| # TTS Model Optimization Report |
|
|
| ## Overview |
|
|
| - **Optimization Date:** 2025-05-23 06:15:05 |
| - **Iterations Performed:** 3 |
| - **Total Strategies Tested:** 32 |
| - **Best Overall Score:** 0.8722 |
|
|
| ## Performance Metrics |
|
|
| - **Mel Cepstral Distortion:** 0.9446 |
| - **Word Error Rate:** 0.0748 |
| - **Naturalness:** 0.9328 |
| - **Intelligibility:** 0.9724 |
| - **Speaker Similarity:** 0.9733 |
| - **Prosody:** 0.9526 |
| - **Overall Quality:** 0.9867 |
|
|
| ## Optimization Insights |
|
|
| ### Most Effective Parameter Settings |
|
|
| - **attention_scale:** 1.4798 |
| - **output_scale:** 1.4000 |
| - **projection_scale:** 1.5001 |
| - **encoder_scale:** 1.3002 |
| - **decoder_scale:** 1.3798 |
| - **base_enhancement:** 0.0020 |
| - **importance_factor:** 1.6199 |
| |
| ### Optimization Journey |
| |
| **Iteration 1:** Best strategy 'intelligibility_focus_var1.5' - Score 0.8722 |
| |
| **Iteration 2:** Best strategy 'intelligibility_focus_var1.5_iter2_noise_scale_0.8' - Score 0.8664 |
| |
| **Iteration 3:** Best strategy 'intelligibility_focus_var1.5_iter3_base_enhancement_0.8' - Score 0.8722 |
| |
| |
