Llmmaster

Added
- Pika.art third-party Text-To-Video model class
- `PikaPikaPikaGeneration`

Added
- `ElevenLabsTextToSpeech`

Added
- support Meshy 3D modeling APIs
- `MeshyTextToTexture`
- `MeshyTextTo3D`
- `MeshyTextTo3DRefine`
- `MeshyTextToVoxel`
- `MeshyImageTo3D`

Changed
- return value changed to `requests.models.Response` class from `response.content` for the following models:
- `StableDiffusionTextToImage`
- `StableDiffusionImageToVideo`
- `StableDiffusionImageToImage`
Added
- `RESULTSTYPE.md` for brief description how to handle generated contents

Added
- Speech-To-Text model class
- `GoogleSpeechToText`

Added
- Image-To-Video model class
- `StableDiffusionImageToVideo`
Changed
- Renamed `OpenAIAudioToText` to `OpenAISpeechToText`
- Renamed `INSTANCE_CLASSES` to `ACTIVE_MODELS` in `llmmaster.py`
- Changed `MAX_TOKENS` to `DEFAULT_TOKENS` in `config.py`
- LLMMaster no longer checks `max_tokens` and sets default if not given properly