Add: new feature to enable Kokoro TTS voice playback in chat messages

[mpetruc]

• Added TTS server configuration to TextToSpeech component in Settings page
• Fixed voice type checking and server config passing for Kokoro voices
• Ensured both browser and Kokoro voices work in chat messages

[olegshulyakov]

@mpetruc There are some changes needed:

• src/types/provider.ts
  You need to introduce types VoiceProvider and TextToSpeechProvider.
• src/types/voice.ts
  Create types for OpenAI Text to Speech like src/types/inference.ts does for inference.
• Create BaseVoiceOpenAIProvider in src/api/providers implementing LLMProvider, VoiceProvider and TextToSpeechProvider. Use can use src/api/providers/BaseOpenAIProvider.ts as reference.
• Add ability to switch between browser and LLM voice generation in Settings -> Voice.
• Create src/components/LLMTextToSpeech.tsx with all logic there.
• src/config/config-default.json should contain ttsBaseUrl and ttsApiKey

I hope it will help https://github.com/openai/openai-node/tree/master/src/resources/audio