silentis-ai 1.2.3

Silentis AI - A powerful AI assistant plugin.

Silentis AI - Core

Silentis Open Source Core Python Plugin
MIT License | Version 1.2.2

A flexible AI plugin system for interacting with multiple large language models, featuring:
✅ Multi-model support (Reasoner, Llama, DeepSeek, Phi-3)
✅ GPU acceleration toggle
✅ Customizable parameters (temperature, max tokens, Top-P, system prompt, instructions)
✅ RAM requirement checks
✅ Persistent configuration
✅ Interactive chat interface (terminal and web-based)
✅ Web API for browser interaction

---

Features

• 🤖 Supported Models:

  • Reasoner v1 (8GB RAM) - Advanced reasoning
  • Llama 3 8B Instruct (8GB RAM) - Instruction specialist
  • DeepSeek-R1-Distill-Qwen-7B (8GB RAM) - Deep insights
  • Phi-3 Mini Instruct (4GB RAM) - Lightweight responses

• ⚙️ Configurable Parameters:

  • Temperature (0-1)
  • Max output tokens (1-1000)
  • Top-P probability
  • System Prompt
  • Instructions (persistent AI behavior customization)
  • GPU usage toggle
  • API settings (host, port, enable/disable)

• 🌐 Web Interface:

  • Multi-page layout (Chat, Settings, Status)
  • Accessible at http://localhost:5000 when API is enabled

---

Installation

From Pypi:

pip install silentis-ai

OR:

1. Clone the repository:

   git clone https://github.com/silentisai/silentis.git
   cd silentis
   

2. Install dependencies:

   pip install -r requirements.txt
   

   Note: Ensure you have llama-cpp-python, flask, flask-cors, requests, and psutil installed. Create a requirements.txt with:

   llama-cpp-python
   flask
   flask-cors
   requests
   psutil
   

3. (Optional) Configure GPU support:

   # Install CUDA toolkit for GPU acceleration
   # Follow instructions at https://developer.nvidia.com/cuda-downloads
   # Then install llama-cpp-python with GPU support:
   pip install llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cu121
   

4. Ensure index.html is in the same directory as silentis.py.

---

Usage

Terminal Interface

1. Run the plugin:

   python silentis.py
   

2. If show_welcome is enabled (default), you’ll see the welcome message:

   ______     __     __         ______     __   __     ______   __     ______
   /\  ___\   /\ \   /\ \       /\  ___\   /\ "-.\ \   /\__  _\ /\ \   /\  ___\   
   \ \___  \  \ \ \  \ \ \____  \ \  __\   \ \ \-.  \  \/_/\ \/ \ \ \  \ \___  \  
    \/\_____\  \ \_\  \ \_____\  \ \_____\  \ \_\\"\_\    \ \_\  \ \_\  \/\_____\ 
     \/_____/   \/_/   \/_____/   \/_____/   \/_/ \/_/     \/_/   \/_/   \/_____/ 
   Silentis AI - Python Plugin
   Developed by: Silentis Team
   MIT License | Version 1.0
   ...
   

3. Main Menu (if disable_model_selection is False):

   --- Main Menu ---
   1: Chat
   2: Settings
   0: Exit
   Enter your choice:
   

   • 1: Chat: Start an interactive chat session with the loaded model. Type quit to return to the menu.
   • 2: Settings: Configure models and parameters (see below).
   • 0: Exit: Close the plugin.

4. Settings Menu:

   --- Settings Menu ---
   1: Load Model
   2: Update Model Parameters
   0: Back to Main Menu
   

   • 1: Load Model: Choose a model to download and load.
   • 2: Update Model Parameters: Adjust settings like temperature, max tokens, instructions, etc.

5. Direct Chat Mode (if disable_model_selection is True and a model is selected):

   • Loads the default model and starts the chat immediately.

6. Example Chat:

   --- Chat Started ---
   Type 'quit' to return to main menu.
   > Hello, how are you?
   Thinking...
   Assistant: I'm Silentis, doing great! How can I assist you today?
   > quit
   Returning to main menu...
   

Web Interface (API Mode)

1. Enable the API:

   • Edit config.json to set "api_enabled": true (or use the Settings menu in the terminal).
   • Default host: 0.0.0.0, port: 5000.

2. Run the plugin:

   python silentis.py
   

   • You’ll see: Starting API server on 0.0.0.0:5000.

3. Open your browser and navigate to:

   http://localhost:5000
   

4. Web Interface Layout:

   • Sidebar:
     • Chat (default page)
     • Settings
     • Status
   • Chat Page:
     • Chat window to send prompts and view responses.
     • Enter a prompt and click "Send".
   • Settings Page:
     • Available Models: List supported models.
     • Load Model: Select and load a model.
     • Configuration: Adjust temperature, max tokens, Top-P, system prompt, and instructions. Click "Apply Config" to save.
   • Status Page:
     • Check model status, selected model, and available RAM.

5. Example Web Interaction:

   • Go to "Settings", set "Instructions" to "Act as a coding assistant", and click "Apply Config".
   • Switch to "Chat", type "Write a Python function", and see a code-focused response.

---

Configuration

Modify config.json or use the terminal/web settings to customize Silentis AI. Changes persist across sessions.

Available Settings

Setting	Description	Range/Options	Default
system_prompt	Base prompt for AI behavior	Text	"You are Silentis..."
instructions	Additional instructions for AI behavior	Text	""
model_params.temp	Response randomness	0.0 - 1.0	0.7
model_params.max_tokens	Max response length	1 - 1000	50
model_params.top_p	Token probability threshold	0.0 - 1.0	0.9
use_gpu	Enable GPU acceleration	True/False	False
selected_model	Default model number	1, 2, 3, 4, or null	null
show_welcome	Show welcome message on start	True/False	True
disable_model_selection	Skip model selection on start	True/False	False
api_enabled	Enable web API	True/False	False
api_host	API server host	IP address	"0.0.0.0"
api_port	API server port	1 - 65535	5000

Terminal Configuration Example

--- Update Model Parameters ---
Current settings: Temp=0.7, Max Tokens=50, Top-P=0.9, GPU=Disabled, Show Welcome=Enabled, ...
Enter Temperature (0-1, default 0.7): 0.8
Enter Max Tokens (1-1000, default 50): 100
Enter Top-P (0-1, default 0.9): 
Enable GPU? (y/n, default No): n
Show Welcome Message? (y/n, default Yes): y
Disable Model Selection? (y/n, default No): y
Please select a default model:
[1] Reasoner v1 (8GB RAM) - Advanced reasoning & logic
...
Enter the number of the default model: 2
Default model set to: Llama 3 8B Instruct
API Enabled? (y/n, default No): y
API Host (default 0.0.0.0): 
API Port (1-65535, default 5000): 
Enter Instructions (or press Enter to keep current): Act as a coding assistant
Settings updated. Restart model for GPU changes to take effect.

Web Configuration

• Navigate to http://localhost:5000, click "Settings" in the sidebar, and adjust values under "Configuration". Click "Apply Config" to save.

---

API Endpoints

When api_enabled is True, the following endpoints are available at http://localhost:5000/api/:

Endpoint	Method	Description	Request Body Example	Response Example
/models	GET	List supported models	-	{"1": {"name": "Reasoner v1", ...}, ...}
/load_model	POST	Load a model	{"model_number": 2}	{"message": "Loaded Llama 3 8B..."}
/chat	POST	Send a chat prompt	{"prompt": "Hello"}	{"response": "Hi there!"}
/config	GET	Get current config	-	{"system_prompt": "...", "instructions": "..."}
/config	POST	Update config	{"model_params": {"temp": 0.8}, "instructions": "Code assistant"}	{"message": "Config updated successfully"}
/status	GET	Check system status	-	{"model_loaded": true, "available_ram": 16.2}

---

Knowledge Base

The plugin downloads models from HuggingFace. Below are the supported models:

Model Name	RAM Required	Description
Reasoner v1	8GB	Advanced reasoning and logic
Llama 3 8B Instruct	8GB	Instruction execution specialist
DeepSeek-R1-Distill-Qwen	8GB	Deep analysis and insights
Phi-3 Mini Instruct	4GB	Lightweight quick responses

---

Acknowledgments

• Model weights from HuggingFace
• Powered by llama-cpp-python
• Web framework: Flask

Note: Ensure sufficient RAM (4GB or 8GB depending on the model). GPU support requires CUDA setup.

---

License

MIT License
Copyright (c) 2025 - Silentis AI

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

---