stack) - return "failed" diff --git a/autogpt/json_utils.py b/autogpt/json_utils.py deleted file mode 100644 index 848f9fff..00000000 --- a/autogpt/json_utils.py +++ /dev/null @@ -1,128 +0,0 @@ -import json -import re -from typing import Optional - -from autogpt.config import Config - -cfg = Config() - - -def extract_char_position(error_message: str) -> int: - """Extract the character position from the JSONDecodeError message. - - Args: - error_message (str): The error message from the JSONDecodeError - exception. - - Returns: - int: The character position. - """ - import re - - char_pattern = re.compile(r"\(char (\d+)\)") - if match := char_pattern.search(error_message): - return int(match[1]) - else: - raise ValueError("Character position not found in the error message.") - - -def add_quotes_to_property_names(json_string: str) -> str: - """ - Add quotes to property names in a JSON string. - - Args: - json_string (str): The JSON string. - - Returns: - str: The JSON string with quotes added to property names. - """ - - def replace_func(match): - return f'"{match.group(1)}":' - - property_name_pattern = re.compile(r"(\w+):") - corrected_json_string = property_name_pattern.sub(replace_func, json_string) - - try: - json.loads(corrected_json_string) - return corrected_json_string - except json.JSONDecodeError as e: - raise e - - -def balance_braces(json_string: str) -> Optional[str]: - """ - Balance the braces in a JSON string. - - Args: - json_string (str): The JSON string. - - Returns: - str: The JSON string with braces balanced. - """ - - open_braces_count = json_string.count("{") - close_braces_count = json_string.count("}") - - while open_braces_count > close_braces_count: - json_string += "}" - close_braces_count += 1 - - while close_braces_count > open_braces_count: - json_string = json_string.rstrip("}") - close_braces_count -= 1 - - try: - json.loads(json_string) - return json_string - except json.JSONDecodeError: - pass - - -def fix_invalid_escape(json_str: str, error_message: str) -> str: - while error_message.startswith("Invalid \\escape"): - bad_escape_location = extract_char_position(error_message) - json_str = json_str[:bad_escape_location] + json_str[bad_escape_location + 1 :] - try: - json.loads(json_str) - return json_str - except json.JSONDecodeError as e: - if cfg.debug_mode: - print("json loads error - fix invalid escape", e) - error_message = str(e) - return json_str - - -def correct_json(json_str: str) -> str: - """ - Correct common JSON errors. - - Args: - json_str (str): The JSON string. - """ - - try: - if cfg.debug_mode: - print("json", json_str) - json.loads(json_str) - return json_str - except json.JSONDecodeError as e: - if cfg.debug_mode: - print("json loads error", e) - error_message = str(e) - if error_message.startswith("Invalid \\escape"): - json_str = fix_invalid_escape(json_str, error_message) - if error_message.startswith( - "Expecting property name enclosed in double quotes" - ): - json_str = add_quotes_to_property_names(json_str) - try: - json.loads(json_str) - return json_str - except json.JSONDecodeError as e: - if cfg.debug_mode: - print("json loads error - add quotes", e) - error_message = str(e) - if balanced_str := balance_braces(json_str): - return balanced_str - return json_str diff --git a/autogpt/llm_utils.py b/autogpt/llm_utils.py index 7503dc44..a8ac2cdb 100644 --- a/autogpt/llm_utils.py +++ b/autogpt/llm_utils.py @@ -1,4 +1,6 @@ +from ast import List import time +from typing import Dict, Optional import openai from openai.error import APIError, RateLimitError @@ -6,30 +8,79 @@ from colorama import Fore from autogpt.config import Config -cfg = Config() +CFG = Config() -openai.api_key = cfg.openai_api_key +openai.api_key = CFG.openai_api_key + + +def call_ai_function( + function: str, args: List, description: str, model: Optional[str] = None +) -> str: + """Call an AI function + + This is a magic function that can do anything with no-code. See + https://github.com/Torantulino/AI-Functions for more info. + + Args: + function (str): The function to call + args (list): The arguments to pass to the function + description (str): The description of the function + model (str, optional): The model to use. Defaults to None. + + Returns: + str: The response from the function + """ + if model is None: + model = CFG.smart_llm_model + # For each arg, if any are None, convert to "None": + args = [str(arg) if arg is not None else "None" for arg in args] + # parse args to comma separated string + args = ", ".join(args) + messages = [ + { + "role": "system", + "content": f"You are now the following python function: ```# {description}" + f"\n{function}```\n\nOnly respond with your `return` value.", + }, + {"role": "user", "content": args}, + ] + + return create_chat_completion(model=model, messages=messages, temperature=0) # Overly simple abstraction until we create something better # simple retry mechanism when getting a rate error or a bad gateway def create_chat_completion( - messages, model=None, temperature=cfg.temperature, max_tokens=None + messages: List, # type: ignore + model: Optional[str] = None, + temperature: float = CFG.temperature, + max_tokens: Optional[int] = None, ) -> str: - """Create a chat completion using the OpenAI API""" + """Create a chat completion using the OpenAI API + + Args: + messages (List[Dict[str, str]]): The messages to send to the chat completion + model (str, optional): The model to use. Defaults to None. + temperature (float, optional): The temperature to use. Defaults to 0.9. + max_tokens (int, optional): The max tokens to use. Defaults to None. + + Returns: + str: The response from the chat completion + """ response = None - num_retries = 5 - if cfg.debug_mode: + num_retries = 10 + if CFG.debug_mode: print( Fore.GREEN + f"Creating chat completion with model {model}, temperature {temperature}," f" max_tokens {max_tokens}" + Fore.RESET ) for attempt in range(num_retries): + backoff = 2 ** (attempt + 2) try: - if cfg.use_azure: + if CFG.use_azure: response = openai.ChatCompletion.create( - deployment_id=cfg.get_azure_deployment_id_for_model(model), + deployment_id=CFG.get_azure_deployment_id_for_model(model), model=model, messages=messages, temperature=temperature, @@ -44,26 +95,57 @@ def create_chat_completion( ) break except RateLimitError: - if cfg.debug_mode: + if CFG.debug_mode: print( Fore.RED + "Error: ", - "API Rate Limit Reached. Waiting 20 seconds..." + Fore.RESET, + f"Reached rate limit, passing..." + Fore.RESET, ) - time.sleep(20) except APIError as e: if e.http_status == 502: - if cfg.debug_mode: - print( - Fore.RED + "Error: ", - "API Bad gateway. Waiting 20 seconds..." + Fore.RESET, - ) - time.sleep(20) + pass else: raise if attempt == num_retries - 1: raise - + if CFG.debug_mode: + print( + Fore.RED + "Error: ", + f"API Bad gateway. Waiting {backoff} seconds..." + Fore.RESET, + ) + time.sleep(backoff) if response is None: - raise RuntimeError("Failed to get response after 5 retries") + raise RuntimeError(f"Failed to get response after {num_retries} retries") return response.choices[0].message["content"] + + +def create_embedding_with_ada(text) -> list: + """Create a embedding with text-ada-002 using the OpenAI SDK""" + num_retries = 10 + for attempt in range(num_retries): + backoff = 2 ** (attempt + 2) + try: + if CFG.use_azure: + return openai.Embedding.create(input=[text], + engine=CFG.get_azure_deployment_id_for_model("text-embedding-ada-002"), + )["data"][0]["embedding"] + else: + return openai.Embedding.create(input=[text], model="text-embedding-ada-002")[ + "data" + ][0]["embedding"] + except RateLimitError: + pass + except APIError as e: + if e.http_status == 502: + pass + else: + raise + if attempt == num_retries - 1: + raise + if CFG.debug_mode: + print( + Fore.RED + "Error: ", + f"API Bad gateway. Waiting {backoff} seconds..." + Fore.RESET, + ) + time.sleep(backoff) + diff --git a/autogpt/logger.py b/autogpt/logs.py similarity index 59% rename from autogpt/logger.py rename to autogpt/logs.py index 1b58fdff..a34d89b1 100644 --- a/autogpt/logger.py +++ b/autogpt/logs.py @@ -1,25 +1,28 @@ +"""Logging module for Auto-GPT.""" +import json import logging import os import random import re import time from logging import LogRecord +import traceback from colorama import Fore, Style -from autogpt import speak +from autogpt.speech import say_text from autogpt.config import Config, Singleton -cfg = Config() - -""" -Logger that handle titles in different colors. -Outputs logs in console, activity.log, and errors.log -For console handler: simulates typing -""" +CFG = Config() class Logger(metaclass=Singleton): + """ + Logger that handle titles in different colors. + Outputs logs in console, activity.log, and errors.log + For console handler: simulates typing + """ + def __init__(self): # create log directory if it doesn't exist this_files_dir_path = os.path.dirname(__file__) @@ -74,8 +77,8 @@ class Logger(metaclass=Singleton): def typewriter_log( self, title="", title_color="", content="", speak_text=False, level=logging.INFO ): - if speak_text and cfg.speak_mode: - speak.say_text(f"{title}. {content}") + if speak_text and CFG.speak_mode: + say_text(f"{title}. {content}") if content: if isinstance(content, list): @@ -193,3 +196,93 @@ def remove_color_codes(s: str) -> str: logger = Logger() + + +def print_assistant_thoughts(ai_name, assistant_reply): + """Prints the assistant's thoughts to the console""" + from autogpt.json_fixes.bracket_termination import ( + attempt_to_fix_json_by_finding_outermost_brackets, + ) + from autogpt.json_fixes.parsing import fix_and_parse_json + + try: + try: + # Parse and print Assistant response + assistant_reply_json = fix_and_parse_json(assistant_reply) + except json.JSONDecodeError: + logger.error("Error: Invalid JSON in assistant thoughts\n", assistant_reply) + assistant_reply_json = attempt_to_fix_json_by_finding_outermost_brackets( + assistant_reply + ) + if isinstance(assistant_reply_json, str): + assistant_reply_json = fix_and_parse_json(assistant_reply_json) + + # Check if assistant_reply_json is a string and attempt to parse + # it into a JSON object + if isinstance(assistant_reply_json, str): + try: + assistant_reply_json = json.loads(assistant_reply_json) + except json.JSONDecodeError: + logger.error("Error: Invalid JSON\n", assistant_reply) + assistant_reply_json = ( + attempt_to_fix_json_by_finding_outermost_brackets( + assistant_reply_json + ) + ) + + assistant_thoughts_reasoning = None + assistant_thoughts_plan = None + assistant_thoughts_speak = None + assistant_thoughts_criticism = None + if not isinstance(assistant_reply_json, dict): + assistant_reply_json = {} + assistant_thoughts = assistant_reply_json.get("thoughts", {}) + assistant_thoughts_text = assistant_thoughts.get("text") + + if assistant_thoughts: + assistant_thoughts_reasoning = assistant_thoughts.get("reasoning") + assistant_thoughts_plan = assistant_thoughts.get("plan") + assistant_thoughts_criticism = assistant_thoughts.get("criticism") + assistant_thoughts_speak = assistant_thoughts.get("speak") + + logger.typewriter_log( + f"{ai_name.upper()} THOUGHTS:", Fore.YELLOW, f"{assistant_thoughts_text}" + ) + logger.typewriter_log( + "REASONING:", Fore.YELLOW, f"{assistant_thoughts_reasoning}" + ) + + if assistant_thoughts_plan: + logger.typewriter_log("PLAN:", Fore.YELLOW, "") + # If it's a list, join it into a string + if isinstance(assistant_thoughts_plan, list): + assistant_thoughts_plan = "\n".join(assistant_thoughts_plan) + elif isinstance(assistant_thoughts_plan, dict): + assistant_thoughts_plan = str(assistant_thoughts_plan) + + # Split the input_string using the newline character and dashes + lines = assistant_thoughts_plan.split("\n") + for line in lines: + line = line.lstrip("- ") + logger.typewriter_log("- ", Fore.GREEN, line.strip()) + + logger.typewriter_log( + "CRITICISM:", Fore.YELLOW, f"{assistant_thoughts_criticism}" + ) + # Speak the assistant's thoughts + if CFG.speak_mode and assistant_thoughts_speak: + say_text(assistant_thoughts_speak) + + return assistant_reply_json + except json.decoder.JSONDecodeError: + logger.error("Error: Invalid JSON\n", assistant_reply) + if CFG.speak_mode: + say_text( + "I have received an invalid JSON response from the OpenAI API." + " I cannot ignore this response." + ) + + # All other errors, return "Error: + error message" + except Exception: + call_stack = traceback.format_exc() + logger.error("Error: \n", call_stack) diff --git a/autogpt/memory/__init__.py b/autogpt/memory/__init__.py index 42ba5a65..ab953527 100644 --- a/autogpt/memory/__init__.py +++ b/autogpt/memory/__init__.py @@ -21,6 +21,12 @@ except ImportError: print("Pinecone not installed. Skipping import.") PineconeMemory = None +try: + from autogpt.memory.milvus import MilvusMemory +except ImportError: + print("pymilvus not installed. Skipping import.") + MilvusMemory = None + def get_memory(cfg, init=False): memory = None @@ -42,6 +48,14 @@ def get_memory(cfg, init=False): ) else: memory = RedisMemory(cfg) + elif cfg.memory_backend == "milvus": + if not MilvusMemory: + print( + "Error: Milvus sdk is not installed." + "Please install pymilvus to use Milvus as memory backend." + ) + else: + memory = MilvusMemory(cfg) elif cfg.memory_backend == "no_memory": memory = NoMemory(cfg) @@ -56,4 +70,11 @@ def get_supported_memory_backends(): return supported_memory -__all__ = ["get_memory", "LocalCache", "RedisMemory", "PineconeMemory", "NoMemory"] +__all__ = [ + "get_memory", + "LocalCache", + "RedisMemory", + "PineconeMemory", + "NoMemory", + "MilvusMemory", +] diff --git a/autogpt/memory/local.py b/autogpt/memory/local.py index 67f040ae..004153c1 100644 --- a/autogpt/memory/local.py +++ b/autogpt/memory/local.py @@ -1,11 +1,12 @@ import dataclasses import os -from typing import Any, List, Optional +from typing import Any, List, Optional, Tuple import numpy as np import orjson -from autogpt.memory.base import MemoryProviderSingleton, get_ada_embedding +from autogpt.memory.base import MemoryProviderSingleton +from autogpt.llm_utils import create_embedding_with_ada EMBED_DIM = 1536 SAVE_OPTIONS = orjson.OPT_SERIALIZE_NUMPY | orjson.OPT_SERIALIZE_DATACLASS @@ -24,8 +25,17 @@ class CacheContent: class LocalCache(MemoryProviderSingleton): - # on load, load our database + """A class that stores the memory in a local file""" + def __init__(self, cfg) -> None: + """Initialize a class instance + + Args: + cfg: Config object + + Returns: + None + """ self.filename = f"{cfg.memory_index}.json" if os.path.exists(self.filename): try: @@ -42,7 +52,8 @@ class LocalCache(MemoryProviderSingleton): self.data = CacheContent() else: print( - f"Warning: The file '{self.filename}' does not exist. Local memory would not be saved to a file." + f"Warning: The file '{self.filename}' does not exist." + "Local memory would not be saved to a file." ) self.data = CacheContent() @@ -60,7 +71,7 @@ class LocalCache(MemoryProviderSingleton): return "" self.data.texts.append(text) - embedding = get_ada_embedding(text) + embedding = create_embedding_with_ada(text) vector = np.array(embedding).astype(np.float32) vector = vector[np.newaxis, :] @@ -108,7 +119,7 @@ class LocalCache(MemoryProviderSingleton): Returns: List[str] """ - embedding = get_ada_embedding(text) + embedding = create_embedding_with_ada(text) scores = np.dot(self.data.embeddings, embedding) @@ -116,7 +127,7 @@ class LocalCache(MemoryProviderSingleton): return [self.data.texts[i] for i in top_k_indices] - def get_stats(self): + def get_stats(self) -> Tuple[int, Tuple[int, ...]]: """ Returns: The stats of the local cache. """ diff --git a/autogpt/memory/milvus.py b/autogpt/memory/milvus.py new file mode 100644 index 00000000..c6e7d5a3 --- /dev/null +++ b/autogpt/memory/milvus.py @@ -0,0 +1,121 @@ +""" Milvus memory storage provider.""" +from pymilvus import ( + connections, + FieldSchema, + CollectionSchema, + DataType, + Collection, +) + +from autogpt.memory.base import MemoryProviderSingleton, get_ada_embedding + + +class MilvusMemory(MemoryProviderSingleton): + """Milvus memory storage provider.""" + + def __init__(self, cfg) -> None: + """Construct a milvus memory storage connection. + + Args: + cfg (Config): Auto-GPT global config. + """ + # connect to milvus server. + connections.connect(address=cfg.milvus_addr) + fields = [ + FieldSchema(name="pk", dtype=DataType.INT64, is_primary=True, auto_id=True), + FieldSchema(name="embeddings", dtype=DataType.FLOAT_VECTOR, dim=1536), + FieldSchema(name="raw_text", dtype=DataType.VARCHAR, max_length=65535), + ] + + # create collection if not exist and load it. + self.milvus_collection = cfg.milvus_collection + self.schema = CollectionSchema(fields, "auto-gpt memory storage") + self.collection = Collection(self.milvus_collection, self.schema) + # create index if not exist. + if not self.collection.has_index(): + self.collection.release() + self.collection.create_index( + "embeddings", + { + "metric_type": "IP", + "index_type": "HNSW", + "params": {"M": 8, "efConstruction": 64}, + }, + index_name="embeddings", + ) + self.collection.load() + + def add(self, data) -> str: + """Add a embedding of data into memory. + + Args: + data (str): The raw text to construct embedding index. + + Returns: + str: log. + """ + embedding = get_ada_embedding(data) + result = self.collection.insert([[embedding], [data]]) + _text = ( + "Inserting data into memory at primary key: " + f"{result.primary_keys[0]}:\n data: {data}" + ) + return _text + + def get(self, data): + """Return the most relevant data in memory. + Args: + data: The data to compare to. + """ + return self.get_relevant(data, 1) + + def clear(self) -> str: + """Drop the index in memory. + + Returns: + str: log. + """ + self.collection.drop() + self.collection = Collection(self.milvus_collection, self.schema) + self.collection.create_index( + "embeddings", + { + "metric_type": "IP", + "index_type": "HNSW", + "params": {"M": 8, "efConstruction": 64}, + }, + index_name="embeddings", + ) + self.collection.load() + return "Obliviated" + + def get_relevant(self, data: str, num_relevant: int = 5): + """Return the top-k relevant data in memory. + Args: + data: The data to compare to. + num_relevant (int, optional): The max number of relevant data. + Defaults to 5. + + Returns: + list: The top-k relevant data. + """ + # search the embedding and return the most relevant text. + embedding = get_ada_embedding(data) + search_params = { + "metrics_type": "IP", + "params": {"nprobe": 8}, + } + result = self.collection.search( + [embedding], + "embeddings", + search_params, + num_relevant, + output_fields=["raw_text"], + ) + return [item.entity.value_of_field("raw_text") for item in result[0]] + + def get_stats(self) -> str: + """ + Returns: The stats of the milvus cache. + """ + return f"Entities num: {self.collection.num_entities}" diff --git a/autogpt/memory/no_memory.py b/autogpt/memory/no_memory.py index 37f00085..0a976690 100644 --- a/autogpt/memory/no_memory.py +++ b/autogpt/memory/no_memory.py @@ -1,9 +1,14 @@ +"""A class that does not store any data. This is the default memory provider.""" from typing import Optional, List, Any from autogpt.memory.base import MemoryProviderSingleton class NoMemory(MemoryProviderSingleton): + """ + A class that does not store any data. This is the default memory provider. + """ + def __init__(self, cfg): """ Initializes the NoMemory provider. diff --git a/autogpt/memory/pinecone.py b/autogpt/memory/pinecone.py index 4e302f6e..d781073e 100644 --- a/autogpt/memory/pinecone.py +++ b/autogpt/memory/pinecone.py @@ -1,8 +1,9 @@ import pinecone from colorama import Fore, Style -from autogpt.logger import logger -from autogpt.memory.base import MemoryProviderSingleton, get_ada_embedding +from autogpt.logs import logger +from autogpt.memory.base import MemoryProviderSingleton +from autogpt.llm_utils import create_embedding_with_ada class PineconeMemory(MemoryProviderSingleton): @@ -16,7 +17,8 @@ class PineconeMemory(MemoryProviderSingleton): table_name = "auto-gpt" # this assumes we don't start with memory. # for now this works. - # we'll need a more complicated and robust system if we want to start with memory. + # we'll need a more complicated and robust system if we want to start with + # memory. self.vec_num = 0 try: @@ -28,8 +30,10 @@ class PineconeMemory(MemoryProviderSingleton): Style.BRIGHT + str(e) + Style.RESET_ALL, ) logger.double_check( - "Please ensure you have setup and configured Pinecone properly for use. " - + f"You can check out {Fore.CYAN + Style.BRIGHT}https://github.com/Torantulino/Auto-GPT#-pinecone-api-key-setup{Style.RESET_ALL} to ensure you've set up everything correctly." + "Please ensure you have setup and configured Pinecone properly for use." + + f"You can check out {Fore.CYAN + Style.BRIGHT}" + "https://github.com/Torantulino/Auto-GPT#-pinecone-api-key-setup" + f"{Style.RESET_ALL} to ensure you've set up everything correctly." ) exit(1) @@ -40,9 +44,9 @@ class PineconeMemory(MemoryProviderSingleton): self.index = pinecone.Index(table_name) def add(self, data): - vector = get_ada_embedding(data) + vector = create_embedding_with_ada(data) # no metadata here. We may wish to change that long term. - resp = self.index.upsert([(str(self.vec_num), vector, {"raw_text": data})]) + self.index.upsert([(str(self.vec_num), vector, {"raw_text": data})]) _text = f"Inserting data into memory at index: {self.vec_num}:\n data: {data}" self.vec_num += 1 return _text @@ -60,7 +64,7 @@ class PineconeMemory(MemoryProviderSingleton): :param data: The data to compare to. :param num_relevant: The number of relevant data to return. Defaults to 5 """ - query_embedding = get_ada_embedding(data) + query_embedding = create_embedding_with_ada(data) results = self.index.query( query_embedding, top_k=num_relevant, include_metadata=True ) diff --git a/autogpt/memory/redismem.py b/autogpt/memory/redismem.py index d36594ba..4d73b741 100644 --- a/autogpt/memory/redismem.py +++ b/autogpt/memory/redismem.py @@ -8,8 +8,9 @@ from redis.commands.search.field import TextField, VectorField from redis.commands.search.indexDefinition import IndexDefinition, IndexType from redis.commands.search.query import Query -from autogpt.logger import logger -from autogpt.memory.base import MemoryProviderSingleton, get_ada_embedding +from autogpt.logs import logger +from autogpt.memory.base import MemoryProviderSingleton +from autogpt.llm_utils import create_embedding_with_ada SCHEMA = [ TextField("data"), @@ -54,7 +55,9 @@ class RedisMemory(MemoryProviderSingleton): ) logger.double_check( "Please ensure you have setup and configured Redis properly for use. " - + f"You can check out {Fore.CYAN + Style.BRIGHT}https://github.com/Torantulino/Auto-GPT#redis-setup{Style.RESET_ALL} to ensure you've set up everything correctly." + + f"You can check out {Fore.CYAN + Style.BRIGHT}" + f"https://github.com/Torantulino/Auto-GPT#redis-setup{Style.RESET_ALL}" + " to ensure you've set up everything correctly." ) exit(1) @@ -83,7 +86,7 @@ class RedisMemory(MemoryProviderSingleton): """ if "Command Error:" in data: return "" - vector = get_ada_embedding(data) + vector = create_embedding_with_ada(data) vector = np.array(vector).astype(np.float32).tobytes() data_dict = {b"data": data, "embedding": vector} pipe = self.redis.pipeline() @@ -125,7 +128,7 @@ class RedisMemory(MemoryProviderSingleton): Returns: A list of the most relevant data. """ - query_embedding = get_ada_embedding(data) + query_embedding = create_embedding_with_ada(data) base_query = f"*=>[KNN {num_relevant} @embedding $vector AS vector_score]" query = ( Query(base_query) diff --git a/autogpt/processing/__init__.py b/autogpt/processing/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/autogpt/processing/text.py b/autogpt/processing/text.py new file mode 100644 index 00000000..d30036d8 --- /dev/null +++ b/autogpt/processing/text.py @@ -0,0 +1,132 @@ +"""Text processing functions""" +from typing import Generator, Optional, Dict +from selenium.webdriver.remote.webdriver import WebDriver +from autogpt.memory import get_memory +from autogpt.config import Config +from autogpt.llm_utils import create_chat_completion + +CFG = Config() +MEMORY = get_memory(CFG) + + +def split_text(text: str, max_length: int = 8192) -> Generator[str, None, None]: + """Split text into chunks of a maximum length + + Args: + text (str): The text to split + max_length (int, optional): The maximum length of each chunk. Defaults to 8192. + + Yields: + str: The next chunk of text + + Raises: + ValueError: If the text is longer than the maximum length + """ + paragraphs = text.split("\n") + current_length = 0 + current_chunk = [] + + for paragraph in paragraphs: + if current_length + len(paragraph) + 1 <= max_length: + current_chunk.append(paragraph) + current_length += len(paragraph) + 1 + else: + yield "\n".join(current_chunk) + current_chunk = [paragraph] + current_length = len(paragraph) + 1 + + if current_chunk: + yield "\n".join(current_chunk) + + +def summarize_text( + url: str, text: str, question: str, driver: Optional[WebDriver] = None +) -> str: + """Summarize text using the OpenAI API + + Args: + url (str): The url of the text + text (str): The text to summarize + question (str): The question to ask the model + driver (WebDriver): The webdriver to use to scroll the page + + Returns: + str: The summary of the text + """ + if not text: + return "Error: No text to summarize" + + text_length = len(text) + print(f"Text length: {text_length} characters") + + summaries = [] + chunks = list(split_text(text)) + scroll_ratio = 1 / len(chunks) + + for i, chunk in enumerate(chunks): + if driver: + scroll_to_percentage(driver, scroll_ratio * i) + print(f"Adding chunk {i + 1} / {len(chunks)} to memory") + + memory_to_add = f"Source: {url}\n" f"Raw content part#{i + 1}: {chunk}" + + MEMORY.add(memory_to_add) + + print(f"Summarizing chunk {i + 1} / {len(chunks)}") + messages = [create_message(chunk, question)] + + summary = create_chat_completion( + model=CFG.fast_llm_model, + messages=messages, + max_tokens=CFG.browse_summary_max_token, + ) + summaries.append(summary) + print(f"Added chunk {i + 1} summary to memory") + + memory_to_add = f"Source: {url}\n" f"Content summary part#{i + 1}: {summary}" + + MEMORY.add(memory_to_add) + + print(f"Summarized {len(chunks)} chunks.") + + combined_summary = "\n".join(summaries) + messages = [create_message(combined_summary, question)] + + return create_chat_completion( + model=CFG.fast_llm_model, + messages=messages, + max_tokens=CFG.browse_summary_max_token, + ) + + +def scroll_to_percentage(driver: WebDriver, ratio: float) -> None: + """Scroll to a percentage of the page + + Args: + driver (WebDriver): The webdriver to use + ratio (float): The percentage to scroll to + + Raises: + ValueError: If the ratio is not between 0 and 1 + """ + if ratio < 0 or ratio > 1: + raise ValueError("Percentage should be between 0 and 1") + driver.execute_script(f"window.scrollTo(0, document.body.scrollHeight * {ratio});") + + +def create_message(chunk: str, question: str) -> Dict[str, str]: + """Create a message for the chat completion + + Args: + chunk (str): The chunk of text to summarize + question (str): The question to answer + + Returns: + Dict[str, str]: The message to send to the chat completion + """ + return { + "role": "user", + "content": f'"""{chunk}""" Using the above text, answer the following' + f' question: "{question}" -- if the question cannot be answered using the text,' + " summarize the text.", + } diff --git a/autogpt/prompt.py b/autogpt/prompt.py index e41312c7..6c51f33e 100644 --- a/autogpt/prompt.py +++ b/autogpt/prompt.py @@ -1,4 +1,13 @@ +from colorama import Fore +from autogpt.config.ai_config import AIConfig +from autogpt.config.config import Config +from autogpt.logs import logger from autogpt.promptgenerator import PromptGenerator +from autogpt.config import Config +from autogpt.setup import prompt_user +from autogpt.utils import clean_input + +CFG = Config() def get_prompt() -> str: @@ -10,6 +19,9 @@ def get_prompt() -> str: str: The generated prompt string. """ + # Initialize the Config object + cfg = Config() + # Initialize the PromptGenerator object prompt_generator = PromptGenerator() @@ -47,6 +59,11 @@ def get_prompt() -> str: ), ("List GPT Agents", "list_agents", {}), ("Delete GPT Agent", "delete_agent", {"key": ""}), + ( + "Clone Repository", + "clone_repository", + {"repository_url": "", "clone_path": ""}, + ), ("Write to file", "write_to_file", {"file": "", "text": ""}), ("Read file", "read_file", {"file": ""}), ("Append to file", "append_to_file", {"file": "", "text": ""}), @@ -64,16 +81,27 @@ def get_prompt() -> str: {"code": "", "focus": ""}, ), ("Execute Python File", "execute_python_file", {"file": ""}), - ( - "Execute Shell Command, non-interactive commands only", - "execute_shell", - {"command_line": ""}, - ), - ("Task Complete (Shutdown)", "task_complete", {"reason": ""}), ("Generate Image", "generate_image", {"prompt": ""}), - ("Do Nothing", "do_nothing", {}), ] + # Only add shell command to the prompt if the AI is allowed to execute it + if cfg.execute_local_commands: + commands.append( + ( + "Execute Shell Command, non-interactive commands only", + "execute_shell", + {"command_line": ""}, + ), + ) + + # Add these command last. + commands.append( + ("Do Nothing", "do_nothing", {}), + ) + commands.append( + ("Task Complete (Shutdown)", "task_complete", {"reason": ""}), + ) + # Add commands to the PromptGenerator object for command_label, command_name, args in commands: prompt_generator.add_command(command_label, command_name, args) @@ -106,3 +134,42 @@ def get_prompt() -> str: # Generate the prompt string return prompt_generator.generate_prompt_string() + + +def construct_prompt() -> str: + """Construct the prompt for the AI to respond to + + Returns: + str: The prompt string + """ + config = AIConfig.load(CFG.ai_settings_file) + if CFG.skip_reprompt and config.ai_name: + logger.typewriter_log("Name :", Fore.GREEN, config.ai_name) + logger.typewriter_log("Role :", Fore.GREEN, config.ai_role) + logger.typewriter_log("Goals:", Fore.GREEN, f"{config.ai_goals}") + elif config.ai_name: + logger.typewriter_log( + "Welcome back! ", + Fore.GREEN, + f"Would you like me to return to being {config.ai_name}?", + speak_text=True, + ) + should_continue = clean_input( + f"""Continue with the last settings? +Name: {config.ai_name} +Role: {config.ai_role} +Goals: {config.ai_goals} +Continue (y/n): """ + ) + if should_continue.lower() == "n": + config = AIConfig() + + if not config.ai_name: + config = prompt_user() + config.save() + + # Get rid of this global: + global ai_name + ai_name = config.ai_name + + return config.construct_full_prompt() diff --git a/autogpt/promptgenerator.py b/autogpt/promptgenerator.py index d6802cec..82dba8e0 100644 --- a/autogpt/promptgenerator.py +++ b/autogpt/promptgenerator.py @@ -1,4 +1,6 @@ +""" A module for generating custom prompt strings.""" import json +from typing import Any, Dict, List class PromptGenerator: @@ -7,7 +9,7 @@ class PromptGenerator: resources, and performance evaluations. """ - def __init__(self): + def __init__(self) -> None: """ Initialize the PromptGenerator object with empty lists of constraints, commands, resources, and performance evaluations. @@ -27,7 +29,7 @@ class PromptGenerator: "command": {"name": "command name", "args": {"arg name": "value"}}, } - def add_constraint(self, constraint): + def add_constraint(self, constraint: str) -> None: """ Add a constraint to the constraints list. @@ -36,7 +38,7 @@ class PromptGenerator: """ self.constraints.append(constraint) - def add_command(self, command_label, command_name, args=None): + def add_command(self, command_label: str, command_name: str, args=None) -> None: """ Add a command to the commands list with a label, name, and optional arguments. @@ -59,7 +61,7 @@ class PromptGenerator: self.commands.append(command) - def _generate_command_string(self, command): + def _generate_command_string(self, command: Dict[str, Any]) -> str: """ Generate a formatted string representation of a command. @@ -92,7 +94,7 @@ class PromptGenerator: """ self.performance_evaluation.append(evaluation) - def _generate_numbered_list(self, items, item_type="list") -> str: + def _generate_numbered_list(self, items: List[Any], item_type="list") -> str: """ Generate a numbered list from given items based on the item_type. diff --git a/autogpt/setup.py b/autogpt/setup.py new file mode 100644 index 00000000..5315c01d --- /dev/null +++ b/autogpt/setup.py @@ -0,0 +1,69 @@ +"""Setup the AI and its goals""" +from colorama import Fore, Style +from autogpt import utils +from autogpt.config.ai_config import AIConfig +from autogpt.logs import logger + + +def prompt_user() -> AIConfig: + """Prompt the user for input + + Returns: + AIConfig: The AIConfig object containing the user's input + """ + ai_name = "" + # Construct the prompt + logger.typewriter_log( + "Welcome to Auto-GPT! ", + Fore.GREEN, + "Enter the name of your AI and its role below. Entering nothing will load" + " defaults.", + speak_text=True, + ) + + # Get AI Name from User + logger.typewriter_log( + "Name your AI: ", Fore.GREEN, "For example, 'Entrepreneur-GPT'" + ) + ai_name = utils.clean_input("AI Name: ") + if ai_name == "": + ai_name = "Entrepreneur-GPT" + + logger.typewriter_log( + f"{ai_name} here!", Fore.LIGHTBLUE_EX, "I am at your service.", speak_text=True + ) + + # Get AI Role from User + logger.typewriter_log( + "Describe your AI's role: ", + Fore.GREEN, + "For example, 'an AI designed to autonomously develop and run businesses with" + " the sole goal of increasing your net worth.'", + ) + ai_role = utils.clean_input(f"{ai_name} is: ") + if ai_role == "": + ai_role = "an AI designed to autonomously develop and run businesses with the" + " sole goal of increasing your net worth." + + # Enter up to 5 goals for the AI + logger.typewriter_log( + "Enter up to 5 goals for your AI: ", + Fore.GREEN, + "For example: \nIncrease net worth, Grow Twitter Account, Develop and manage" + " multiple businesses autonomously'", + ) + print("Enter nothing to load defaults, enter nothing when finished.", flush=True) + ai_goals = [] + for i in range(5): + ai_goal = utils.clean_input(f"{Fore.LIGHTBLUE_EX}Goal{Style.RESET_ALL} {i+1}: ") + if ai_goal == "": + break + ai_goals.append(ai_goal) + if not ai_goals: + ai_goals = [ + "Increase net worth", + "Grow Twitter Account", + "Develop and manage multiple businesses autonomously", + ] + + return AIConfig(ai_name, ai_role, ai_goals) diff --git a/autogpt/speak.py b/autogpt/speak.py deleted file mode 100644 index 41b3ac9f..00000000 --- a/autogpt/speak.py +++ /dev/null @@ -1,120 +0,0 @@ -import os - -import requests -from playsound import playsound - -from autogpt.config import Config - -import threading -from threading import Lock, Semaphore - -import gtts - -cfg = Config() - -# Default voice IDs -default_voices = ["ErXwobaYiN019PkySvjV", "EXAVITQu4vr4xnSDxMaL"] - -# Retrieve custom voice IDs from the Config class -custom_voice_1 = cfg.elevenlabs_voice_1_id -custom_voice_2 = cfg.elevenlabs_voice_2_id - -# Placeholder values that should be treated as empty -placeholders = {"your-voice-id"} - -# Use custom voice IDs if provided and not placeholders, otherwise use default voice IDs -voices = [ - custom_voice_1 - if custom_voice_1 and custom_voice_1 not in placeholders - else default_voices[0], - custom_voice_2 - if custom_voice_2 and custom_voice_2 not in placeholders - else default_voices[1], -] - -tts_headers = {"Content-Type": "application/json", "xi-api-key": cfg.elevenlabs_api_key} - -mutex_lock = Lock() # Ensure only one sound is played at a time -queue_semaphore = Semaphore( - 1 -) # The amount of sounds to queue before blocking the main thread - - -def eleven_labs_speech(text, voice_index=0): - """Speak text using elevenlabs.io's API""" - tts_url = "https://api.elevenlabs.io/v1/text-to-speech/{voice_id}".format( - voice_id=voices[voice_index] - ) - formatted_message = {"text": text} - response = requests.post(tts_url, headers=tts_headers, json=formatted_message) - - if response.status_code == 200: - with mutex_lock: - with open("speech.mpeg", "wb") as f: - f.write(response.content) - playsound("speech.mpeg", True) - os.remove("speech.mpeg") - return True - else: - print("Request failed with status code:", response.status_code) - print("Response content:", response.content) - return False - - -def brian_speech(text): - """Speak text using Brian with the streamelements API""" - tts_url = f"https://api.streamelements.com/kappa/v2/speech?voice=Brian&text={text}" - response = requests.get(tts_url) - - if response.status_code == 200: - with mutex_lock: - with open("speech.mp3", "wb") as f: - f.write(response.content) - playsound("speech.mp3") - os.remove("speech.mp3") - return True - else: - print("Request failed with status code:", response.status_code) - print("Response content:", response.content) - return False - - -def gtts_speech(text): - tts = gtts.gTTS(text) - with mutex_lock: - tts.save("speech.mp3") - playsound("speech.mp3", True) - os.remove("speech.mp3") - - -def macos_tts_speech(text, voice_index=0): - if voice_index == 0: - os.system(f'say "{text}"') - else: - if voice_index == 1: - os.system(f'say -v "Ava (Premium)" "{text}"') - else: - os.system(f'say -v Samantha "{text}"') - - -def say_text(text, voice_index=0): - def speak(): - if not cfg.elevenlabs_api_key: - if cfg.use_mac_os_tts == "True": - macos_tts_speech(text) - elif cfg.use_brian_tts == "True": - success = brian_speech(text) - if not success: - gtts_speech(text) - else: - gtts_speech(text) - else: - success = eleven_labs_speech(text, voice_index) - if not success: - gtts_speech(text) - - queue_semaphore.release() - - queue_semaphore.acquire(True) - thread = threading.Thread(target=speak) - thread.start() diff --git a/autogpt/speech/__init__.py b/autogpt/speech/__init__.py new file mode 100644 index 00000000..2ff0d2bf --- /dev/null +++ b/autogpt/speech/__init__.py @@ -0,0 +1,4 @@ +"""This module contains the speech recognition and speech synthesis functions.""" +from autogpt.speech.say import say_text + +__all__ = ["say_text"] diff --git a/autogpt/speech/base.py b/autogpt/speech/base.py new file mode 100644 index 00000000..d74fa51b --- /dev/null +++ b/autogpt/speech/base.py @@ -0,0 +1,50 @@ +"""Base class for all voice classes.""" +import abc +from threading import Lock + +from autogpt.config import AbstractSingleton + + +class VoiceBase(AbstractSingleton): + """ + Base class for all voice classes. + """ + + def __init__(self): + """ + Initialize the voice class. + """ + self._url = None + self._headers = None + self._api_key = None + self._voices = [] + self._mutex = Lock() + self._setup() + + def say(self, text: str, voice_index: int = 0) -> bool: + """ + Say the given text. + + Args: + text (str): The text to say. + voice_index (int): The index of the voice to use. + """ + with self._mutex: + return self._speech(text, voice_index) + + @abc.abstractmethod + def _setup(self) -> None: + """ + Setup the voices, API key, etc. + """ + pass + + @abc.abstractmethod + def _speech(self, text: str, voice_index: int = 0) -> bool: + """ + Play the given text. + + Args: + text (str): The text to play. + """ + pass diff --git a/autogpt/speech/brian.py b/autogpt/speech/brian.py new file mode 100644 index 00000000..e581bbcc --- /dev/null +++ b/autogpt/speech/brian.py @@ -0,0 +1,39 @@ +""" Brian speech module for autogpt """ +import os +import requests +from playsound import playsound + +from autogpt.speech.base import VoiceBase + + +class BrianSpeech(VoiceBase): + """Brian speech module for autogpt""" + + def _setup(self) -> None: + """Setup the voices, API key, etc.""" + pass + + def _speech(self, text: str) -> bool: + """Speak text using Brian with the streamelements API + + Args: + text (str): The text to speak + + Returns: + bool: True if the request was successful, False otherwise + """ + tts_url = ( + f"https://api.streamelements.com/kappa/v2/speech?voice=Brian&text={text}" + ) + response = requests.get(tts_url) + + if response.status_code == 200: + with open("speech.mp3", "wb") as f: + f.write(response.content) + playsound("speech.mp3") + os.remove("speech.mp3") + return True + else: + print("Request failed with status code:", response.status_code) + print("Response content:", response.content) + return False diff --git a/autogpt/speech/eleven_labs.py b/autogpt/speech/eleven_labs.py new file mode 100644 index 00000000..2c54ffad --- /dev/null +++ b/autogpt/speech/eleven_labs.py @@ -0,0 +1,71 @@ +"""ElevenLabs speech module""" +import os +from playsound import playsound + +import requests + +from autogpt.config import Config +from autogpt.speech.base import VoiceBase + +PLACEHOLDERS = {"your-voice-id"} + + +class ElevenLabsSpeech(VoiceBase): + """ElevenLabs speech class""" + + def _setup(self) -> None: + """Setup the voices, API key, etc. + + Returns: + None: None + """ + + cfg = Config() + default_voices = ["ErXwobaYiN019PkySvjV", "EXAVITQu4vr4xnSDxMaL"] + self._headers = { + "Content-Type": "application/json", + "xi-api-key": cfg.elevenlabs_api_key, + } + self._voices = default_voices.copy() + self._use_custom_voice(cfg.elevenlabs_voice_1_id, 0) + self._use_custom_voice(cfg.elevenlabs_voice_2_id, 1) + + def _use_custom_voice(self, voice, voice_index) -> None: + """Use a custom voice if provided and not a placeholder + + Args: + voice (str): The voice ID + voice_index (int): The voice index + + Returns: + None: None + """ + # Placeholder values that should be treated as empty + if voice and voice not in PLACEHOLDERS: + self._voices[voice_index] = voice + + def _speech(self, text: str, voice_index: int = 0) -> bool: + """Speak text using elevenlabs.io's API + + Args: + text (str): The text to speak + voice_index (int, optional): The voice to use. Defaults to 0. + + Returns: + bool: True if the request was successful, False otherwise + """ + tts_url = ( + f"https://api.elevenlabs.io/v1/text-to-speech/{self._voices[voice_index]}" + ) + response = requests.post(tts_url, headers=self._headers, json={"text": text}) + + if response.status_code == 200: + with open("speech.mpeg", "wb") as f: + f.write(response.content) + playsound("speech.mpeg", True) + os.remove("speech.mpeg") + return True + else: + print("Request failed with status code:", response.status_code) + print("Response content:", response.content) + return False diff --git a/autogpt/speech/gtts.py b/autogpt/speech/gtts.py new file mode 100644 index 00000000..37497075 --- /dev/null +++ b/autogpt/speech/gtts.py @@ -0,0 +1,21 @@ +""" GTTS Voice. """ +import os +from playsound import playsound +import gtts + +from autogpt.speech.base import VoiceBase + + +class GTTSVoice(VoiceBase): + """GTTS Voice.""" + + def _setup(self) -> None: + pass + + def _speech(self, text: str, _: int = 0) -> bool: + """Play the given text.""" + tts = gtts.gTTS(text) + tts.save("speech.mp3") + playsound("speech.mp3", True) + os.remove("speech.mp3") + return True diff --git a/autogpt/speech/macos_tts.py b/autogpt/speech/macos_tts.py new file mode 100644 index 00000000..4c072ce2 --- /dev/null +++ b/autogpt/speech/macos_tts.py @@ -0,0 +1,21 @@ +""" MacOS TTS Voice. """ +import os + +from autogpt.speech.base import VoiceBase + + +class MacOSTTS(VoiceBase): + """MacOS TTS Voice.""" + + def _setup(self) -> None: + pass + + def _speech(self, text: str, voice_index: int = 0) -> bool: + """Play the given text.""" + if voice_index == 0: + os.system(f'say "{text}"') + elif voice_index == 1: + os.system(f'say -v "Ava (Premium)" "{text}"') + else: + os.system(f'say -v Samantha "{text}"') + return True diff --git a/autogpt/speech/say.py b/autogpt/speech/say.py new file mode 100644 index 00000000..78b75b21 --- /dev/null +++ b/autogpt/speech/say.py @@ -0,0 +1,42 @@ +""" Text to speech module """ +from autogpt.config import Config + +import threading +from threading import Semaphore +from autogpt.speech.brian import BrianSpeech +from autogpt.speech.macos_tts import MacOSTTS +from autogpt.speech.gtts import GTTSVoice +from autogpt.speech.eleven_labs import ElevenLabsSpeech + + +CFG = Config() +DEFAULT_VOICE_ENGINE = GTTSVoice() +VOICE_ENGINE = None +if CFG.elevenlabs_api_key: + VOICE_ENGINE = ElevenLabsSpeech() +elif CFG.use_mac_os_tts == "True": + VOICE_ENGINE = MacOSTTS() +elif CFG.use_brian_tts == "True": + VOICE_ENGINE = BrianSpeech() +else: + VOICE_ENGINE = GTTSVoice() + + +QUEUE_SEMAPHORE = Semaphore( + 1 +) # The amount of sounds to queue before blocking the main thread + + +def say_text(text: str, voice_index: int = 0) -> None: + """Speak the given text using the given voice index""" + + def speak() -> None: + success = VOICE_ENGINE.say(text, voice_index) + if not success: + DEFAULT_VOICE_ENGINE.say(text) + + QUEUE_SEMAPHORE.release() + + QUEUE_SEMAPHORE.acquire(True) + thread = threading.Thread(target=speak) + thread.start() diff --git a/autogpt/spinner.py b/autogpt/spinner.py index 65679d56..56b4f20a 100644 --- a/autogpt/spinner.py +++ b/autogpt/spinner.py @@ -1,3 +1,4 @@ +"""A simple spinner module""" import itertools import sys import threading @@ -7,15 +8,20 @@ import time class Spinner: """A simple spinner class""" - def __init__(self, message="Loading...", delay=0.1): - """Initialize the spinner class""" + def __init__(self, message: str = "Loading...", delay: float = 0.1) -> None: + """Initialize the spinner class + + Args: + message (str): The message to display. + delay (float): The delay between each spinner update. + """ self.spinner = itertools.cycle(["-", "/", "|", "\\"]) self.delay = delay self.message = message self.running = False self.spinner_thread = None - def spin(self): + def spin(self) -> None: """Spin the spinner""" while self.running: sys.stdout.write(f"{next(self.spinner)} {self.message}\r") @@ -23,14 +29,20 @@ class Spinner: time.sleep(self.delay) sys.stdout.write(f"\r{' ' * (len(self.message) + 2)}\r") - def __enter__(self): + def __enter__(self) -> None: """Start the spinner""" self.running = True self.spinner_thread = threading.Thread(target=self.spin) self.spinner_thread.start() - def __exit__(self, exc_type, exc_value, exc_traceback): - """Stop the spinner""" + def __exit__(self, exc_type, exc_value, exc_traceback) -> None: + """Stop the spinner + + Args: + exc_type (Exception): The exception type. + exc_value (Exception): The exception value. + exc_traceback (Exception): The exception traceback. + """ self.running = False if self.spinner_thread is not None: self.spinner_thread.join() diff --git a/autogpt/summary.py b/autogpt/summary.py deleted file mode 100644 index 38f0aadc..00000000 --- a/autogpt/summary.py +++ /dev/null @@ -1,69 +0,0 @@ -from autogpt.llm_utils import create_chat_completion - - -def summarize_text(driver, text, question): - if not text: - return "Error: No text to summarize" - - text_length = len(text) - print(f"Text length: {text_length} characters") - - summaries = [] - chunks = list(split_text(text)) - - scroll_ratio = 1 / len(chunks) - for i, chunk in enumerate(chunks): - scroll_to_percentage(driver, scroll_ratio * i) - print(f"Summarizing chunk {i + 1} / {len(chunks)}") - messages = [create_message(chunk, question)] - - summary = create_chat_completion( - model="gpt-3.5-turbo", - messages=messages, - max_tokens=300, - ) - summaries.append(summary) - - print(f"Summarized {len(chunks)} chunks.") - - combined_summary = "\n".join(summaries) - messages = [create_message(combined_summary, question)] - - return create_chat_completion( - model="gpt-3.5-turbo", - messages=messages, - max_tokens=300, - ) - - -def split_text(text, max_length=8192): - paragraphs = text.split("\n") - current_length = 0 - current_chunk = [] - - for paragraph in paragraphs: - if current_length + len(paragraph) + 1 <= max_length: - current_chunk.append(paragraph) - current_length += len(paragraph) + 1 - else: - yield "\n".join(current_chunk) - current_chunk = [paragraph] - current_length = len(paragraph) + 1 - - if current_chunk: - yield "\n".join(current_chunk) - - -def create_message(chunk, question): - return { - "role": "user", - "content": f'"""{chunk}""" Using the above text, please answer the following' - f' question: "{question}" -- if the question cannot be answered using the text,' - " please summarize the text.", - } - - -def scroll_to_percentage(driver, ratio): - if ratio < 0 or ratio > 1: - raise ValueError("Percentage should be between 0 and 1") - driver.execute_script(f"window.scrollTo(0, document.body.scrollHeight * {ratio});") diff --git a/autogpt/token_counter.py b/autogpt/token_counter.py index 875e8d42..a85a54be 100644 --- a/autogpt/token_counter.py +++ b/autogpt/token_counter.py @@ -1,8 +1,9 @@ +"""Functions for counting the number of tokens in a message or string.""" from typing import Dict, List import tiktoken -from autogpt.logger import logger +from autogpt.logs import logger def count_message_tokens( @@ -12,13 +13,13 @@ def count_message_tokens( Returns the number of tokens used by a list of messages. Args: - messages (list): A list of messages, each of which is a dictionary - containing the role and content of the message. - model (str): The name of the model to use for tokenization. - Defaults to "gpt-3.5-turbo-0301". + messages (list): A list of messages, each of which is a dictionary + containing the role and content of the message. + model (str): The name of the model to use for tokenization. + Defaults to "gpt-3.5-turbo-0301". Returns: - int: The number of tokens used by the list of messages. + int: The number of tokens used by the list of messages. """ try: encoding = tiktoken.encoding_for_model(model) @@ -26,7 +27,7 @@ def count_message_tokens( logger.warn("Warning: model not found. Using cl100k_base encoding.") encoding = tiktoken.get_encoding("cl100k_base") if model == "gpt-3.5-turbo": - # !Node: gpt-3.5-turbo may change over time. + # !Note: gpt-3.5-turbo may change over time. # Returning num tokens assuming gpt-3.5-turbo-0301.") return count_message_tokens(messages, model="gpt-3.5-turbo-0301") elif model == "gpt-4": @@ -62,11 +63,11 @@ def count_string_tokens(string: str, model_name: str) -> int: Returns the number of tokens in a text string. Args: - string (str): The text string. - model_name (str): The name of the encoding to use. (e.g., "gpt-3.5-turbo") + string (str): The text string. + model_name (str): The name of the encoding to use. (e.g., "gpt-3.5-turbo") Returns: - int: The number of tokens in the text string. + int: The number of tokens in the text string. """ encoding = tiktoken.encoding_for_model(model_name) return len(encoding.encode(string)) diff --git a/autogpt/web.py b/autogpt/web.py deleted file mode 100644 index 355f7fd3..00000000 --- a/autogpt/web.py +++ /dev/null @@ -1,85 +0,0 @@ -from selenium import webdriver -import autogpt.summary as summary -from bs4 import BeautifulSoup -from selenium.webdriver.common.by import By -from selenium.webdriver.support.wait import WebDriverWait -from selenium.webdriver.support import expected_conditions as EC -from webdriver_manager.chrome import ChromeDriverManager -from selenium.webdriver.chrome.options import Options -import logging -from pathlib import Path -from autogpt.config import Config - -file_dir = Path(__file__).parent -cfg = Config() - - -def browse_website(url, question): - driver, text = scrape_text_with_selenium(url) - add_header(driver) - summary_text = summary.summarize_text(driver, text, question) - links = scrape_links_with_selenium(driver) - - # Limit links to 5 - if len(links) > 5: - links = links[:5] - close_browser(driver) - return f"Answer gathered from website: {summary_text} \n \n Links: {links}", driver - - -def scrape_text_with_selenium(url): - logging.getLogger("selenium").setLevel(logging.CRITICAL) - - options = Options() - options.add_argument( - "user-agent=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/112.0.5615.49 Safari/537.36" - ) - driver = webdriver.Chrome( - executable_path=ChromeDriverManager().install(), options=options - ) - driver.get(url) - - WebDriverWait(driver, 10).until( - EC.presence_of_element_located((By.TAG_NAME, "body")) - ) - - # Get the HTML content directly from the browser's DOM - page_source = driver.execute_script("return document.body.outerHTML;") - soup = BeautifulSoup(page_source, "html.parser") - - for script in soup(["script", "style"]): - script.extract() - - text = soup.get_text() - lines = (line.strip() for line in text.splitlines()) - chunks = (phrase.strip() for line in lines for phrase in line.split(" ")) - text = "\n".join(chunk for chunk in chunks if chunk) - return driver, text - - -def scrape_links_with_selenium(driver): - page_source = driver.page_source - soup = BeautifulSoup(page_source, "html.parser") - - for script in soup(["script", "style"]): - script.extract() - - hyperlinks = extract_hyperlinks(soup) - - return format_hyperlinks(hyperlinks) - - -def close_browser(driver): - driver.quit() - - -def extract_hyperlinks(soup): - return [(link.text, link["href"]) for link in soup.find_all("a", href=True)] - - -def format_hyperlinks(hyperlinks): - return [f"{link_text} ({link_url})" for link_text, link_url in hyperlinks] - - -def add_header(driver): - driver.execute_script(open(f"{file_dir}/js/overlay.js", "r").read()) diff --git a/docker-compose.yml b/docker-compose.yml index 79f20bb5..be225d7d 100644 --- a/docker-compose.yml +++ b/docker-compose.yml @@ -7,6 +7,8 @@ services: depends_on: - redis build: ./ + env_file: + - .env volumes: - "./autogpt:/app" - ".env:/app/.env" diff --git a/docs/imgs/openai-api-key-billing-paid-account.png b/docs/imgs/openai-api-key-billing-paid-account.png new file mode 100644 index 00000000..8948505a Binary files /dev/null and b/docs/imgs/openai-api-key-billing-paid-account.png differ diff --git a/requirements-docker.txt b/requirements-docker.txt new file mode 100644 index 00000000..fecbd0a6 --- /dev/null +++ b/requirements-docker.txt @@ -0,0 +1,25 @@ +beautifulsoup4 +colorama==0.4.6 +openai==0.27.2 +playsound==1.2.2 +python-dotenv==1.0.0 +pyyaml==6.0 +readability-lxml==0.8.1 +requests +tiktoken==0.3.3 +gTTS==2.3.1 +docker +duckduckgo-search +google-api-python-client #(https://developers.google.com/custom-search/v1/overview) +pinecone-client==2.2.1 +redis +orjson +Pillow +selenium +webdriver-manager +coverage +flake8 +numpy +pre-commit +black +isort diff --git a/requirements.txt b/requirements.txt index 44cd600f..dbfa7741 100644 --- a/requirements.txt +++ b/requirements.txt @@ -24,3 +24,4 @@ pre-commit black sourcery isort +gitpython==3.1.31 diff --git a/run.bat b/run.bat new file mode 100644 index 00000000..afbab57a --- /dev/null +++ b/run.bat @@ -0,0 +1,8 @@ +@echo off +python scripts/check_requirements.py requirements.txt +if errorlevel 1 ( + echo Installing missing packages... + pip install -r requirements.txt +) +python -m autogpt %* +pause diff --git a/run_continuous.bat b/run_continuous.bat new file mode 100644 index 00000000..812aa01c --- /dev/null +++ b/run_continuous.bat @@ -0,0 +1,3 @@ +@echo off +set argument=--continuous +call run.bat %argument% diff --git a/scripts/check_requirements.py b/scripts/check_requirements.py new file mode 100644 index 00000000..d1f23504 --- /dev/null +++ b/scripts/check_requirements.py @@ -0,0 +1,31 @@ +import pkg_resources +import sys + + +def main(): + requirements_file = sys.argv[1] + with open(requirements_file, "r") as f: + required_packages = [ + line.strip().split("#")[0].strip() for line in f.readlines() + ] + + installed_packages = [package.key for package in pkg_resources.working_set] + + missing_packages = [] + for package in required_packages: + if not package: # Skip empty lines + continue + package_name = package.strip().split("==")[0] + if package_name.lower() not in installed_packages: + missing_packages.append(package_name) + + if missing_packages: + print("Missing packages:") + print(", ".join(missing_packages)) + sys.exit(1) + else: + print("All packages are installed.") + + +if __name__ == "__main__": + main() diff --git a/scripts/main.py b/scripts/main.py deleted file mode 100644 index b6871d1c..00000000 --- a/scripts/main.py +++ /dev/null @@ -1,11 +0,0 @@ -from colorama import Style, init - -# Initialize colorama -init(autoreset=True) - -# Use the bold ANSI style -print( - f"""{Style.BRIGHT}Please run: -python -m autogpt -""" -) diff --git a/tests.py b/tests.py index 46f802f6..67ba1c8e 100644 --- a/tests.py +++ b/tests.py @@ -1,8 +1,20 @@ import unittest +import coverage if __name__ == "__main__": + # Start coverage collection + cov = coverage.Coverage() + cov.start() + # Load all tests from the 'autogpt/tests' package - suite = unittest.defaultTestLoader.discover("autogpt/tests") + suite = unittest.defaultTestLoader.discover("./tests") # Run the tests unittest.TextTestRunner().run(suite) + + # Stop coverage collection + cov.stop() + cov.save() + + # Report the coverage + cov.report(show_missing=True) diff --git a/tests/integration/milvus_memory_tests.py b/tests/integration/milvus_memory_tests.py new file mode 100644 index 00000000..96934cd6 --- /dev/null +++ b/tests/integration/milvus_memory_tests.py @@ -0,0 +1,48 @@ +import random +import string +import unittest + +from autogpt.config import Config +from autogpt.memory.milvus import MilvusMemory + + +class TestMilvusMemory(unittest.TestCase): + def random_string(self, length): + return "".join(random.choice(string.ascii_letters) for _ in range(length)) + + def setUp(self): + cfg = Config() + cfg.milvus_addr = "localhost:19530" + self.memory = MilvusMemory(cfg) + self.memory.clear() + + # Add example texts to the cache + self.example_texts = [ + "The quick brown fox jumps over the lazy dog", + "I love machine learning and natural language processing", + "The cake is a lie, but the pie is always true", + "ChatGPT is an advanced AI model for conversation", + ] + + for text in self.example_texts: + self.memory.add(text) + + # Add some random strings to test noise + for _ in range(5): + self.memory.add(self.random_string(10)) + + def test_get_relevant(self): + query = "I'm interested in artificial intelligence and NLP" + k = 3 + relevant_texts = self.memory.get_relevant(query, k) + + print(f"Top {k} relevant texts for the query '{query}':") + for i, text in enumerate(relevant_texts, start=1): + print(f"{i}. {text}") + + self.assertEqual(len(relevant_texts), k) + self.assertIn(self.example_texts[1], relevant_texts) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/milvus_memory_test.py b/tests/milvus_memory_test.py new file mode 100644 index 00000000..0113fa1c --- /dev/null +++ b/tests/milvus_memory_test.py @@ -0,0 +1,63 @@ +import os +import sys +import unittest + +from autogpt.memory.milvus import MilvusMemory + + +def MockConfig(): + return type( + "MockConfig", + (object,), + { + "debug_mode": False, + "continuous_mode": False, + "speak_mode": False, + "milvus_collection": "autogpt", + "milvus_addr": "localhost:19530", + }, + ) + + +class TestMilvusMemory(unittest.TestCase): + def setUp(self): + self.cfg = MockConfig() + self.memory = MilvusMemory(self.cfg) + + def test_add(self): + text = "Sample text" + self.memory.clear() + self.memory.add(text) + result = self.memory.get(text) + self.assertEqual([text], result) + + def test_clear(self): + self.memory.clear() + self.assertEqual(self.memory.collection.num_entities, 0) + + def test_get(self): + text = "Sample text" + self.memory.clear() + self.memory.add(text) + result = self.memory.get(text) + self.assertEqual(result, [text]) + + def test_get_relevant(self): + text1 = "Sample text 1" + text2 = "Sample text 2" + self.memory.clear() + self.memory.add(text1) + self.memory.add(text2) + result = self.memory.get_relevant(text1, 1) + self.assertEqual(result, [text1]) + + def test_get_stats(self): + text = "Sample text" + self.memory.clear() + self.memory.add(text) + stats = self.memory.get_stats() + self.assertEqual(15, len(stats)) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/smoke_test.py b/tests/smoke_test.py new file mode 100644 index 00000000..8d33cf89 --- /dev/null +++ b/tests/smoke_test.py @@ -0,0 +1,63 @@ +import os +import subprocess +import sys +import unittest + +from autogpt.file_operations import delete_file, read_file + +env_vars = {"MEMORY_BACKEND": "no_memory", "TEMPERATURE": "0"} + + +class TestCommands(unittest.TestCase): + def test_write_file(self): + # Test case to check if the write_file command can successfully write 'Hello World' to a file + # named 'hello_world.txt'. + + # Read the current ai_settings.yaml file and store its content. + ai_settings = None + if os.path.exists("ai_settings.yaml"): + with open("ai_settings.yaml", "r") as f: + ai_settings = f.read() + os.remove("ai_settings.yaml") + + try: + if os.path.exists("hello_world.txt"): + # Clean up any existing 'hello_world.txt' file before testing. + delete_file("hello_world.txt") + # Prepare input data for the test. + input_data = """write_file-GPT +an AI designed to use the write_file command to write 'Hello World' into a file named "hello_world.txt" and then use the task_complete command to complete the task. +Use the write_file command to write 'Hello World' into a file named "hello_world.txt". +Use the task_complete command to complete the task. +Do not use any other commands. + +y -5 +EOF""" + command = f"{sys.executable} -m autogpt" + + # Execute the script with the input data. + process = subprocess.Popen( + command, + stdin=subprocess.PIPE, + shell=True, + env={**os.environ, **env_vars}, + ) + process.communicate(input_data.encode()) + + # Read the content of the 'hello_world.txt' file created during the test. + content = read_file("hello_world.txt") + finally: + if ai_settings: + # Restore the original ai_settings.yaml file. + with open("ai_settings.yaml", "w") as f: + f.write(ai_settings) + + # Check if the content of the 'hello_world.txt' file is equal to 'Hello World'. + self.assertEqual( + content, "Hello World", f"Expected 'Hello World', got {content}" + ) + + +# Run the test case. +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_config.py b/tests/test_config.py index 52613b30..b472a24c 100644 --- a/tests/test_config.py +++ b/tests/test_config.py @@ -1,59 +1,84 @@ -import unittest +from unittest import TestCase from autogpt.config import Config -class TestConfig(unittest.TestCase): +class TestConfig(TestCase): + """ + Test cases for the Config class, which handles the configuration settings + for the AI and ensures it behaves as a singleton. + """ + + def setUp(self): + """ + Set up the test environment by creating an instance of the Config class. + """ + self.config = Config() + def test_singleton(self): - config1 = Config() + """ + Test if the Config class behaves as a singleton by ensuring that two instances are the same. + """ config2 = Config() - self.assertIs(config1, config2) + self.assertIs(self.config, config2) def test_initial_values(self): - config = Config() - self.assertFalse(config.debug_mode) - self.assertFalse(config.continuous_mode) - self.assertFalse(config.speak_mode) - self.assertEqual(config.fast_llm_model, "gpt-3.5-turbo") - self.assertEqual(config.smart_llm_model, "gpt-4") - self.assertEqual(config.fast_token_limit, 4000) - self.assertEqual(config.smart_token_limit, 8000) + """ + Test if the initial values of the Config class attributes are set correctly. + """ + self.assertFalse(self.config.debug_mode) + self.assertFalse(self.config.continuous_mode) + self.assertFalse(self.config.speak_mode) + self.assertEqual(self.config.fast_llm_model, "gpt-3.5-turbo") + self.assertEqual(self.config.smart_llm_model, "gpt-4") + self.assertEqual(self.config.fast_token_limit, 4000) + self.assertEqual(self.config.smart_token_limit, 8000) def test_set_continuous_mode(self): - config = Config() - config.set_continuous_mode(True) - self.assertTrue(config.continuous_mode) + """ + Test if the set_continuous_mode() method updates the continuous_mode attribute. + """ + self.config.set_continuous_mode(True) + self.assertTrue(self.config.continuous_mode) def test_set_speak_mode(self): - config = Config() - config.set_speak_mode(True) - self.assertTrue(config.speak_mode) + """ + Test if the set_speak_mode() method updates the speak_mode attribute. + """ + self.config.set_speak_mode(True) + self.assertTrue(self.config.speak_mode) def test_set_fast_llm_model(self): - config = Config() - config.set_fast_llm_model("gpt-3.5-turbo-test") - self.assertEqual(config.fast_llm_model, "gpt-3.5-turbo-test") + """ + Test if the set_fast_llm_model() method updates the fast_llm_model attribute. + """ + self.config.set_fast_llm_model("gpt-3.5-turbo-test") + self.assertEqual(self.config.fast_llm_model, "gpt-3.5-turbo-test") def test_set_smart_llm_model(self): - config = Config() - config.set_smart_llm_model("gpt-4-test") - self.assertEqual(config.smart_llm_model, "gpt-4-test") + """ + Test if the set_smart_llm_model() method updates the smart_llm_model attribute. + """ + self.config.set_smart_llm_model("gpt-4-test") + self.assertEqual(self.config.smart_llm_model, "gpt-4-test") def test_set_fast_token_limit(self): - config = Config() - config.set_fast_token_limit(5000) - self.assertEqual(config.fast_token_limit, 5000) + """ + Test if the set_fast_token_limit() method updates the fast_token_limit attribute. + """ + self.config.set_fast_token_limit(5000) + self.assertEqual(self.config.fast_token_limit, 5000) def test_set_smart_token_limit(self): - config = Config() - config.set_smart_token_limit(9000) - self.assertEqual(config.smart_token_limit, 9000) + """ + Test if the set_smart_token_limit() method updates the smart_token_limit attribute. + """ + self.config.set_smart_token_limit(9000) + self.assertEqual(self.config.smart_token_limit, 9000) def test_set_debug_mode(self): - config = Config() - config.set_debug_mode(True) - self.assertTrue(config.debug_mode) - - -if __name__ == "__main__": - unittest.main() + """ + Test if the set_debug_mode() method updates the debug_mode attribute. + """ + self.config.set_debug_mode(True) + self.assertTrue(self.config.debug_mode) diff --git a/tests/test_json_parser.py b/tests/test_json_parser.py index 7e3d4f51..2862034b 100644 --- a/tests/test_json_parser.py +++ b/tests/test_json_parser.py @@ -1,7 +1,7 @@ import unittest import tests.context -from autogpt.json_parser import fix_and_parse_json +from autogpt.json_fixes.parsing import fix_and_parse_json class TestParseJson(unittest.TestCase): diff --git a/tests/promptgenerator_tests.py b/tests/test_prompt_generator.py similarity index 67% rename from tests/promptgenerator_tests.py rename to tests/test_prompt_generator.py index b700af49..6a0bfd6c 100644 --- a/tests/promptgenerator_tests.py +++ b/tests/test_prompt_generator.py @@ -1,25 +1,35 @@ -# Import the required libraries for unit testing -import os -import sys -import unittest +from unittest import TestCase from autogpt.promptgenerator import PromptGenerator -# Create a test class for the PromptGenerator, subclassed from unittest.TestCase -class promptgenerator_tests(unittest.TestCase): - # Set up the initial state for each test method by creating an instance of PromptGenerator - def setUp(self): - self.generator = PromptGenerator() +class TestPromptGenerator(TestCase): + """ + Test cases for the PromptGenerator class, which is responsible for generating + prompts for the AI with constraints, commands, resources, and performance evaluations. + """ + + @classmethod + def setUpClass(cls): + """ + Set up the initial state for each test method by creating an instance of PromptGenerator. + """ + cls.generator = PromptGenerator() # Test whether the add_constraint() method adds a constraint to the generator's constraints list def test_add_constraint(self): + """ + Test if the add_constraint() method adds a constraint to the generator's constraints list. + """ constraint = "Constraint1" self.generator.add_constraint(constraint) self.assertIn(constraint, self.generator.constraints) # Test whether the add_command() method adds a command to the generator's commands list def test_add_command(self): + """ + Test if the add_command() method adds a command to the generator's commands list. + """ command_label = "Command Label" command_name = "command_name" args = {"arg1": "value1", "arg2": "value2"} @@ -31,20 +41,29 @@ class promptgenerator_tests(unittest.TestCase): } self.assertIn(command, self.generator.commands) - # Test whether the add_resource() method adds a resource to the generator's resources list def test_add_resource(self): + """ + Test if the add_resource() method adds a resource to the generator's resources list. + """ resource = "Resource1" self.generator.add_resource(resource) self.assertIn(resource, self.generator.resources) - # Test whether the add_performance_evaluation() method adds an evaluation to the generator's performance_evaluation list def test_add_performance_evaluation(self): + """ + Test if the add_performance_evaluation() method adds an evaluation to the generator's + performance_evaluation list. + """ evaluation = "Evaluation1" self.generator.add_performance_evaluation(evaluation) self.assertIn(evaluation, self.generator.performance_evaluation) - # Test whether the generate_prompt_string() method generates a prompt string with all the added constraints, commands, resources and evaluations def test_generate_prompt_string(self): + """ + Test if the generate_prompt_string() method generates a prompt string with all the added + constraints, commands, resources, and evaluations. + """ + # Define the test data constraints = ["Constraint1", "Constraint2"] commands = [ { @@ -61,7 +80,7 @@ class promptgenerator_tests(unittest.TestCase): resources = ["Resource1", "Resource2"] evaluations = ["Evaluation1", "Evaluation2"] - # Add all the constraints, commands, resources, and evaluations to the generator + # Add test data to the generator for constraint in constraints: self.generator.add_constraint(constraint) for command in commands: @@ -76,24 +95,20 @@ class promptgenerator_tests(unittest.TestCase): # Generate the prompt string and verify its correctness prompt_string = self.generator.generate_prompt_string() self.assertIsNotNone(prompt_string) + + # Check if all constraints, commands, resources, and evaluations are present in the prompt string for constraint in constraints: self.assertIn(constraint, prompt_string) for command in commands: self.assertIn(command["name"], prompt_string) - - # Check for each key-value pair in the command args dictionary - for key, value in command["args"].items(): - self.assertIn(f'"{key}": "{value}"', prompt_string) + for key, value in command["args"].items(): + self.assertIn(f'"{key}": "{value}"', prompt_string) for resource in resources: self.assertIn(resource, prompt_string) for evaluation in evaluations: self.assertIn(evaluation, prompt_string) + self.assertIn("constraints", prompt_string.lower()) self.assertIn("commands", prompt_string.lower()) self.assertIn("resources", prompt_string.lower()) self.assertIn("performance evaluation", prompt_string.lower()) - - -# Run the tests when this script is executed -if __name__ == "__main__": - unittest.main() diff --git a/tests/test_token_counter.py b/tests/test_token_counter.py new file mode 100644 index 00000000..81e68277 --- /dev/null +++ b/tests/test_token_counter.py @@ -0,0 +1,62 @@ +import unittest +import tests.context +from autogpt.token_counter import count_message_tokens, count_string_tokens + + +class TestTokenCounter(unittest.TestCase): + def test_count_message_tokens(self): + messages = [ + {"role": "user", "content": "Hello"}, + {"role": "assistant", "content": "Hi there!"}, + ] + self.assertEqual(count_message_tokens(messages), 17) + + def test_count_message_tokens_with_name(self): + messages = [ + {"role": "user", "content": "Hello", "name": "John"}, + {"role": "assistant", "content": "Hi there!"}, + ] + self.assertEqual(count_message_tokens(messages), 17) + + def test_count_message_tokens_empty_input(self): + self.assertEqual(count_message_tokens([]), 3) + + def test_count_message_tokens_invalid_model(self): + messages = [ + {"role": "user", "content": "Hello"}, + {"role": "assistant", "content": "Hi there!"}, + ] + with self.assertRaises(KeyError): + count_message_tokens(messages, model="invalid_model") + + def test_count_message_tokens_gpt_4(self): + messages = [ + {"role": "user", "content": "Hello"}, + {"role": "assistant", "content": "Hi there!"}, + ] + self.assertEqual(count_message_tokens(messages, model="gpt-4-0314"), 15) + + def test_count_string_tokens(self): + string = "Hello, world!" + self.assertEqual( + count_string_tokens(string, model_name="gpt-3.5-turbo-0301"), 4 + ) + + def test_count_string_tokens_empty_input(self): + self.assertEqual(count_string_tokens("", model_name="gpt-3.5-turbo-0301"), 0) + + def test_count_message_tokens_invalid_model(self): + messages = [ + {"role": "user", "content": "Hello"}, + {"role": "assistant", "content": "Hi there!"}, + ] + with self.assertRaises(NotImplementedError): + count_message_tokens(messages, model="invalid_model") + + def test_count_string_tokens_gpt_4(self): + string = "Hello, world!" + self.assertEqual(count_string_tokens(string, model_name="gpt-4-0314"), 4) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/unit/test_browse_scrape_text.py b/tests/unit/test_browse_scrape_text.py index be075ad7..98f5f558 100644 --- a/tests/unit/test_browse_scrape_text.py +++ b/tests/unit/test_browse_scrape_text.py @@ -2,7 +2,7 @@ import requests -from autogpt.browse import scrape_text +from autogpt.commands.web_requests import scrape_text """ Code Analysis diff --git a/tests/unit/test_commands.py b/tests/unit/test_commands.py new file mode 100644 index 00000000..21982f7e --- /dev/null +++ b/tests/unit/test_commands.py @@ -0,0 +1,18 @@ +import autogpt.agent.agent_manager as agent_manager +from autogpt.app import start_agent, list_agents +import unittest +from unittest.mock import patch, MagicMock + + +class TestCommands(unittest.TestCase): + def test_make_agent(self): + with patch("openai.ChatCompletion.create") as mock: + obj = MagicMock() + obj.response.choices[0].messages[0].content = "Test message" + mock.return_value = obj + start_agent("Test Agent", "chat", "Hello, how are you?", "gpt2") + agents = list_agents() + self.assertEqual("List of agents:\n0: chat", agents) + start_agent("Test Agent 2", "write", "Hello, how are you?", "gpt2") + agents = list_agents() + self.assertEqual("List of agents:\n0: chat\n1: write", agents)