zombie602 / optimization_report.md
memevis's picture
Upload folder using huggingface_hub
bf811a0 verified

TTS Model Optimization Report

Overview

  • Optimization Date: 2025-04-29 04:39:26
  • Iterations Performed: 3
  • Total Strategies Tested: 32
  • Best Overall Score: 0.8787

Performance Metrics

  • Mel Cepstral Distortion: 0.9405
  • Word Error Rate: 0.0685
  • Naturalness: 0.9465
  • Intelligibility: 0.9851
  • Speaker Similarity: 0.9711
  • Prosody: 0.9403
  • Overall Quality: 0.9837

Optimization Insights

Most Effective Parameter Settings

  • attention_scale: 1.3000
  • output_scale: 1.4000
  • projection_scale: 1.7294
  • encoder_scale: 1.6194
  • decoder_scale: 1.2728
  • base_enhancement: 0.0031
  • importance_factor: 1.5596

Optimization Journey

Iteration 1: Best strategy 'intelligibility_focus_var1.5' - Score 0.8714

Iteration 2: Best strategy 'intelligibility_focus_var1.5_iter2_noise_scale_0.8' - Score 0.8677

Iteration 3: Best strategy 'intelligibility_focus_var1.5_iter3_clarity_boost' - Score 0.8787