Diagnostics fails on last step on vanilla macOS install

[machineglow]

Hey guys,

Just discovered this project and tried it out. Install went smoothly but the diagnostics fails on last step.

Starting diagnostic tests...
========================================
DREAM SERVER - FUNCTIONAL TESTS
Tests actual functionality, not ports
========================================
> Testing LLM Functional Generation
�[0;31m✗�[0m LLM returned no response
> Testing TTS Audio Generation
�[0;31m✗�[0m TTS returned HTTP 000
> Testing Embeddings Vector Generation
�[0;31m✗�[0m Embeddings returned no response
> Testing Whisper Transcription
�[1;33m⚠�[0m Could not generate test audio for Whisper
�[1;33m⚠�[0m Skipping Whisper functional test (TTS dependency)
========================================
Results: 0 passed, 3 failed
========================================
�[0;31m✗ Some functional tests failed�[0m
Some tests failed.

I checked the speeches.ai front end. which shows a blank in the model drop down but didn't really know how to fix it. Does it auto download the models on first use or something?

Any advice?

Thanks!