OuteTTS Lib v0.4

OuteTTS Lib v0.4 Release Notes

Interface Improvements

• Consolidated all interface versions into a single interface.py file for centralized management
• Implemented isolated model handling in separate version folders while maintaining core functionality for cross-compatibility
• Added Interface Version 3 implementation to support OuteTTS v1.0 models

New Features

• Smart text chunking for generating long audio clips from large text inputs
• Added DAC interface code to handle OuteTTS 1.0 audio encoding and decoding
• Added metadata for interface version compatibility in speaker files

Transformers Backend Patch for OuteTTS 1.0

• Implemented windowed repetition penalty processor (RepetitionPenaltyLogitsProcessorPatch) for improved text generation quality
• Applies penalties only to recent tokens (64-token window) rather than full context
• Addresses key quality issues in speech synthesis applications
• Maintains backward compatibility with standard HuggingFace interfaces

Streamlined Usage

Simplified code usage with a more modular and compact implementation:

output = interface.generate(
    config=outetts.GenerationConfig(
        text="Hello, how are you doing?",
        generation_type=outetts.GenerationType.CHUNKED,
        speaker=speaker,
        sampler_config=outetts.SamplerConfig(
            temperature=0.4
            # Additional sampler parameters
        ),
    )
)

Automatic Configuration

Added support for automatic config and model loading for v1.0 models:

# Auto-configuration approach
interface = outetts.Interface(
    config=outetts.ModelConfig.auto_config(
        model=outetts.Models.VERSION_1_0_SIZE_1B,
        backend=outetts.Backend.LLAMACPP,
        quantization=outetts.LlamaCppQuantization.FP16
    )
)

Manual configuration remains available:

# Manual configuration approach
interface = outetts.Interface(
    config=outetts.ModelConfig(
        model_path="...",
        tokenizer_path="...",
        backend=outetts.Backend.LLAMACPP,
        interface_version=outetts.InterfaceVersion.V3
    )
)

Performance and Dependencies

• Improved loading times by dynamically loading only required components
• Removed unused dependencies (further optimizations pending, particularly for WavTokenizer implementation)

Documentation

Full usage documentation is available at:
🔗 interface_usage.md

OuteTTS v0.3.2

Update 0.3

• Implement v2 interface with simplified structure.
• Split documentation for interface v1 and interface v2.
• Add compatibility for OuteTTS-0.3 1B and 500M models.
• Restructure codebase for better maintainability.

OuteTTS v0.2.3

Release Notes v0.2.3

• Split WavTokenizer into encoder (82MB) and decoder (248MB) components
• [WIP] Streaming support

OuteTTS v0.2.1

Release Notes v0.2.1

New Features and Improvements:

1. Support for ExLlamaV2

   • Integrated support for ExLlamaV2
   • Pull request: #37

2. Whisper Integration for Speaker Generation

   • Added Whisper-based transcription for generating speakers when no transcript is provided.
   • Suggested in: #28
   • Now, if transcript is set to None, the text will be automatically transcribed using Whisper.

   def create_speaker(
       self, 
       audio_path: str, 
       transcript: str = None, 
       whisper_model: str = "turbo",
       whisper_device = None
   )

OuteTTS v0.2.0 Release

OuteTTS v0.2.0 Release Notes

Major Changes

• New Model Support: Added support for OuteTTS-0.2-500M model
• Speaker Management: Introduced default speaker presets for each supported language
• Breaking Changes:
  • Speaker files from previous versions (<0.2.0) are not compatible
  • Interface usage has been significantly revised (see README.md for new implementation)

New Features

• Added voice cloning guidelines and interface usage recommendations in README.md
• Implemented Gradio example playground for OuteTTS-0.2-500M
• Multi-language alignment support
• Enhanced speaker management:
  • New methods: interface.print_default_speakers() and interface.load_default_speaker(name="male_1")
  • Switched from pickle to JSON format for speaker saving
  • Added speaker language information in saved files
• Option to load WavTokenizer from custom path (resolves issue #24)
• Multiple interface version initialization in a single function

Improvements

• Restructured library files for better organization
• Implemented hash verification for WavTokenizer downloads (resolves issue #3)
• Reworked interface for better usability
• Made sounddevice optional with improved error handling for sound playback
• Added data preparation examples for training

Error Handling

• Added validation for audio token detection
• Improved error messages for long input text and early EOS cases
• Enhanced overall library error handling and feedback

How to Upgrade

• Update your library via pip:

  pip install --upgrade outetts