AppWish 2.1 #196
pwgit-create
announced in
Announcements
AppWish 2.1
#196
Replies: 2 comments
-
The install script has been updated to ensure that the Ollama Serve command runs in the background. The installation script's efficiency will be enhanced. There have been no modifications made to any other files in the release archive. |
Beta Was this translation helpful? Give feedback.
0 replies
-
Good news, the Raspberry Pi 5 version (ARM) is now available for this release :) |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
AppWish 2.1
News
• A feature has been added that allows you to place external libraries (jars) in a designated folder for the purpose of compiling more advanced applications. To enhance security, the default mode is set to off.
Please note that you need to set USE_EXTERNAL_LIBRARIES from off to "on" to enable external libraries.
Ollama-config
USE_EXTERNAL_LIBRARIES=on
• The installation script has been improved/fixed.
• In the AI output, there should be fewer references to code comments when continuing an existing application.
Install AppWish Ollama
Start Appwish Ollama
Linux x64 amd
Model: codestral:22b-v0.1-q5_0
Java: In case Java 19 isn't already installed, it'll be installed using apt.
Raspberry Pie 5
Model: gemma2:2b
Java: JDK21
Comment: Our only available Java option for the Raspberry OS is Java 17, so we use bellsoft-jdk21.0.4+9-linux-aarch64.deb instead
Java 19 is needed for the jar to run, but you can create code that runs on Java 21 (such as KEM Apps).
Considerations regarding the Raspberry Pie 5 version:
The model that we use is not as good as CodeStral for code base generation, but it's awesome for generating single apps. I can run it flawlessly on Raspberry Pie 5 with 8gb of ram :) Dont except as good results as you would have with AMD version running on wsl on a decent pc though 😉
Proper cooling:
It would be great if you could review this video and post about cooling tips for a Raspberry Pie 5. The instructions on proper cooling for your Raspberry Pi 5 are essential since you will be running your own AI server on it. Running your own AI server is advantageous because you can generate your code offline. It's easy to set everything up because the install script installs everything for you.
There are default external libraries included
To use external_libraries, please ensure that external_libraries are turned on (see the text on top of this release).
There are three libraries included by default. At the mvn repository, the licenses can be viewed.
https://mvnrepository.com/artifact/com.google.code.gson/gson/2.4
https://mvnrepository.com/artifact/org.json/json
https://mvnrepository.com/artifact/org.seleniumhq.selenium/selenium-java
This is always included in the installation script, no matter if it's for the arm or x64 editions
Note: If you already have Java or Curl on your system, the install script won't install them. By running the install script again, it is also possible to update the ollama version. Even if you have already installed Ollama or AppWish, you can still do this and it won't cause any issues.
This discussion was created from the release AppWish 2.1.
Beta Was this translation helpful? Give feedback.
All reactions