fix: Update deprecated Groq models to recommended replacements

[mccullya]

This PR updates references to deprecated Groq LLM models with their recommended replacements as documented in Groq's deprecation notice: https://console.groq.com/docs/deprecations#deprecation-history

Changes Updated deprecated llama3-70b-8192 and llama3-8b-8192 models to use the recommended replacements: llama-3.3-70b-versatile meta-llama/llama-4-scout-17b-16e-instruct Also updated references to deprecated llama-3.1-70b-versatile to the newer llama-3.3-70b-versatile Updated deprecated qwen-2.5-32b to qwen-qwq-32b These changes ensure the application continues to function correctly as Groq deprecates older models according to their deprecation timeline, as the deprecated models will eventually return errors.

Files Changed Configuration files (config.py) Utility functions (utils.py) Documentation files and examples Example code Testing Code changes reflect Groq's official model deprecation recommendations. The updated models maintain the same functionality but provide better performance and reliability.