Blue Moon Release

[pwgit-create]

Release Notes 1.7.2

There have been three additional AI /ollama configurations that can be edited within the same file. Those are:

• TOP_K=40
• NUM_PREDICT=-1
• TEMPERATURE=0.8

Install AppWish Ollama

                          ./install_ollama.sh

Start Appwish Ollama

                         sudo  java -jar appwish.jar

Do you have less than 8gb of ram?

If you plan on running the Linux AMD x64 version with less than 8GB of ram, you may experience slow response times from the AI models. In order to achieve faster speeds, consider running a lighter model (llama3 8b). The raspberry pie version already has that model defaulted to.

How can I change the model to LLama 3 for the Linux AMD X64 version?

1. Run the install script that installed ollama before typing ollama pull llama3:latest in your terminal.

2. Edit the text in the file using the path: src/main/resources/ollama_model.props
   From
   MODEL_NAME=codestral:22b
   Into
   MODEL_NAME=llama3:latest

Helper script for Windows Subsystem for Linux (WSL)

This script is designed to help you run Appwish Ollama using WSL (wsl_helper_script.sh).

If you have a decent Nvidia GPU, you can run NvidiaCUDA with WSL without much setup. If you're looking for a very fast app generation, this option is a good choice.

Running this script is not recommended if you have no intention of using WSL.

Have fun with the release and generate apps in a responsible manner. 🐲 🔮 🌌

---

This discussion was created from the release Blue Moon Release.