Results 2 issues of Celestial Rose

This PR addresses a critical bug in the implementation that caused non-output when streaming responses from models like Llama4 that can return numeric chunks. The issue stemmed from an inadequate...

Handle base64 image data for Llama 4 in chat messages. As per the documentation found here https://developers.cloudflare.com/workers-ai/models/llama-4-scout-17b-16e-instruct/