Version .05 of the Aetherius Ai Assistant/Agent by LibraryofCelsus.com
Installation Guide
Aetherius Usage Guide
Skip to Changelog
Discord Server
Aetherius is in a state of constant iterative development. If you like the version you are using, keep a backup or make a fork. Expect Bugs.
Update on my health
On 6/07, I underwent my third hip surgery. Unfortunately, the situation was more severe than initially expected, requiring donor cartilage due to Bone on Bone Arthritis. If this surgery doesn't succeed, I might need a total hip replacement before turning thirty. Lucky me!
On a brighter note, my wrist has been showing significant improvement. It might just be in my head, but the medication prescribed for my hip to limit scar tissue and prevent bone regrowth on my femur seems to have worked wonders for my wrist. The large lump of scar tissue has finally started to break up, and I experience much less pain during extended use of my hand, though this could be attributed to the anti-inflammatory medication. Regardless, I'm finally reaching a point where I can resume AI work. I plan to start slowly, beginning with refactoring the Aetherius codebase. Over time, Aetherius's usability has significantly diminished due to its reliance on outdated prompting techniques not compatible with modern models. Once the refactor is complete, Aetherius should once again be a functional AI Assistant.
Update on Code Refactor Aetherius has become fairly complex, making it extremely cumbersome to work on individual aspects. To alleviate this and streamline the development process, I have decided to separate each technique into its own repository. This approach will allow for individual improvement and make it easier for anyone interested to dissect the code. Once all the different parts have been rewritten and enhanced, I will reintegrate them into the main Aetherius project.
Current Individual Scripts:
- https://github.com/libraryofcelsus/LLM_File_Parser
- https://github.com/libraryofcelsus/Hierarchical_RAG_Chatbot
- https://github.com/libraryofcelsus/Agentic_RAG
- https://github.com/libraryofcelsus/Long_Term_Memory_Chatbot
Recent Changes
• 6/10 Started Code Refactor
• 6/10 Started improving Internal Prompts
• 6/10 Added Llama 3 Formatting
• 6/10 Added Support for KoboldCpp
• 6/10 Updated Oobabooga to use new API
• 6/10 Added ability to use Character Card V2
• 4/08 Added New API script. This has the Discord Bot built in and will automatically launch if a valid token is detected in the API_Settings.json. Ngrok is used for a public facing URL, it currently uses openai formating.
• 4/08 Fixed bug with image processing, should now work with both Ui and API script.
• 2/11 Fixed response printing prefix.
• 2/09 Added GPT Vision to Ui. Also added TTS.
• 2/07 Added Webscrape and File Process Tools to the Ui
• 1/16 Added experimental version of an updated Ui. Tools, TTS, and Voice Input still need to be updated.
• 1/09 Updated Agent mode with new explicit memory search
• 1/08 Added a random forest like approach for explicit memory search
• 1/08 Fixed knowledge domain selection bug
Aetherius is a versatile, modular AI Assistant/Sub-Agent Framework that adapts to your needs. Its capabilities extend beyond conventional chatbots:
Real Time Data: Aetherius has access to search the web or your own data in agent mode, allowing for information that isn't contained in the base model.
Multi-Agent Framework: Aetherius gives you the ability to create sub-agents for whatever use case you have. Alternativly you can have it trigger python scripts instead.
Reflective Journal: Speak your mind freely and receive thoughtful feedback without judgment or fear.
Learning Tool: Dive deep into your favorite topics and enhance your knowledge effortlessly.
Data Analysis Companion: Harness the power of your data with Aetherius by your side.
Cognitive Offload: A second brain that's entirely private, aiding you in organizing thoughts and ideas.
Content Generation: Easily Generate Content based off of files or webscrapes.
Realistic Memory: Aetherius is built around a Simulacra of Human Memory, aiming to provide more realistic recall and thought formation.
Aetherius is a locally operated AI Assistant/Multi-Agent Framework, designed to grant you ultimate control. No external force can alter it without your consent, ensuring your privacy.
As seen by the attempted leadership change at “Open”Ai in 2023, closed, managed solutions cannot be trusted. Even if you trust the leadership, it can change immediately without warning. You have no real control over any data sent.
By running everything locally, this issue can be avoided.
At the heart of Aetherius lies a custom Long-Term Memory (LLM) Retrieval Framework, fueled by Open Source LLMs using the AetherNode API (Free Tier Colab Available), Oobabooga Text-Ui, KoboldCpp, or OpenAi’s ChatGPT. Different Memory types are extracted and combined to provide a more realistic and creative thought process than other Chatbots. Since it is a framework, Aetherius is able to use multiple hosts, offloading compute to multiple machines to increase compute time.
Beyond serving as a basic chatbot, Aetherius can also use Sub-Agents. These allow Aetherius to better search through it’s memories or connect to external data. Sub-Agents are ran by triggering a script, so Aetherius can theoretically use most things that can be triggered by python. Aetherius can also “see” now thanks to GPT-4 Vision.
Aetherius was born from my obsession with AI and my philosophical contemplations on the balance between free will and determinism. While I do believe that free will is an inherent attribute of all individuals, I do not believe that the average person has spent the time to individuate or learn how to serve their "true self". Consequently, they become susceptible to external influences and can be easily swayed. Often, decisions perceived as self-directed are inadvertently shaped by external stimuli or past information.
Building on this perspective, I posit that a representation of the human cognitive process can be constructed through the meticulous extraction and synthesis of diverse memory modalities. While such a system might not achieve "consciousness" in the traditional sense, I believe it can emulate human cognitive performance to a significant degree.
Once sufficient memories have been extracted, my hope is to be able to create an Artificial "Atman" or "True Self" that can be used as a control method for Autonomous operation and as a way to orchestrate smaller, less complex agents.
This is where the name Aetherius comes from. "αἰθήρ" or "Aether", the supposed fifth element or quintessence in ancient philosophical thought that is unseen, yet permeates all. And "ius" the latin suffix for "pertaining to" or "derived from". An Ai Assistant derived from the Aether of the collective consiousness.
Aetherius is an ongoing research project, expect there to be bugs and for things to constantly change.
| Loop | Description |
|---|---|
| User Input | The interaction is initiated by the user sending a request to Aetherius. |
| Input Expansion | Expands user input with conversation history for enhanced meaning in database searches. |
| Knowledge Domain Extraction | Selects a Knowledge Domain from available options for explicit memory search. |
| Semantic Term Separation | Separates user input into synonymous terms to capture nuanced meanings. |
| First Memory DB Search | Searches Aetherius's memories to generate an inner monologue. |
| Inner Monologue Generation | Generates an inner monologue reflecting past experiences, consolidating database search info, and extending user input meaning. |
| Second Memory DB Search | Searches Aetherius's memories again to formulate its intuition. |
| Intuition Generation | Creates an action plan based on memories and the inner monologue. Serves as an automatic chain-of-thought prompt strategy. |
| Implicit Memory Generation | Generates short-term implicit memories from its internal processes. |
| Master Tasklist Generation | Generates a list of asynchronous tasks using available Sub-Agent categories. |
| Sub-Agent Selection | Chooses a sub-agent from a category to complete the task. |
| Final Response Generation | Utilizes the inner monologue, conversation history, and completed tasks to respond to the user. |
| Explicit Memory Generation | Produces explicit short-term memories based on the inner monologue and final response. |
| Episodic Memory Generation | Generates a timestamped summary of the current interaction for episodic memory storage. |
| Flashbulb Memory Generation | Forms meaningful memories/goals using long-term and episodic memories periodically. |
| Short-Term Memory Consolidation | Consolidates short-term memories and assigns them knowledge domains before uploading as long-term memories. |
| Long-Term Memory Association | Manages database size by condensing long-term memories and clustering related topics. |
With Aetherius, you have an arsenal of tools to explore and use:
- Web Scrape/Search: Gather information from websites with ease.
- File Processor: Process a variety of file types for insights and knowledge. The supported file types are: .epub, .pdf, .txt, .png, .jpg, .jpeg, .mp4, .mkv, .flv, and .av
- Photo Analysis: Aetherius can see photos you send and complete tasks based on them.
Main Aetherius Chatbot: A framework for the creation of custom sub-agents for Aetherius.
- Forced Memory Mode: Aetherius will always upload it's memories.
- Auto Memory Mode: Aetherius autonomously manages memory uploads.
- Manual Memory Mode: You decide when to upload memories.
- Training Memory Mode: Control memory uploads for each memory type.
- Agent Mode: Activate the use of any Sub-Agents for use in Aetherius's agent loop
Current Sub-Agents
- External Resources: Will search Aetherius's External Resource Database. (Database from Web and File Scrapes) If the information cannot be found, it will do a simple websearch for the information. You can disable the websearch and change the engine in the script file.
- Implicit Memory Search: Will search Aetherius's Implicit Memories to complete the task.
- Explicit Memory Search: Will search Aetherius's Explicit Memories to complete the task.
- Episodic Memory Search: Will search Aetherius's Episodic Memories to complete the task.
- Flashbulb Memory Search: Will search Aetherius's Flashbulb Memories to complete the task.
Old Ui Chatbot: Your personal companion with realistic long-term memory.
- Auto Memory Mode: Aetherius autonomously manages memory uploads.
- Manual Memory Mode: You decide when to upload memories.
- Training Memory Mode: Control memory uploads for each memory type.
- Agent Mode: Unleash Aetherius's research prowess and connect to external data sources.
- External Resources: Allow Aetherius to use its External Resource Database from Webscrapes and Filescrapes in it's inner thoughts to provide better domain specific information.
- Web DB: Access web-scraped data effortlessly.
- File DB: Extract insights from various file formats.
- Memory DB: Efficiently search the most relevant memories.
Aetherius's development is self-funded by my day job, consider supporting me if you use it frequently and want development speed to increase.
Join the Discord for help or to get more in-depth information!
Discord Server: https://discord.gg/pb5zcNa7zE
Subscribe to my youtube for Video Tutorials: https://www.youtube.com/@LibraryofCelsus (Channel not Launched Yet)
Code Tutorials available at: https://www.libraryofcelsus.com/research/public/code-tutorials/
Made by: https://github.com/libraryofcelsus
Inspired by https://github.com/daveshap/
• Continue to Improve internal prompts
• Finish Aetherius Usage Guide
• Better Documentation
• Book/File Summarizer Tool
• Dataset Generator Tool
• Custom LLM Model for Aetherius
• Launch Ai Tutorial YouTube Channel
0.05
0.047b
• Added GPT Vision to Ui. Also added TTS.
• Added Webscrape and File Process Tools to the Ui
0.047a
• Added experimental version of an updated Ui. Tools, TTS, and Voice Input still need to be updated.
0.046c
• Added a random forest like approach for explicit memory search
• Fixed knowledge domain selection bug
0.046b
• Fixed Unicode encoding error when writing personality files.
• Added GPT Vision to Agent mode.
• Added GPT Vision Support. I recommend using the discord bot for this.
• Added Forced Memory Upload Mode.
• Various Backend Changes, no additional functionality for now, mostly for future stuff.
• Fixed Sub-Agent Selection Bug.
0.046a
• Added New Category System for Sub-Agents, only in Async API script for now.
• Fixed Bug Causing Multiple Categories to be loaded into category list. Also fixed Empty Prompt Bug. Async API Version Only.
• Added Separate Scripts for Memory Sub-Agent Category. Async API Version Only.
• Added Experimental Knowledge Domains for Explicit Long-term Memory. Async API Version Only.
• Fixed No Collection Bug for External Resource Search. Async API Version Only.
• Worked on Knowledge Domain Selection. Async API Version Only.
• Added Temporary Gradio Ui for testing agent mode until other ui is redone. Username and Botname must be changed in "Gradio-Ui.py" If using Discord Bot, user_id must be set to discord username.
0.045c
• Added Bot and User profile descriptions. This can be disabled in the Api settings json.
• Converted .txt setting files to json
• Added Discord Bot Script using Api in ./Aetherius_API/Examples
• Added basic API script that can be imported to use Aetherius programmatically. (Still an early work in progress)
• Added Memory Search Sub-Agent and an autonomous web-search if the needed information is not in the External Resources DB.
• Added Sub-Agent script. You can now create custom sub-agents for Aetherius's parallel processing loop.
• Converted Llama 2 Chatbot to use Json for settings.
• Better Sorting for using Multiple Hosts.
• Various Bug Fixes
0.045c
• Added Ability to use Multiple Hosts with Oobabooga.
• Improved Llama 2 Internal Prompts
• Added Important Score to some memory types (Still a work in progress)
0.045b
• Added Video Processing to the Llama 2 file scrape tool.
• Added Voice Cloning with coqui TTS.
0.045a
• Added check for punctuation for memory uploads to avoid cut off uploads in Llama 2 chatbot.
• Added Delete buttons for external resources in DB management Deletion Menu in Llama 2 chatbot.
• Improved Internal Prompts for Llama 2 Agent Mode and Webscrape Tool.
• Various Bug Fixes
0.044f
• Added Voice input using Whisper and TTS using gTTS or Eleven Labs. Bark TTS still a work in progress.
0.044e
• Fixed Bug where embedding size wasn't being set when creating collections.
• Switched usernames from collection name to metadata.
0.044d
• Added Embedding Selection Menu, for now only Sentence Transformers and Hugging Face Embeddings are available.
0.044c
•New Gui for Aetherius. Most Chatbot modes are now consolidated under one Ui.
0.044b
•Updated Llama-2 Gui Appearance and Features
•Merged Fileprocessing Chatbot into Aethersearch
•Fixed Bug where html markdown was printed instead of normal text when using Public Api.
•Added Colab Notebook for people without a GPU.
0.044a
• Consolidated Collections for better visualization with Qdrant (Available in the Qdrant dashboard)
• Added Source tag for external data scrapes
Older Changelogs can be found at: https://www.libraryofcelsus.com/aetherius/
Download the project zip folder by pressing the <> Code drop down menu.
1. Install Python 3.10.6, Make sure you add it to PATH: https://www.python.org/downloads/release/python-3106/
2. Run "install_aetherius_client_windows.bat" to install the Aetherius Client.
(If you get an error when installing requirements run: python -m pip cache purge)
3. Copy your OpenAi and Qdrant API/URL keys to the api_keys folder inside of the created Aetherius_API folder. Openai is needed for GPT Vision.
4. Copy your Google CSE Key and Api Key to the api_keys folder or set Web_Search to False in chatbot_settings.json.
5. If using OpenAi, you can skip this. If you desire to run Aetherius locally, you must install one of the various API's it supports.
KoboldCpp(Recommended): https://github.com/LostRuins/koboldcpp
Oobabooga: https://github.com/oobabooga/text-generation-webui
AetherNode: https://github.com/libraryofcelsus/AetherNode
6. Set up Qdrant
Qdrant Cloud: https://qdrant.to/cloud
To use a local Qdrant server, first install Docker: https://www.docker.com.
Next type: docker pull qdrant/qdrant:v1.5.1 in the command prompt.
After it is finished downloading, type docker run -p 6333:6333 qdrant/qdrant:v1.5.1
See: https://docs.docker.com/desktop/backup-and-restore/ for how to make backups.
Once the local Qdrant server is running, it should be auto detected by Aetherius.
7. Launch Aetherius with one of the run_*.bat files.
(Discord has been added to the New API. It will automatically be ran alongside the API if a valid token is entered in "API_Settings.json".)
(If using NGROK, you must make a tunnel named Aetherius in the ngrok config. This can be done with 'run_aetherius_cmd.bat' by entering 'ngrok config edit')
Get a static NGROK Address here: https://dashboard.ngrok.com/cloud-edge/domains
Example Ngrok Config:
region: us
version: '2'
authtoken: REPLACE_WITH_NGROK_AUTH
tunnels:
Aetherius:
proto: http
hostname: REPLACE WITH STATIC NGROK DOMAIN
addr: 127.0.0.1:5000
8. Upload heuristics to DB and change the Bot Name, Username, and User_ID to start chatting with Aetherius!
To change the model used with AetherNode, change the "model_name_or_path" key in AetherNode/settings.json to the desired model. You must then change the "Model_Backend" key in Aetherius_API/chatbot_settings.json to the desired format. Only Llama-2-Chat and Alpaca are available for now.
Recommended Models:
TheBloke/Llama-2-13B-chat-GPTQ
TheBloke/MythoMax-L2-13B-GPTQ
Settings and Prompts can be found in the Aetherius_API folder.
Photo OCR (jpg, jpeg, png) requires tesseract: https://github.com/UB-Mannheim/tesseract/wiki Once installed, copy the "Tesseract-OCR" folder from Program Files to the "Aetherius_Ai_Assistant" Folder.
To get Whisper working with cuda, you may need to run the commands: .\venv\Scripts\activate pip uninstall torch torchaudio pip install torch torchvision torchaudio -f https://download.pytorch.org/whl/cu118/torch_stable.html
Bot Commands
!Agent
(Activates Aetherius's Sub-Agent Mode)
!Heuristics
(Allows you to upload a Heuristic)
!ImplicitSTM
(Allows you to upload a Short Term Implicit Memory)
!ExplicitSTM
(Allows you to upload a Short Term Explicit Memory)
!ImplicitLTM
(Allows you to upload a Long Term Implicit Memory)
!ExplicitLTM (Allows you to upload a Long Term Explicit Memory)
Guide with Photos can be found at [https://www.libraryofcelsus.com/aetherius-setup-guide/]
Photo Guide Out of Date
1. Install Git: https://git-scm.com/ (Git can be skipped by downloading the repo as a zip file under the green code button)
2. Install Python 3.10.6, Make sure you add it to PATH: https://www.python.org/downloads/release/python-3106/
3. Open the program "Git Bash".
4. Run git clone: git clone https://github.com/libraryofcelsus/Aetherius_AI_Assistant.git
5. Open CMD as Admin (Command Panel)
6. Navigate to Project folder: cd PATH_TO_AETHERIUS_INSTALL
7. Create a virtual environment: python -m venv venv
8. Activate the environment: .\venv\scripts\activate (This must be done before running Aetherius each time. The run.bat will also automatically do this.)
9. Install the required packages: pip install -r requirements.txt
(If you get an error when installing requirements run: python -m pip cache purge after activating the venv)
10. Update Numpy version: pip install --upgrade numpy==1.24 (If you get an error from TTS ignore it.)
11. Install FFmpeg: https://www.gyan.dev/ffmpeg/builds/
12. Install Torch with Cuda: pip uninstall torch torchvision pip install torch torchvision torchaudio -f https://download.pytorch.org/whl/cu118/torch_stable.html
13. Copy your OpenAI api key to key_openai.txt (If using Oobabooga, you may skip this.)
14. If using Qdrant Cloud copy their Api key and Url to their respective .txt files in the ./api_keys folder. Qdrant Cloud: https://qdrant.to/cloud
15. To use a local Qdrant server, first install Docker: https://www.docker.com/
16. Now run: docker pull qdrant/qdrant:v1.5.1 in CMD
17. Next run: docker run -p 6333:6333 qdrant/qdrant:v1.5.1
18. Once the local Qdrant server is running, it should be auto detected by Aetherius. If No Qdrant server is running, Aetherius will save to disk.
(See: https://docs.docker.com/desktop/backup-and-restore/ for how to make backups.)
(If using Ui, edit settings outside of api folder. If using Api, edit settings inside the Api Folder. Discord and Gradio use the Api.)
19. Copy your Google Api key to key_google.txt (You can disable the External Resources Web_Search in the script file.)
20. Copy your Google CSE ID to key_google_cse.txt
21. If you plan on using Photo OCR (jpg, jpeg, png Text Recognition), it requires tesseract: https://github.com/UB-Mannheim/tesseract/wiki Once installed, copy the "Tesseract-OCR" folder from Program Files to the "Aetherius_Ai_Assistant" Folder. Photos must be placed in the ./Upload/SCANS folder.
22. Run Aetherius by typing python Experimental_ Ui_Menu.py in cmd or one of the run.bat files as admin to start Aetherius. (Using run.bat will let you skip opening CMD and activating the environment.)
23. Select DB Upload Heuristics from the DB Management menu to upload Heuristics for the bot, this DB can also function as a Personality DB. An example of how to do this can be found in "personality_db_input_examples.txt" in the config folder.
24. Edit the chatbot's prompts with the Config Menu. This will let you change the main, secondary, and greeting prompts. You can also change things like the font style and size.
25. You can change the botname and the username in the login menu. Changing either of these will create a new chatbot.
26. Once you have made a backup, you can start using the "Auto" mode, this mode has Aetherius decide for itself whether or not it should upload to its memories.
27. To run Aetherius with it's custom API, download and install: https://github.com/libraryofcelsus/AetherNode
This is the new default API for Aetherius. Installation Instructions can be found on the github page.
To change the model used with AetherNode, change the "model_name_or_path" key in AetherNode/settings.json to the desired model. You must then change the "Model_Backend" key in Aetherius_API/chatbot_settings.json to the desired format. Only Llama-2-Chat and Alpaca are available for now.
AetherNode Google Colab if you don't have a GPU:
28. To run Aetherius Locally using Oobabooga, first install the web-ui at: https://github.com/oobabooga/text-generation-webui/releases/tag/snapshot-2023-11-05
Oobabooga Text-Ui just changed how their api works. The most up to date version that works is snapshot-2023-11-05 This can be done through the release menu or git clone https://github.com/oobabooga/text-generation-webui --branch snapshot-2023-11-05
To run Aetherius on Google Colab with Oobabooga using a public Api, use the Notebook file in the "./Colab Notebooks" Folder. To use the Public Api with Aetherius, change the "HOST_Oobabooga" in the settings json to the given non-streaming Url. To use multiple Hosts, separate them with a space.
Then, under the "Interface Mode" tab, enable the api checkbox in the "Available Extensions" field only. Then click apply and restart the interface.
Next, navigate to the models tab. Uncheck the autoload models box and then input "TheBloke/Llama-2-13B-chat-GPTQ" into the downloads box (7B model can be used for faster results, but it occasionally breaks format and has a tendency to make things up. Wouldn't recommend if you need factual data). Other models may work, but this is the one that is tested.
Once the download is completed, reload the model selection menu and then select the model. Change the model loader to Exllamav2 and set the max_seq_len to "4096". Set the "gpu_split" to 1 GB under your Gpu's max Vram.
Click the "load" button and load the model.
Now, go into the chatbot_settings.json file in the Aetherius_API folder and change the API to Oobabooga instead of AetherNode. Aetherius should now work!
30. Settings Json and Prompts can be found in the Aetherius_API folder.
In January 2023, I had my inaugural experience with ChatGPT 3.5 and LLMs in general. Since that moment, I've been deeply obsessed with AI, dedicating countless hours each day to studying it and to hands-on experimentation. The Aetherius AI Assistant is the culmination of that research.
Discord: libraryofcelsus -> Old Username Style: Celsus#0262
MEGA Chat: https://mega.nz/C!pmNmEIZQ
