Full Release v1.0

Instructions:

• Must be on Debian-based Linux (Ubuntu, Debian, Linux Mint)
• Download source code (zip) and extract
• Port forward network on port 8000
• Run ./setup.sh in the root of the project files
• When the server is running, connect to https://[your-ip-here]:8000/
• Default admin username: admin, default password: password1234

Test Deployment Build v0.3 Fast Response Time

Changelog:

• Response time has been massively improved. It now responds in about 6-12 seconds on an NVIDIA GeForce RTX 3060 with 8 GB of VRAM using Qwen3-4B.
• Hotfix to add user information instead of dummy information.

Instructions:

• Must be on Debian-based Linux (Ubuntu, Debian, Linux Mint)
• Download source code (zip) and extract
• Port forward network on port 8000
• Run ./setup.sh in the root of the project files
• When the server is running, connect to https://[your-ip-here]:8000/
• Default admin username: admin, default password: password1234

Notes:

• Response time fix only applies to Qwen module currently. This will be fixed soon. Mistral and Llama will still take a long time to respond.
• First time opening the chat page after launching the program will take a long time, this is intentional. Launching the chat page again will be instant.
• Known issue: Session creation is buggy. Try to stick to one session for now.
• Known issue: Submitting a quiz will say "No active quiz" and won't give you the answer key. Simply reopen the session and it'll show you the answer key.
• Known issue: System prompt exposed when you switch sessions.

TestDeployment Build v0.2.1 - RAG

Hotfix: Setup.sh didn't install django or other requirements properly

Test Deployment Build v0.2.1- No RAG

Hotfix: Error with requirements not being installed by setup.sh

Test Deployment Build v0.2 - No RAG

Changes:

• Model now uses Qwen by default. To change to Mistral or LLaMA, edit "LLM_MODULE" in settings.py to be either "mistral" or "llama" instead of "qwen".
• RAG has been updated to be more efficient (This version has RAG disabled; check the other DeploymentTest-V2 release for the RAG version)
• Data export implemented for CSV and JSON
• Signup now has a name box

Instructions:

• Must be on Debian-based Linux (Ubuntu, Debian, Linux Mint)
• Download source code (zip) and extract
• Port forward network on port 8000
• Run ./setup.sh in the root of the project files
• When the server is running, connect to https://[your-ip-here]:8000/
• Default admin username: admin, default password: password1234

Test Deployment Build v0.1

Instructions:

• Must be on Debian-based Linux (Ubuntu, Debian, Linux Mint)
• Download source code (zip) and extract
• Port forward network on port 8000
• Run ./setup.sh in the root of the project files
• When the server is running, connect to https://[your-ip-here]:8000/
• Default admin username: admin, default password: password1234

Test Deployment Build v0.2 - RAG Included

Changes:

• Model now uses Qwen by default. To change to Mistral or LLaMA, edit "LLM_MODULE" in settings.py to be either "mistral" or "llama" instead of "qwen".
• RAG has been updated to be more efficient
• Data export implemented for CSV and JSON
• Signup now has a name box

Instructions:

• Must be on Debian-based Linux (Ubuntu, Debian, Linux Mint)
• Download source code (zip) and extract
• Port forward network on port 8000
• Run ./setup.sh in the root of the project files
• When the server is running, connect to https://[your-ip-here]:8000/
• Default admin username: admin, default password: password1234