Antonis Makropoulos

In the GPU argument parsing of the llama.cpp server, the FLAG_gpu is considered an error if it is equal to -1. However the correct value of error should be -2...

Based on the Sample FSDP Run output provided (and the timings) the block_size should be 128 instead of 12

Models not using espeak

2

Hi, thank you for this awesome library 💥! I'd like to use it in the Unity engine but I'm concerned about the license issue (#93). Are there any good quality...

llama.cpp integration with DLL

6

### Describe the feature LLM for Unity uses the llamafile server for the LLM functionality. This approach can't be used for mobile integrations due to security limitations. Purpose of this...

Test Windows IL2CPP

2

Test if the Process works for Windows IL2CPP: https://forum.unity.com/threads/solved-il2cpp-and-process-start.533988/ otherwise implement a specific solution like: https://github.com/josh4364/IL2cppStartProcess/tree/master

Add Android support

8

Purpose of this features is to add support of LLMUnity for Android Issues: - llamafile does not support Android, the llama.cpp server binary will need to be separately compiled -...

Persist user interaction to disk so that it can be continued

1

llama.cpp allows to store the conversation along with the processed tokens to disk so that the interaction can be resumed at the next run. Purpose of this feature is to...

### Describe the feature Integration with the Playmaker Unity plugin: https://assetstore.unity.com/packages/tools/visual-scripting/playmaker-368