lm studio

[Shehab-Ecstasyandfire]

the instructions aren't clear on how to run it on lm studio server

[greggft]

According to Mervin Praison in his youtub video "RawDog: How I Cut My Work Hours in Half with Context-Aware AI?" (https://www.youtube.com/watch?v=5r86vtZcvJ4)
rawdog --llm-base-url http://:1234/v1 --llm-custom-provider openai --llm-model

This forces a setting of openai key because the provider is openai

Due to a "bug" with rawdog you can NOT specify it on the command line, it will still complain, even if you specify the key on the command line

To make this work you must set the environment variable for openai key to get this to work
[change for your operating system]
export OPENAI_API_KEY=XXXXXXXXXXXXXXXXXXXXXXXXX
The number of characters I believe doesn't matter

The rawdog command should work now

example:
export OPENAI_API_KEY=XXXXXXXXXXXXXXXXXXXXXXXXX
rawdog --retries 1 --llm-base-url http://192.168.1.38:1234/v1 --llm-custom-provider openai --llm-model Orenguteng/Llama-3.1-8B-Lexi-Uncensored-GGUF count the number of files in the directory