zombie5xx / optimization_report.md
memevis's picture
Upload folder using huggingface_hub
839f026 verified

TTS Model Optimization Report

Overview

  • Optimization Date: 2025-05-19 05:42:20
  • Iterations Performed: 3
  • Total Strategies Tested: 36
  • Best Overall Score: 0.8813

Performance Metrics

  • Mel Cepstral Distortion: 0.9250
  • Word Error Rate: 0.0512
  • Naturalness: 0.9429
  • Intelligibility: 0.9781
  • Speaker Similarity: 0.9580
  • Prosody: 0.9594
  • Overall Quality: 0.9037

Optimization Insights

Most Effective Parameter Settings

  • attention_scale: 1.4000
  • output_scale: 1.4000
  • projection_scale: 1.4000
  • encoder_scale: 1.4051
  • decoder_scale: 1.3000
  • base_enhancement: 0.0021
  • importance_factor: 1.7289

Optimization Journey

Iteration 1: Best strategy 'balanced_enhancement_var1.5' - Score 0.8699

Iteration 2: Best strategy 'balanced_enhancement_var1.5_iter2_importance_factor_1.2' - Score 0.8750

Iteration 3: Best strategy 'balanced_enhancement_var1.5_iter2_importance_factor_1.2_iter3_encoder_boost' - Score 0.8813