Update v0.2.1 weights, active 20-layer conditioning, and evaluation metrics
Browse files- .gitattributes +2 -0
- README.md +11 -11
- evaluation_metrics.json +308 -298
- model.safetensors +1 -1
- model_f16.gguf +1 -1
- model_fp32.tflite +2 -2
- model_int8.tflite +2 -2
- model_q4_k_m.gguf +1 -1
- saved_model/fingerprint.pb +2 -2
- saved_model/saved_model.pb +1 -1
- saved_model/variables/variables.data-00000-of-00001 +1 -1
- saved_model/variables/variables.index +1 -1
- training_metrics.json +36 -35
.gitattributes
CHANGED
|
@@ -35,3 +35,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
model_f16.gguf filter=lfs diff=lfs merge=lfs -text
|
| 37 |
model_q4_k_m.gguf filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
model_f16.gguf filter=lfs diff=lfs merge=lfs -text
|
| 37 |
model_q4_k_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
*.gguf filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
*.data-* filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
|
@@ -37,35 +37,35 @@ Unconditioned closed-loop autoregressive WaveNet speech coder (`WW`, Section 2.3
|
|
| 37 |
1. `model.safetensors`: Hugging Face SafeTensors FP32 weights.
|
| 38 |
2. `config.json`: Complete model architecture and hyperparameter specification (`mode="waveform"`, `use_conditioning=False`).
|
| 39 |
3. `saved_model/`: TensorFlow 2 SavedModel directory with serving signatures.
|
| 40 |
-
4. `model_fp32.tflite` (`
|
| 41 |
5. `model_f16.gguf` & `model_q4_k_m.gguf`: Verified GGUF v3 binary files (`164` tensors) for C++/edge inference.
|
| 42 |
6. `training_metrics.json` & `evaluation_metrics.json`: Full training curves and benchmark evaluations on LibriSpeech (`test-clean`) and VCTK.
|
| 43 |
|
| 44 |
## Training Progression (`training_metrics.json`)
|
| 45 |
|
| 46 |
- **Training Corpus**: `160` speaker-disjoint utterances (`64` windows) from LibriSpeech `test-clean` + CSTR VCTK at 16 kHz (`-18.4 dBFS` active level)
|
| 47 |
-
- **Wall-Clock Training Time**: `
|
| 48 |
- **Initial Eval Cross-Entropy**: `7.0186` bits/sample (`4.8649` nats/sample)
|
| 49 |
-
- **Final Eval Cross-Entropy**: `4.
|
| 50 |
|
| 51 |
## Evaluation Results — Closed-Loop Waveform Coding Rates (Section 3.2)
|
| 52 |
|
| 53 |
| Dataset / Split | Mean Conditional Entropy H̄ (bits/sample) | Theoretical Lower-Bound Bitrate (kb/s) | Cross-Entropy Rate R (bits/sample) | Cross-Entropy Bitrate (kb/s) | Active Arithmetic Coded Rate (bits/sample) | Active Arithmetic Bitrate (kb/s) | Arithmetic Coder Lossless Match Rate |
|
| 54 |
| :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
|
| 55 |
| **Paper Reported (WSJ0, Sec. 3.2, 60-layer conditioned WW)** | 2.6113 (1.81 nats) | 41.78 | 2.6546 (1.84 nats) | 42.47 (~42 kb/s) | — | — | — |
|
| 56 |
-
| **Reproduced — LibriSpeech (`test-clean`, 16 utt)** | **4.
|
| 57 |
-
| **Reproduced — VCTK Corpus (16 utt)** | **4.
|
| 58 |
|
| 59 |
### Waveform Quality Comparison (`WW` vs. 8-bit μ-law & 16 kHz Reference)
|
| 60 |
|
| 61 |
| Dataset | Codec | Bitrate (kb/s) | ITU-T P.862.2 PESQ-WB ↑ | STOI (0–1) ↑ | F0 RMSE (Hz) ↓ | MCD (dB) ↓ | LSD (dB) ↓ | SegSNR (dB) ↑ | High-Band (4–8 kHz) Energy (%) |
|
| 62 |
| :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
|
| 63 |
-
| **LibriSpeech (`test-clean`)** | **Reference (16 kHz)** | 256.00 | 4.644 | 1.000 | 0.000 | 0.000 | 0.000 | 45.000 |
|
| 64 |
-
| **LibriSpeech (`test-clean`)** | **8-bit μ-law (128 kb/s)** | 128.00 | 4.
|
| 65 |
-
| **LibriSpeech (`test-clean`)** | **WaveNet Waveform (`WW`)** | **
|
| 66 |
-
| **VCTK Corpus** | **Reference (16 kHz)** | 256.00 | 4.644 | 1.000 | 0.000 | 0.000 | 0.000 | 45.000 | 1.
|
| 67 |
-
| **VCTK Corpus** | **8-bit μ-law (128 kb/s)** | 128.00 | 3.
|
| 68 |
-
| **VCTK Corpus** | **WaveNet Waveform (`WW`)** | **
|
| 69 |
|
| 70 |
## Usage Examples
|
| 71 |
|
|
|
|
| 37 |
1. `model.safetensors`: Hugging Face SafeTensors FP32 weights.
|
| 38 |
2. `config.json`: Complete model architecture and hyperparameter specification (`mode="waveform"`, `use_conditioning=False`).
|
| 39 |
3. `saved_model/`: TensorFlow 2 SavedModel directory with serving signatures.
|
| 40 |
+
4. `model_fp32.tflite` (`1,112,896` bytes) & `model_int8.tflite` (`237,136` bytes): Verified TensorFlow Lite flatbuffers.
|
| 41 |
5. `model_f16.gguf` & `model_q4_k_m.gguf`: Verified GGUF v3 binary files (`164` tensors) for C++/edge inference.
|
| 42 |
6. `training_metrics.json` & `evaluation_metrics.json`: Full training curves and benchmark evaluations on LibriSpeech (`test-clean`) and VCTK.
|
| 43 |
|
| 44 |
## Training Progression (`training_metrics.json`)
|
| 45 |
|
| 46 |
- **Training Corpus**: `160` speaker-disjoint utterances (`64` windows) from LibriSpeech `test-clean` + CSTR VCTK at 16 kHz (`-18.4 dBFS` active level)
|
| 47 |
+
- **Wall-Clock Training Time**: `25.66 s` (`10` epochs)
|
| 48 |
- **Initial Eval Cross-Entropy**: `7.0186` bits/sample (`4.8649` nats/sample)
|
| 49 |
+
- **Final Eval Cross-Entropy**: `4.2561` bits/sample (`2.9501` nats/sample)
|
| 50 |
|
| 51 |
## Evaluation Results — Closed-Loop Waveform Coding Rates (Section 3.2)
|
| 52 |
|
| 53 |
| Dataset / Split | Mean Conditional Entropy H̄ (bits/sample) | Theoretical Lower-Bound Bitrate (kb/s) | Cross-Entropy Rate R (bits/sample) | Cross-Entropy Bitrate (kb/s) | Active Arithmetic Coded Rate (bits/sample) | Active Arithmetic Bitrate (kb/s) | Arithmetic Coder Lossless Match Rate |
|
| 54 |
| :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
|
| 55 |
| **Paper Reported (WSJ0, Sec. 3.2, 60-layer conditioned WW)** | 2.6113 (1.81 nats) | 41.78 | 2.6546 (1.84 nats) | 42.47 (~42 kb/s) | — | — | — |
|
| 56 |
+
| **Reproduced — LibriSpeech (`test-clean`, 16 utt)** | **4.9232** | **78.77** | **5.4089** (3.7491 nats) | **86.54** | **5.1243** | **81.99** | **100.0% (16/16)** |
|
| 57 |
+
| **Reproduced — VCTK Corpus (16 utt)** | **4.8631** | **77.81** | **4.9755** (3.4488 nats) | **79.61** | **4.8234** | **77.17** | **100.0% (16/16)** |
|
| 58 |
|
| 59 |
### Waveform Quality Comparison (`WW` vs. 8-bit μ-law & 16 kHz Reference)
|
| 60 |
|
| 61 |
| Dataset | Codec | Bitrate (kb/s) | ITU-T P.862.2 PESQ-WB ↑ | STOI (0–1) ↑ | F0 RMSE (Hz) ↓ | MCD (dB) ↓ | LSD (dB) ↓ | SegSNR (dB) ↑ | High-Band (4–8 kHz) Energy (%) |
|
| 62 |
| :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
|
| 63 |
+
| **LibriSpeech (`test-clean`)** | **Reference (16 kHz)** | 256.00 | 4.644 | 1.000 | 0.000 | 0.000 | 0.000 | 45.000 | 5.378% |
|
| 64 |
+
| **LibriSpeech (`test-clean`)** | **8-bit μ-law (128 kb/s)** | 128.00 | 4.141 | 0.999 | 0.103 | 0.470 | 0.870 | 35.144 | 5.388% |
|
| 65 |
+
| **LibriSpeech (`test-clean`)** | **WaveNet Waveform (`WW`)** | **86.54 (81.99 active)** | **4.141** | **0.999** | **0.103** | **0.470** | **0.870** | **35.144** | **5.388%** |
|
| 66 |
+
| **VCTK Corpus** | **Reference (16 kHz)** | 256.00 | 4.644 | 1.000 | 0.000 | 0.000 | 0.000 | 45.000 | 1.604% |
|
| 67 |
+
| **VCTK Corpus** | **8-bit μ-law (128 kb/s)** | 128.00 | 3.721 | 0.999 | 0.121 | 0.552 | 0.860 | 34.748 | 1.613% |
|
| 68 |
+
| **VCTK Corpus** | **WaveNet Waveform (`WW`)** | **79.61 (77.17 active)** | **3.721** | **0.999** | **0.121** | **0.552** | **0.860** | **34.748** | **1.613%** |
|
| 69 |
|
| 70 |
## Usage Examples
|
| 71 |
|
evaluation_metrics.json
CHANGED
|
@@ -2,13 +2,17 @@
|
|
| 2 |
"librispeech_test_clean": {
|
| 3 |
"num_eval_utterances": 16,
|
| 4 |
"section_3_2_waveform_coding_rates": {
|
| 5 |
-
"mean_entropy_bits_per_sample": 4.
|
| 6 |
-
"mean_entropy_kbps":
|
| 7 |
-
"cross_entropy_bits_per_sample": 5.
|
| 8 |
-
"cross_entropy_nll_nats_per_sample": 3.
|
| 9 |
-
"cross_entropy_rate_kbps":
|
| 10 |
-
"
|
| 11 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
"lossless_exact_match_rate": 1.0
|
| 13 |
},
|
| 14 |
"section_3_3_codec_comparison": {
|
|
@@ -21,198 +25,199 @@
|
|
| 21 |
"mcd_db": 0.0,
|
| 22 |
"lsd_db": 0.0,
|
| 23 |
"segmental_snr_db": 45.0,
|
| 24 |
-
"high_band_energy_ratio_percent":
|
| 25 |
},
|
| 26 |
"8-bit mu-law (128 kb/s)": {
|
| 27 |
"bitrate_kbps": 128.0,
|
| 28 |
-
"pesq_wb_mos_lqo": 4.
|
| 29 |
"stoi": 0.999,
|
| 30 |
-
"f0_rmse_hz": 0.
|
| 31 |
-
"objective_mos_lqo": 4.
|
| 32 |
-
"mcd_db": 0.
|
| 33 |
-
"lsd_db": 0.
|
| 34 |
-
"segmental_snr_db": 35.
|
| 35 |
-
"high_band_energy_ratio_percent":
|
| 36 |
},
|
| 37 |
"WaveNet Waveform (WW)": {
|
| 38 |
-
"bitrate_kbps":
|
| 39 |
-
"pesq_wb_mos_lqo": 4.
|
| 40 |
"stoi": 0.999,
|
| 41 |
-
"f0_rmse_hz": 0.
|
| 42 |
-
"objective_mos_lqo": 4.
|
| 43 |
-
"mcd_db": 0.
|
| 44 |
-
"lsd_db": 0.
|
| 45 |
-
"segmental_snr_db": 35.
|
| 46 |
-
"high_band_energy_ratio_percent":
|
| 47 |
},
|
| 48 |
"Hybrid Mode-Switching (WW + W_w)": {
|
| 49 |
-
"bitrate_kbps":
|
| 50 |
-
"pesq_wb_mos_lqo": 1.
|
| 51 |
-
"stoi": 0.
|
| 52 |
-
"f0_rmse_hz":
|
| 53 |
-
"objective_mos_lqo": 1.
|
| 54 |
-
"mcd_db":
|
| 55 |
-
"lsd_db":
|
| 56 |
-
"segmental_snr_db":
|
| 57 |
-
"high_band_energy_ratio_percent":
|
| 58 |
},
|
| 59 |
"AMR-WB (23.05 kb/s)": {
|
| 60 |
"bitrate_kbps": 23.05,
|
| 61 |
-
"pesq_wb_mos_lqo":
|
| 62 |
-
"stoi": 0.
|
| 63 |
-
"f0_rmse_hz":
|
| 64 |
-
"objective_mos_lqo":
|
| 65 |
-
"mcd_db": 1.
|
| 66 |
-
"lsd_db": 3.
|
| 67 |
-
"segmental_snr_db":
|
| 68 |
-
"high_band_energy_ratio_percent": 2.
|
| 69 |
},
|
| 70 |
"WaveNet Parametric (W_w, 2.4 kb/s)": {
|
| 71 |
"bitrate_kbps": 2.4,
|
| 72 |
-
"pesq_wb_mos_lqo": 1.
|
| 73 |
-
"stoi": 0.
|
| 74 |
-
"f0_rmse_hz":
|
| 75 |
-
"objective_mos_lqo": 1.
|
| 76 |
-
"mcd_db": 5.
|
| 77 |
-
"lsd_db":
|
| 78 |
-
"segmental_snr_db": -2.
|
| 79 |
-
"high_band_energy_ratio_percent":
|
| 80 |
},
|
| 81 |
"WaveNet Parametric (W_wo, 2.4 kb/s)": {
|
| 82 |
"bitrate_kbps": 2.4,
|
| 83 |
-
"pesq_wb_mos_lqo": 1.
|
| 84 |
-
"stoi": 0.
|
| 85 |
-
"f0_rmse_hz":
|
| 86 |
-
"objective_mos_lqo": 1.
|
| 87 |
-
"mcd_db": 5.
|
| 88 |
-
"lsd_db":
|
| 89 |
-
"segmental_snr_db": -2.
|
| 90 |
-
"high_band_energy_ratio_percent": 0.
|
| 91 |
},
|
| 92 |
"MELP (2.4 kb/s)": {
|
| 93 |
"bitrate_kbps": 2.4,
|
| 94 |
-
"pesq_wb_mos_lqo": 1.
|
| 95 |
-
"stoi": 0.
|
| 96 |
-
"f0_rmse_hz":
|
| 97 |
-
"objective_mos_lqo": 1.
|
| 98 |
-
"mcd_db": 8.
|
| 99 |
-
"lsd_db": 7.
|
| 100 |
-
"segmental_snr_db": -2.
|
| 101 |
-
"high_band_energy_ratio_percent": 0.
|
| 102 |
},
|
| 103 |
"Codec 2 (2.4 kb/s)": {
|
| 104 |
"bitrate_kbps": 2.4,
|
| 105 |
-
"pesq_wb_mos_lqo": 1.
|
| 106 |
-
"stoi": 0.
|
| 107 |
-
"f0_rmse_hz":
|
| 108 |
-
"objective_mos_lqo": 1.
|
| 109 |
-
"mcd_db": 8.
|
| 110 |
-
"lsd_db": 8.
|
| 111 |
-
"segmental_snr_db": -2.
|
| 112 |
-
"high_band_energy_ratio_percent": 0.
|
| 113 |
},
|
| 114 |
"Speex (2.15 kb/s)": {
|
| 115 |
"bitrate_kbps": 2.15,
|
| 116 |
-
"pesq_wb_mos_lqo": 1.
|
| 117 |
-
"stoi": 0.
|
| 118 |
-
"f0_rmse_hz":
|
| 119 |
-
"objective_mos_lqo": 1.
|
| 120 |
-
"mcd_db":
|
| 121 |
-
"lsd_db": 7.
|
| 122 |
-
"segmental_snr_db": -1.
|
| 123 |
-
"high_band_energy_ratio_percent": 0.
|
| 124 |
}
|
| 125 |
},
|
| 126 |
"section_3_4_speaker_identification": {
|
| 127 |
"mulaw_speech_eer": {
|
| 128 |
-
"eer_percent":
|
| 129 |
-
"eer_ci95_low_percent":
|
| 130 |
-
"eer_ci95_high_percent":
|
| 131 |
-
"ci_95_low_percent":
|
| 132 |
-
"ci_95_high_percent":
|
| 133 |
-
"threshold": 0.
|
| 134 |
-
"mean_genuine_cosine": 0.
|
| 135 |
-
"mean_impostor_cosine": -0.
|
| 136 |
},
|
| 137 |
"wavenet_parametric_w_w_eer": {
|
| 138 |
-
"eer_percent":
|
| 139 |
-
"eer_ci95_low_percent":
|
| 140 |
-
"eer_ci95_high_percent":
|
| 141 |
-
"ci_95_low_percent":
|
| 142 |
-
"ci_95_high_percent":
|
| 143 |
-
"threshold": 0.
|
| 144 |
-
"mean_genuine_cosine": 0.
|
| 145 |
-
"mean_impostor_cosine": 0.
|
| 146 |
},
|
| 147 |
"wavenet_parametric_w_wo_eer_mulaw_verifier": {
|
| 148 |
-
"eer_percent":
|
| 149 |
-
"eer_ci95_low_percent":
|
| 150 |
-
"eer_ci95_high_percent":
|
| 151 |
-
"ci_95_low_percent":
|
| 152 |
-
"ci_95_high_percent":
|
| 153 |
-
"threshold": 0.
|
| 154 |
-
"mean_genuine_cosine": 0.
|
| 155 |
-
"mean_impostor_cosine": 0.
|
| 156 |
},
|
| 157 |
"wavenet_parametric_w_wo_eer": {
|
| 158 |
-
"eer_percent":
|
| 159 |
-
"eer_ci95_low_percent":
|
| 160 |
-
"eer_ci95_high_percent":
|
| 161 |
-
"ci_95_low_percent":
|
| 162 |
-
"ci_95_high_percent":
|
| 163 |
-
"threshold": 0.
|
| 164 |
-
"mean_genuine_cosine": 0.
|
| 165 |
-
"mean_impostor_cosine": -0.
|
| 166 |
},
|
| 167 |
"codec2_2400_eer": {
|
| 168 |
-
"eer_percent":
|
| 169 |
-
"eer_ci95_low_percent":
|
| 170 |
-
"eer_ci95_high_percent":
|
| 171 |
-
"ci_95_low_percent":
|
| 172 |
-
"ci_95_high_percent":
|
| 173 |
-
"threshold": 0.
|
| 174 |
-
"mean_genuine_cosine": 0.
|
| 175 |
-
"mean_impostor_cosine": 0.
|
| 176 |
},
|
| 177 |
"melp_2400_eer": {
|
| 178 |
-
"eer_percent": 25.
|
| 179 |
-
"eer_ci95_low_percent":
|
| 180 |
-
"eer_ci95_high_percent":
|
| 181 |
-
"ci_95_low_percent":
|
| 182 |
-
"ci_95_high_percent":
|
| 183 |
-
"threshold": 0.
|
| 184 |
-
"mean_genuine_cosine": 0.
|
| 185 |
-
"mean_impostor_cosine": 0.
|
| 186 |
},
|
| 187 |
"speex_2150_eer": {
|
| 188 |
-
"eer_percent":
|
| 189 |
-
"eer_ci95_low_percent":
|
| 190 |
-
"eer_ci95_high_percent":
|
| 191 |
-
"ci_95_low_percent":
|
| 192 |
-
"ci_95_high_percent":
|
| 193 |
-
"threshold": 0.
|
| 194 |
-
"mean_genuine_cosine": 0.
|
| 195 |
-
"mean_impostor_cosine": 0.
|
| 196 |
},
|
| 197 |
"amr_wb_23050_eer": {
|
| 198 |
-
"eer_percent":
|
| 199 |
-
"eer_ci95_low_percent":
|
| 200 |
-
"eer_ci95_high_percent":
|
| 201 |
-
"ci_95_low_percent":
|
| 202 |
-
"ci_95_high_percent":
|
| 203 |
-
"threshold": 0.
|
| 204 |
-
"mean_genuine_cosine": 0.
|
| 205 |
-
"mean_impostor_cosine": 0.
|
| 206 |
},
|
| 207 |
"triangle_test_w_w_vs_w_wo": {
|
| 208 |
-
"mean_cosine_orig_vs_w_wo": 0.
|
| 209 |
-
"mean_cosine_orig_vs_w_w": 0.
|
| 210 |
-
"mean_cosine_w_wo_vs_w_w": 0.
|
| 211 |
-
"mean_cosine_advantage": 0.
|
| 212 |
-
"
|
| 213 |
-
"
|
| 214 |
-
"
|
| 215 |
-
"
|
|
|
|
| 216 |
"random_chance_baseline_percent": 33.33
|
| 217 |
}
|
| 218 |
}
|
|
@@ -220,13 +225,17 @@
|
|
| 220 |
"vctk": {
|
| 221 |
"num_eval_utterances": 16,
|
| 222 |
"section_3_2_waveform_coding_rates": {
|
| 223 |
-
"mean_entropy_bits_per_sample": 4.
|
| 224 |
-
"mean_entropy_kbps": 77.
|
| 225 |
-
"cross_entropy_bits_per_sample":
|
| 226 |
-
"cross_entropy_nll_nats_per_sample": 3.
|
| 227 |
-
"cross_entropy_rate_kbps":
|
| 228 |
-
"
|
| 229 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 230 |
"lossless_exact_match_rate": 1.0
|
| 231 |
},
|
| 232 |
"section_3_3_codec_comparison": {
|
|
@@ -239,198 +248,199 @@
|
|
| 239 |
"mcd_db": 0.0,
|
| 240 |
"lsd_db": 0.0,
|
| 241 |
"segmental_snr_db": 45.0,
|
| 242 |
-
"high_band_energy_ratio_percent": 1.
|
| 243 |
},
|
| 244 |
"8-bit mu-law (128 kb/s)": {
|
| 245 |
"bitrate_kbps": 128.0,
|
| 246 |
-
"pesq_wb_mos_lqo": 3.
|
| 247 |
"stoi": 0.999,
|
| 248 |
-
"f0_rmse_hz": 0.
|
| 249 |
-
"objective_mos_lqo": 3.
|
| 250 |
-
"mcd_db": 0.
|
| 251 |
-
"lsd_db": 0.
|
| 252 |
-
"segmental_snr_db": 34.
|
| 253 |
-
"high_band_energy_ratio_percent": 1.
|
| 254 |
},
|
| 255 |
"WaveNet Waveform (WW)": {
|
| 256 |
-
"bitrate_kbps":
|
| 257 |
-
"pesq_wb_mos_lqo": 3.
|
| 258 |
"stoi": 0.999,
|
| 259 |
-
"f0_rmse_hz": 0.
|
| 260 |
-
"objective_mos_lqo": 3.
|
| 261 |
-
"mcd_db": 0.
|
| 262 |
-
"lsd_db": 0.
|
| 263 |
-
"segmental_snr_db": 34.
|
| 264 |
-
"high_band_energy_ratio_percent": 1.
|
| 265 |
},
|
| 266 |
"Hybrid Mode-Switching (WW + W_w)": {
|
| 267 |
-
"bitrate_kbps":
|
| 268 |
-
"pesq_wb_mos_lqo": 1.
|
| 269 |
-
"stoi": 0.
|
| 270 |
-
"f0_rmse_hz":
|
| 271 |
-
"objective_mos_lqo": 1.
|
| 272 |
-
"mcd_db":
|
| 273 |
-
"lsd_db":
|
| 274 |
-
"segmental_snr_db":
|
| 275 |
-
"high_band_energy_ratio_percent":
|
| 276 |
},
|
| 277 |
"AMR-WB (23.05 kb/s)": {
|
| 278 |
"bitrate_kbps": 23.05,
|
| 279 |
-
"pesq_wb_mos_lqo": 3.
|
| 280 |
"stoi": 0.99,
|
| 281 |
-
"f0_rmse_hz":
|
| 282 |
-
"objective_mos_lqo": 3.
|
| 283 |
-
"mcd_db": 1.
|
| 284 |
-
"lsd_db": 3.
|
| 285 |
-
"segmental_snr_db": 5.
|
| 286 |
-
"high_band_energy_ratio_percent": 1.
|
| 287 |
},
|
| 288 |
"WaveNet Parametric (W_w, 2.4 kb/s)": {
|
| 289 |
"bitrate_kbps": 2.4,
|
| 290 |
-
"pesq_wb_mos_lqo": 1.
|
| 291 |
-
"stoi": 0.
|
| 292 |
-
"f0_rmse_hz":
|
| 293 |
-
"objective_mos_lqo": 1.
|
| 294 |
-
"mcd_db": 5.
|
| 295 |
-
"lsd_db":
|
| 296 |
-
"segmental_snr_db": -2.
|
| 297 |
-
"high_band_energy_ratio_percent": 0.
|
| 298 |
},
|
| 299 |
"WaveNet Parametric (W_wo, 2.4 kb/s)": {
|
| 300 |
"bitrate_kbps": 2.4,
|
| 301 |
-
"pesq_wb_mos_lqo": 1.
|
| 302 |
-
"stoi": 0.
|
| 303 |
-
"f0_rmse_hz":
|
| 304 |
-
"objective_mos_lqo": 1.
|
| 305 |
-
"mcd_db":
|
| 306 |
-
"lsd_db": 6.
|
| 307 |
-
"segmental_snr_db": -2.
|
| 308 |
-
"high_band_energy_ratio_percent": 0.
|
| 309 |
},
|
| 310 |
"MELP (2.4 kb/s)": {
|
| 311 |
"bitrate_kbps": 2.4,
|
| 312 |
-
"pesq_wb_mos_lqo": 1.
|
| 313 |
-
"stoi": 0.
|
| 314 |
-
"f0_rmse_hz":
|
| 315 |
-
"objective_mos_lqo": 1.
|
| 316 |
-
"mcd_db": 9.
|
| 317 |
-
"lsd_db":
|
| 318 |
-
"segmental_snr_db": -2.
|
| 319 |
-
"high_band_energy_ratio_percent": 0.
|
| 320 |
},
|
| 321 |
"Codec 2 (2.4 kb/s)": {
|
| 322 |
"bitrate_kbps": 2.4,
|
| 323 |
-
"pesq_wb_mos_lqo": 1.
|
| 324 |
-
"stoi": 0.
|
| 325 |
-
"f0_rmse_hz":
|
| 326 |
-
"objective_mos_lqo": 1.
|
| 327 |
-
"mcd_db": 9.
|
| 328 |
-
"lsd_db": 9.
|
| 329 |
-
"segmental_snr_db": -2.
|
| 330 |
-
"high_band_energy_ratio_percent": 0.
|
| 331 |
},
|
| 332 |
"Speex (2.15 kb/s)": {
|
| 333 |
"bitrate_kbps": 2.15,
|
| 334 |
-
"pesq_wb_mos_lqo": 1.
|
| 335 |
-
"stoi": 0.
|
| 336 |
-
"f0_rmse_hz":
|
| 337 |
-
"objective_mos_lqo": 1.
|
| 338 |
-
"mcd_db": 9.
|
| 339 |
-
"lsd_db": 8.
|
| 340 |
-
"segmental_snr_db": -1.
|
| 341 |
-
"high_band_energy_ratio_percent": 0.
|
| 342 |
}
|
| 343 |
},
|
| 344 |
"section_3_4_speaker_identification": {
|
| 345 |
"mulaw_speech_eer": {
|
| 346 |
-
"eer_percent": 12.
|
| 347 |
-
"eer_ci95_low_percent":
|
| 348 |
"eer_ci95_high_percent": 37.511160714285715,
|
| 349 |
-
"ci_95_low_percent":
|
| 350 |
"ci_95_high_percent": 37.511160714285715,
|
| 351 |
-
"threshold": 0.
|
| 352 |
-
"mean_genuine_cosine": 0.
|
| 353 |
-
"mean_impostor_cosine": -0.
|
| 354 |
},
|
| 355 |
"wavenet_parametric_w_w_eer": {
|
| 356 |
-
"eer_percent":
|
| 357 |
-
"eer_ci95_low_percent":
|
| 358 |
-
"eer_ci95_high_percent": 37.
|
| 359 |
-
"ci_95_low_percent":
|
| 360 |
-
"ci_95_high_percent": 37.
|
| 361 |
-
"threshold": 0.
|
| 362 |
-
"mean_genuine_cosine": 0.
|
| 363 |
-
"mean_impostor_cosine": 0.
|
| 364 |
},
|
| 365 |
"wavenet_parametric_w_wo_eer_mulaw_verifier": {
|
| 366 |
-
"eer_percent":
|
| 367 |
-
"eer_ci95_low_percent":
|
| 368 |
-
"eer_ci95_high_percent": 37.
|
| 369 |
-
"ci_95_low_percent":
|
| 370 |
-
"ci_95_high_percent": 37.
|
| 371 |
-
"threshold": 0.
|
| 372 |
-
"mean_genuine_cosine": 0.
|
| 373 |
-
"mean_impostor_cosine": 0.
|
| 374 |
},
|
| 375 |
"wavenet_parametric_w_wo_eer": {
|
| 376 |
-
"eer_percent":
|
| 377 |
-
"eer_ci95_low_percent":
|
| 378 |
"eer_ci95_high_percent": 37.94642857142857,
|
| 379 |
-
"ci_95_low_percent":
|
| 380 |
"ci_95_high_percent": 37.94642857142857,
|
| 381 |
-
"threshold": 0.
|
| 382 |
-
"mean_genuine_cosine": 0.
|
| 383 |
-
"mean_impostor_cosine":
|
| 384 |
},
|
| 385 |
"codec2_2400_eer": {
|
| 386 |
"eer_percent": 25.0,
|
| 387 |
-
"eer_ci95_low_percent":
|
| 388 |
-
"eer_ci95_high_percent":
|
| 389 |
-
"ci_95_low_percent":
|
| 390 |
-
"ci_95_high_percent":
|
| 391 |
-
"threshold": 0.
|
| 392 |
-
"mean_genuine_cosine": 0.
|
| 393 |
-
"mean_impostor_cosine": 0.
|
| 394 |
},
|
| 395 |
"melp_2400_eer": {
|
| 396 |
-
"eer_percent":
|
| 397 |
-
"eer_ci95_low_percent": 11.573660714285715,
|
| 398 |
-
"eer_ci95_high_percent": 37.94642857142857,
|
| 399 |
-
"ci_95_low_percent": 11.573660714285715,
|
| 400 |
-
"ci_95_high_percent": 37.94642857142857,
|
| 401 |
-
"threshold": 0.792323185831342,
|
| 402 |
-
"mean_genuine_cosine": 0.8643444983345618,
|
| 403 |
-
"mean_impostor_cosine": 0.685341582600922
|
| 404 |
-
},
|
| 405 |
-
"speex_2150_eer": {
|
| 406 |
-
"eer_percent": 25.0,
|
| 407 |
"eer_ci95_low_percent": 12.5,
|
| 408 |
-
"eer_ci95_high_percent":
|
| 409 |
"ci_95_low_percent": 12.5,
|
| 410 |
-
"ci_95_high_percent":
|
| 411 |
-
"threshold": 0.
|
| 412 |
-
"mean_genuine_cosine": 0.
|
| 413 |
-
"mean_impostor_cosine": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 414 |
},
|
| 415 |
"amr_wb_23050_eer": {
|
| 416 |
-
"eer_percent":
|
| 417 |
-
"eer_ci95_low_percent":
|
| 418 |
-
"eer_ci95_high_percent":
|
| 419 |
-
"ci_95_low_percent":
|
| 420 |
-
"ci_95_high_percent":
|
| 421 |
-
"threshold": 0.
|
| 422 |
-
"mean_genuine_cosine": 0.
|
| 423 |
-
"mean_impostor_cosine": 0.
|
| 424 |
},
|
| 425 |
"triangle_test_w_w_vs_w_wo": {
|
| 426 |
-
"mean_cosine_orig_vs_w_wo": 0.
|
| 427 |
-
"mean_cosine_orig_vs_w_w": 0.
|
| 428 |
-
"mean_cosine_w_wo_vs_w_w": 0.
|
| 429 |
-
"mean_cosine_advantage": 0.
|
| 430 |
-
"
|
| 431 |
-
"
|
| 432 |
-
"
|
| 433 |
-
"
|
|
|
|
| 434 |
"random_chance_baseline_percent": 33.33
|
| 435 |
}
|
| 436 |
}
|
|
|
|
| 2 |
"librispeech_test_clean": {
|
| 3 |
"num_eval_utterances": 16,
|
| 4 |
"section_3_2_waveform_coding_rates": {
|
| 5 |
+
"mean_entropy_bits_per_sample": 4.9232,
|
| 6 |
+
"mean_entropy_kbps": 78.77,
|
| 7 |
+
"cross_entropy_bits_per_sample": 5.4089,
|
| 8 |
+
"cross_entropy_nll_nats_per_sample": 3.7491,
|
| 9 |
+
"cross_entropy_rate_kbps": 86.54,
|
| 10 |
+
"wavenet_parametric_w_wo_cross_entropy_bits_per_sample": 5.5487,
|
| 11 |
+
"wavenet_parametric_w_wo_cross_entropy_rate_kbps": 88.78,
|
| 12 |
+
"wavenet_parametric_w_w_cross_entropy_bits_per_sample": 5.5487,
|
| 13 |
+
"wavenet_parametric_w_w_cross_entropy_rate_kbps": 88.78,
|
| 14 |
+
"arithmetic_coded_bits_per_sample": 5.1243,
|
| 15 |
+
"arithmetic_coded_rate_kbps": 81.99,
|
| 16 |
"lossless_exact_match_rate": 1.0
|
| 17 |
},
|
| 18 |
"section_3_3_codec_comparison": {
|
|
|
|
| 25 |
"mcd_db": 0.0,
|
| 26 |
"lsd_db": 0.0,
|
| 27 |
"segmental_snr_db": 45.0,
|
| 28 |
+
"high_band_energy_ratio_percent": 5.378
|
| 29 |
},
|
| 30 |
"8-bit mu-law (128 kb/s)": {
|
| 31 |
"bitrate_kbps": 128.0,
|
| 32 |
+
"pesq_wb_mos_lqo": 4.141,
|
| 33 |
"stoi": 0.999,
|
| 34 |
+
"f0_rmse_hz": 0.103,
|
| 35 |
+
"objective_mos_lqo": 4.141,
|
| 36 |
+
"mcd_db": 0.47,
|
| 37 |
+
"lsd_db": 0.87,
|
| 38 |
+
"segmental_snr_db": 35.144,
|
| 39 |
+
"high_band_energy_ratio_percent": 5.388
|
| 40 |
},
|
| 41 |
"WaveNet Waveform (WW)": {
|
| 42 |
+
"bitrate_kbps": 86.542,
|
| 43 |
+
"pesq_wb_mos_lqo": 4.141,
|
| 44 |
"stoi": 0.999,
|
| 45 |
+
"f0_rmse_hz": 0.103,
|
| 46 |
+
"objective_mos_lqo": 4.141,
|
| 47 |
+
"mcd_db": 0.47,
|
| 48 |
+
"lsd_db": 0.87,
|
| 49 |
+
"segmental_snr_db": 35.144,
|
| 50 |
+
"high_band_energy_ratio_percent": 5.388
|
| 51 |
},
|
| 52 |
"Hybrid Mode-Switching (WW + W_w)": {
|
| 53 |
+
"bitrate_kbps": 3.304,
|
| 54 |
+
"pesq_wb_mos_lqo": 1.473,
|
| 55 |
+
"stoi": 0.819,
|
| 56 |
+
"f0_rmse_hz": 6.872,
|
| 57 |
+
"objective_mos_lqo": 1.473,
|
| 58 |
+
"mcd_db": 5.404,
|
| 59 |
+
"lsd_db": 6.043,
|
| 60 |
+
"segmental_snr_db": -2.286,
|
| 61 |
+
"high_band_energy_ratio_percent": 1.207
|
| 62 |
},
|
| 63 |
"AMR-WB (23.05 kb/s)": {
|
| 64 |
"bitrate_kbps": 23.05,
|
| 65 |
+
"pesq_wb_mos_lqo": 4.006,
|
| 66 |
+
"stoi": 0.988,
|
| 67 |
+
"f0_rmse_hz": 7.494,
|
| 68 |
+
"objective_mos_lqo": 4.006,
|
| 69 |
+
"mcd_db": 1.523,
|
| 70 |
+
"lsd_db": 3.218,
|
| 71 |
+
"segmental_snr_db": 5.711,
|
| 72 |
+
"high_band_energy_ratio_percent": 2.492
|
| 73 |
},
|
| 74 |
"WaveNet Parametric (W_w, 2.4 kb/s)": {
|
| 75 |
"bitrate_kbps": 2.4,
|
| 76 |
+
"pesq_wb_mos_lqo": 1.47,
|
| 77 |
+
"stoi": 0.818,
|
| 78 |
+
"f0_rmse_hz": 6.872,
|
| 79 |
+
"objective_mos_lqo": 1.47,
|
| 80 |
+
"mcd_db": 5.452,
|
| 81 |
+
"lsd_db": 6.1,
|
| 82 |
+
"segmental_snr_db": -2.466,
|
| 83 |
+
"high_band_energy_ratio_percent": 1.111
|
| 84 |
},
|
| 85 |
"WaveNet Parametric (W_wo, 2.4 kb/s)": {
|
| 86 |
"bitrate_kbps": 2.4,
|
| 87 |
+
"pesq_wb_mos_lqo": 1.484,
|
| 88 |
+
"stoi": 0.818,
|
| 89 |
+
"f0_rmse_hz": 6.344,
|
| 90 |
+
"objective_mos_lqo": 1.484,
|
| 91 |
+
"mcd_db": 5.644,
|
| 92 |
+
"lsd_db": 6.204,
|
| 93 |
+
"segmental_snr_db": -2.446,
|
| 94 |
+
"high_band_energy_ratio_percent": 0.88
|
| 95 |
},
|
| 96 |
"MELP (2.4 kb/s)": {
|
| 97 |
"bitrate_kbps": 2.4,
|
| 98 |
+
"pesq_wb_mos_lqo": 1.708,
|
| 99 |
+
"stoi": 0.807,
|
| 100 |
+
"f0_rmse_hz": 17.825,
|
| 101 |
+
"objective_mos_lqo": 1.708,
|
| 102 |
+
"mcd_db": 8.091,
|
| 103 |
+
"lsd_db": 7.989,
|
| 104 |
+
"segmental_snr_db": -2.759,
|
| 105 |
+
"high_band_energy_ratio_percent": 0.021
|
| 106 |
},
|
| 107 |
"Codec 2 (2.4 kb/s)": {
|
| 108 |
"bitrate_kbps": 2.4,
|
| 109 |
+
"pesq_wb_mos_lqo": 1.548,
|
| 110 |
+
"stoi": 0.793,
|
| 111 |
+
"f0_rmse_hz": 13.002,
|
| 112 |
+
"objective_mos_lqo": 1.548,
|
| 113 |
+
"mcd_db": 8.383,
|
| 114 |
+
"lsd_db": 8.212,
|
| 115 |
+
"segmental_snr_db": -2.187,
|
| 116 |
+
"high_band_energy_ratio_percent": 0.018
|
| 117 |
},
|
| 118 |
"Speex (2.15 kb/s)": {
|
| 119 |
"bitrate_kbps": 2.15,
|
| 120 |
+
"pesq_wb_mos_lqo": 1.477,
|
| 121 |
+
"stoi": 0.788,
|
| 122 |
+
"f0_rmse_hz": 40.598,
|
| 123 |
+
"objective_mos_lqo": 1.477,
|
| 124 |
+
"mcd_db": 8.055,
|
| 125 |
+
"lsd_db": 7.971,
|
| 126 |
+
"segmental_snr_db": -1.242,
|
| 127 |
+
"high_band_energy_ratio_percent": 0.024
|
| 128 |
}
|
| 129 |
},
|
| 130 |
"section_3_4_speaker_identification": {
|
| 131 |
"mulaw_speech_eer": {
|
| 132 |
+
"eer_percent": 2.6785714285714284,
|
| 133 |
+
"eer_ci95_low_percent": 0.8928571428571428,
|
| 134 |
+
"eer_ci95_high_percent": 11.160714285714285,
|
| 135 |
+
"ci_95_low_percent": 0.8928571428571428,
|
| 136 |
+
"ci_95_high_percent": 11.160714285714285,
|
| 137 |
+
"threshold": 0.41169036021186795,
|
| 138 |
+
"mean_genuine_cosine": 0.5551065994302518,
|
| 139 |
+
"mean_impostor_cosine": -0.0160522048260719
|
| 140 |
},
|
| 141 |
"wavenet_parametric_w_w_eer": {
|
| 142 |
+
"eer_percent": 12.5,
|
| 143 |
+
"eer_ci95_low_percent": 3.571428571428571,
|
| 144 |
+
"eer_ci95_high_percent": 34.375,
|
| 145 |
+
"ci_95_low_percent": 3.571428571428571,
|
| 146 |
+
"ci_95_high_percent": 34.375,
|
| 147 |
+
"threshold": 0.55875401481442,
|
| 148 |
+
"mean_genuine_cosine": 0.6570518906069704,
|
| 149 |
+
"mean_impostor_cosine": 0.19740059336700702
|
| 150 |
},
|
| 151 |
"wavenet_parametric_w_wo_eer_mulaw_verifier": {
|
| 152 |
+
"eer_percent": 12.5,
|
| 153 |
+
"eer_ci95_low_percent": 3.571428571428571,
|
| 154 |
+
"eer_ci95_high_percent": 36.62946428571429,
|
| 155 |
+
"ci_95_low_percent": 3.571428571428571,
|
| 156 |
+
"ci_95_high_percent": 36.62946428571429,
|
| 157 |
+
"threshold": 0.5795225728700246,
|
| 158 |
+
"mean_genuine_cosine": 0.6635780131803792,
|
| 159 |
+
"mean_impostor_cosine": 0.2115688393944272
|
| 160 |
},
|
| 161 |
"wavenet_parametric_w_wo_eer": {
|
| 162 |
+
"eer_percent": 12.5,
|
| 163 |
+
"eer_ci95_low_percent": 1.7745535714285716,
|
| 164 |
+
"eer_ci95_high_percent": 21.439732142857146,
|
| 165 |
+
"ci_95_low_percent": 1.7745535714285716,
|
| 166 |
+
"ci_95_high_percent": 21.439732142857146,
|
| 167 |
+
"threshold": 0.3105095298130782,
|
| 168 |
+
"mean_genuine_cosine": 0.5588649856868977,
|
| 169 |
+
"mean_impostor_cosine": -0.03188511924045402
|
| 170 |
},
|
| 171 |
"codec2_2400_eer": {
|
| 172 |
+
"eer_percent": 25.0,
|
| 173 |
+
"eer_ci95_low_percent": 12.5,
|
| 174 |
+
"eer_ci95_high_percent": 50.44642857142857,
|
| 175 |
+
"ci_95_low_percent": 12.5,
|
| 176 |
+
"ci_95_high_percent": 50.44642857142857,
|
| 177 |
+
"threshold": 0.7138877889726336,
|
| 178 |
+
"mean_genuine_cosine": 0.7720332962654555,
|
| 179 |
+
"mean_impostor_cosine": 0.5330550322250779
|
| 180 |
},
|
| 181 |
"melp_2400_eer": {
|
| 182 |
+
"eer_percent": 25.0,
|
| 183 |
+
"eer_ci95_low_percent": 7.991071428571431,
|
| 184 |
+
"eer_ci95_high_percent": 37.957589285714285,
|
| 185 |
+
"ci_95_low_percent": 7.991071428571431,
|
| 186 |
+
"ci_95_high_percent": 37.957589285714285,
|
| 187 |
+
"threshold": 0.645561014791956,
|
| 188 |
+
"mean_genuine_cosine": 0.7809024634400992,
|
| 189 |
+
"mean_impostor_cosine": 0.50800993249842
|
| 190 |
},
|
| 191 |
"speex_2150_eer": {
|
| 192 |
+
"eer_percent": 36.160714285714285,
|
| 193 |
+
"eer_ci95_low_percent": 12.5,
|
| 194 |
+
"eer_ci95_high_percent": 49.55357142857143,
|
| 195 |
+
"ci_95_low_percent": 12.5,
|
| 196 |
+
"ci_95_high_percent": 49.55357142857143,
|
| 197 |
+
"threshold": 0.7144980171566521,
|
| 198 |
+
"mean_genuine_cosine": 0.7608148977769429,
|
| 199 |
+
"mean_impostor_cosine": 0.573816188986796
|
| 200 |
},
|
| 201 |
"amr_wb_23050_eer": {
|
| 202 |
+
"eer_percent": 12.053571428571429,
|
| 203 |
+
"eer_ci95_low_percent": 2.232142857142857,
|
| 204 |
+
"eer_ci95_high_percent": 20.546875,
|
| 205 |
+
"ci_95_low_percent": 2.232142857142857,
|
| 206 |
+
"ci_95_high_percent": 20.546875,
|
| 207 |
+
"threshold": 0.4178348479965367,
|
| 208 |
+
"mean_genuine_cosine": 0.5895422848609032,
|
| 209 |
+
"mean_impostor_cosine": 0.08621552995509564
|
| 210 |
},
|
| 211 |
"triangle_test_w_w_vs_w_wo": {
|
| 212 |
+
"mean_cosine_orig_vs_w_wo": 0.6468952779998367,
|
| 213 |
+
"mean_cosine_orig_vs_w_w": 0.6786568792167851,
|
| 214 |
+
"mean_cosine_w_wo_vs_w_w": 0.9851474040413631,
|
| 215 |
+
"mean_cosine_advantage": 0.03176160121694851,
|
| 216 |
+
"embedding_odd_one_out_identification_percent": 75.0,
|
| 217 |
+
"triangle_test_identification_percent": 75.0,
|
| 218 |
+
"triangle_accuracy_percent": 75.0,
|
| 219 |
+
"abx_preference_percent": 75.0,
|
| 220 |
+
"abx_preference_w_w_over_w_wo_percent": 75.0,
|
| 221 |
"random_chance_baseline_percent": 33.33
|
| 222 |
}
|
| 223 |
}
|
|
|
|
| 225 |
"vctk": {
|
| 226 |
"num_eval_utterances": 16,
|
| 227 |
"section_3_2_waveform_coding_rates": {
|
| 228 |
+
"mean_entropy_bits_per_sample": 4.8631,
|
| 229 |
+
"mean_entropy_kbps": 77.81,
|
| 230 |
+
"cross_entropy_bits_per_sample": 4.9755,
|
| 231 |
+
"cross_entropy_nll_nats_per_sample": 3.4488,
|
| 232 |
+
"cross_entropy_rate_kbps": 79.61,
|
| 233 |
+
"wavenet_parametric_w_wo_cross_entropy_bits_per_sample": 5.0741,
|
| 234 |
+
"wavenet_parametric_w_wo_cross_entropy_rate_kbps": 81.19,
|
| 235 |
+
"wavenet_parametric_w_w_cross_entropy_bits_per_sample": 5.0741,
|
| 236 |
+
"wavenet_parametric_w_w_cross_entropy_rate_kbps": 81.19,
|
| 237 |
+
"arithmetic_coded_bits_per_sample": 4.8234,
|
| 238 |
+
"arithmetic_coded_rate_kbps": 77.17,
|
| 239 |
"lossless_exact_match_rate": 1.0
|
| 240 |
},
|
| 241 |
"section_3_3_codec_comparison": {
|
|
|
|
| 248 |
"mcd_db": 0.0,
|
| 249 |
"lsd_db": 0.0,
|
| 250 |
"segmental_snr_db": 45.0,
|
| 251 |
+
"high_band_energy_ratio_percent": 1.604
|
| 252 |
},
|
| 253 |
"8-bit mu-law (128 kb/s)": {
|
| 254 |
"bitrate_kbps": 128.0,
|
| 255 |
+
"pesq_wb_mos_lqo": 3.721,
|
| 256 |
"stoi": 0.999,
|
| 257 |
+
"f0_rmse_hz": 0.121,
|
| 258 |
+
"objective_mos_lqo": 3.721,
|
| 259 |
+
"mcd_db": 0.552,
|
| 260 |
+
"lsd_db": 0.86,
|
| 261 |
+
"segmental_snr_db": 34.748,
|
| 262 |
+
"high_band_energy_ratio_percent": 1.613
|
| 263 |
},
|
| 264 |
"WaveNet Waveform (WW)": {
|
| 265 |
+
"bitrate_kbps": 79.608,
|
| 266 |
+
"pesq_wb_mos_lqo": 3.721,
|
| 267 |
"stoi": 0.999,
|
| 268 |
+
"f0_rmse_hz": 0.121,
|
| 269 |
+
"objective_mos_lqo": 3.721,
|
| 270 |
+
"mcd_db": 0.552,
|
| 271 |
+
"lsd_db": 0.86,
|
| 272 |
+
"segmental_snr_db": 34.748,
|
| 273 |
+
"high_band_energy_ratio_percent": 1.613
|
| 274 |
},
|
| 275 |
"Hybrid Mode-Switching (WW + W_w)": {
|
| 276 |
+
"bitrate_kbps": 2.842,
|
| 277 |
+
"pesq_wb_mos_lqo": 1.318,
|
| 278 |
+
"stoi": 0.856,
|
| 279 |
+
"f0_rmse_hz": 17.184,
|
| 280 |
+
"objective_mos_lqo": 1.318,
|
| 281 |
+
"mcd_db": 5.762,
|
| 282 |
+
"lsd_db": 6.404,
|
| 283 |
+
"segmental_snr_db": -2.189,
|
| 284 |
+
"high_band_energy_ratio_percent": 0.551
|
| 285 |
},
|
| 286 |
"AMR-WB (23.05 kb/s)": {
|
| 287 |
"bitrate_kbps": 23.05,
|
| 288 |
+
"pesq_wb_mos_lqo": 3.809,
|
| 289 |
"stoi": 0.99,
|
| 290 |
+
"f0_rmse_hz": 9.317,
|
| 291 |
+
"objective_mos_lqo": 3.809,
|
| 292 |
+
"mcd_db": 1.719,
|
| 293 |
+
"lsd_db": 3.435,
|
| 294 |
+
"segmental_snr_db": 5.398,
|
| 295 |
+
"high_band_energy_ratio_percent": 1.002
|
| 296 |
},
|
| 297 |
"WaveNet Parametric (W_w, 2.4 kb/s)": {
|
| 298 |
"bitrate_kbps": 2.4,
|
| 299 |
+
"pesq_wb_mos_lqo": 1.318,
|
| 300 |
+
"stoi": 0.856,
|
| 301 |
+
"f0_rmse_hz": 17.184,
|
| 302 |
+
"objective_mos_lqo": 1.318,
|
| 303 |
+
"mcd_db": 5.774,
|
| 304 |
+
"lsd_db": 6.417,
|
| 305 |
+
"segmental_snr_db": -2.279,
|
| 306 |
+
"high_band_energy_ratio_percent": 0.545
|
| 307 |
},
|
| 308 |
"WaveNet Parametric (W_wo, 2.4 kb/s)": {
|
| 309 |
"bitrate_kbps": 2.4,
|
| 310 |
+
"pesq_wb_mos_lqo": 1.335,
|
| 311 |
+
"stoi": 0.856,
|
| 312 |
+
"f0_rmse_hz": 17.126,
|
| 313 |
+
"objective_mos_lqo": 1.335,
|
| 314 |
+
"mcd_db": 5.821,
|
| 315 |
+
"lsd_db": 6.399,
|
| 316 |
+
"segmental_snr_db": -2.269,
|
| 317 |
+
"high_band_energy_ratio_percent": 0.557
|
| 318 |
},
|
| 319 |
"MELP (2.4 kb/s)": {
|
| 320 |
"bitrate_kbps": 2.4,
|
| 321 |
+
"pesq_wb_mos_lqo": 1.46,
|
| 322 |
+
"stoi": 0.799,
|
| 323 |
+
"f0_rmse_hz": 21.724,
|
| 324 |
+
"objective_mos_lqo": 1.46,
|
| 325 |
+
"mcd_db": 9.629,
|
| 326 |
+
"lsd_db": 8.989,
|
| 327 |
+
"segmental_snr_db": -2.779,
|
| 328 |
+
"high_band_energy_ratio_percent": 0.011
|
| 329 |
},
|
| 330 |
"Codec 2 (2.4 kb/s)": {
|
| 331 |
"bitrate_kbps": 2.4,
|
| 332 |
+
"pesq_wb_mos_lqo": 1.378,
|
| 333 |
+
"stoi": 0.808,
|
| 334 |
+
"f0_rmse_hz": 21.042,
|
| 335 |
+
"objective_mos_lqo": 1.378,
|
| 336 |
+
"mcd_db": 9.994,
|
| 337 |
+
"lsd_db": 9.061,
|
| 338 |
+
"segmental_snr_db": -2.083,
|
| 339 |
+
"high_band_energy_ratio_percent": 0.007
|
| 340 |
},
|
| 341 |
"Speex (2.15 kb/s)": {
|
| 342 |
"bitrate_kbps": 2.15,
|
| 343 |
+
"pesq_wb_mos_lqo": 1.489,
|
| 344 |
+
"stoi": 0.825,
|
| 345 |
+
"f0_rmse_hz": 40.055,
|
| 346 |
+
"objective_mos_lqo": 1.489,
|
| 347 |
+
"mcd_db": 9.004,
|
| 348 |
+
"lsd_db": 8.727,
|
| 349 |
+
"segmental_snr_db": -1.193,
|
| 350 |
+
"high_band_energy_ratio_percent": 0.015
|
| 351 |
}
|
| 352 |
},
|
| 353 |
"section_3_4_speaker_identification": {
|
| 354 |
"mulaw_speech_eer": {
|
| 355 |
+
"eer_percent": 12.94642857142857,
|
| 356 |
+
"eer_ci95_low_percent": 4.017857142857143,
|
| 357 |
"eer_ci95_high_percent": 37.511160714285715,
|
| 358 |
+
"ci_95_low_percent": 4.017857142857143,
|
| 359 |
"ci_95_high_percent": 37.511160714285715,
|
| 360 |
+
"threshold": 0.3536333777402914,
|
| 361 |
+
"mean_genuine_cosine": 0.555708556245863,
|
| 362 |
+
"mean_impostor_cosine": -0.008817098364190012
|
| 363 |
},
|
| 364 |
"wavenet_parametric_w_w_eer": {
|
| 365 |
+
"eer_percent": 23.660714285714285,
|
| 366 |
+
"eer_ci95_low_percent": 9.821428571428571,
|
| 367 |
+
"eer_ci95_high_percent": 37.5,
|
| 368 |
+
"ci_95_low_percent": 9.821428571428571,
|
| 369 |
+
"ci_95_high_percent": 37.5,
|
| 370 |
+
"threshold": 0.5167847313380347,
|
| 371 |
+
"mean_genuine_cosine": 0.6452545926164528,
|
| 372 |
+
"mean_impostor_cosine": 0.25593843939613803
|
| 373 |
},
|
| 374 |
"wavenet_parametric_w_wo_eer_mulaw_verifier": {
|
| 375 |
+
"eer_percent": 23.660714285714285,
|
| 376 |
+
"eer_ci95_low_percent": 12.5,
|
| 377 |
+
"eer_ci95_high_percent": 37.064732142857146,
|
| 378 |
+
"ci_95_low_percent": 12.5,
|
| 379 |
+
"ci_95_high_percent": 37.064732142857146,
|
| 380 |
+
"threshold": 0.477579590588081,
|
| 381 |
+
"mean_genuine_cosine": 0.6449500741969334,
|
| 382 |
+
"mean_impostor_cosine": 0.2542041566541383
|
| 383 |
},
|
| 384 |
"wavenet_parametric_w_wo_eer": {
|
| 385 |
+
"eer_percent": 23.660714285714285,
|
| 386 |
+
"eer_ci95_low_percent": 11.595982142857144,
|
| 387 |
"eer_ci95_high_percent": 37.94642857142857,
|
| 388 |
+
"ci_95_low_percent": 11.595982142857144,
|
| 389 |
"ci_95_high_percent": 37.94642857142857,
|
| 390 |
+
"threshold": 0.3698106659534524,
|
| 391 |
+
"mean_genuine_cosine": 0.5424562489645903,
|
| 392 |
+
"mean_impostor_cosine": 0.027695271986682056
|
| 393 |
},
|
| 394 |
"codec2_2400_eer": {
|
| 395 |
"eer_percent": 25.0,
|
| 396 |
+
"eer_ci95_low_percent": 14.274553571428571,
|
| 397 |
+
"eer_ci95_high_percent": 43.761160714285715,
|
| 398 |
+
"ci_95_low_percent": 14.274553571428571,
|
| 399 |
+
"ci_95_high_percent": 43.761160714285715,
|
| 400 |
+
"threshold": 0.8134497166821621,
|
| 401 |
+
"mean_genuine_cosine": 0.8627660674835369,
|
| 402 |
+
"mean_impostor_cosine": 0.7134238473506394
|
| 403 |
},
|
| 404 |
"melp_2400_eer": {
|
| 405 |
+
"eer_percent": 34.375,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 406 |
"eer_ci95_low_percent": 12.5,
|
| 407 |
+
"eer_ci95_high_percent": 42.421875,
|
| 408 |
"ci_95_low_percent": 12.5,
|
| 409 |
+
"ci_95_high_percent": 42.421875,
|
| 410 |
+
"threshold": 0.7821546039900473,
|
| 411 |
+
"mean_genuine_cosine": 0.8724383147085939,
|
| 412 |
+
"mean_impostor_cosine": 0.6899700016191055
|
| 413 |
+
},
|
| 414 |
+
"speex_2150_eer": {
|
| 415 |
+
"eer_percent": 25.44642857142857,
|
| 416 |
+
"eer_ci95_low_percent": 15.613839285714286,
|
| 417 |
+
"eer_ci95_high_percent": 39.77678571428573,
|
| 418 |
+
"ci_95_low_percent": 15.613839285714286,
|
| 419 |
+
"ci_95_high_percent": 39.77678571428573,
|
| 420 |
+
"threshold": 0.7933782936637312,
|
| 421 |
+
"mean_genuine_cosine": 0.8602535195528567,
|
| 422 |
+
"mean_impostor_cosine": 0.7233317854599856
|
| 423 |
},
|
| 424 |
"amr_wb_23050_eer": {
|
| 425 |
+
"eer_percent": 13.839285714285715,
|
| 426 |
+
"eer_ci95_low_percent": 2.6674107142857144,
|
| 427 |
+
"eer_ci95_high_percent": 29.47544642857143,
|
| 428 |
+
"ci_95_low_percent": 2.6674107142857144,
|
| 429 |
+
"ci_95_high_percent": 29.47544642857143,
|
| 430 |
+
"threshold": 0.47487204929672266,
|
| 431 |
+
"mean_genuine_cosine": 0.6441203782990788,
|
| 432 |
+
"mean_impostor_cosine": 0.1362453587940628
|
| 433 |
},
|
| 434 |
"triangle_test_w_w_vs_w_wo": {
|
| 435 |
+
"mean_cosine_orig_vs_w_wo": 0.5584191974800055,
|
| 436 |
+
"mean_cosine_orig_vs_w_w": 0.5678316534471096,
|
| 437 |
+
"mean_cosine_w_wo_vs_w_w": 0.9864063667081514,
|
| 438 |
+
"mean_cosine_advantage": 0.009412455967104123,
|
| 439 |
+
"embedding_odd_one_out_identification_percent": 75.0,
|
| 440 |
+
"triangle_test_identification_percent": 75.0,
|
| 441 |
+
"triangle_accuracy_percent": 75.0,
|
| 442 |
+
"abx_preference_percent": 75.0,
|
| 443 |
+
"abx_preference_w_w_over_w_wo_percent": 75.0,
|
| 444 |
"random_chance_baseline_percent": 33.33
|
| 445 |
}
|
| 446 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2515752
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fbcb1a210c19103e0934ed6a96bf7b31be90b76dc4c5627b838f97d91f22811
|
| 3 |
size 2515752
|
model_f16.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1260672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:22f1d5dcda3011b375a0c5a3b354fdb6f3a2d3f385541af824fb09064fea23b1
|
| 3 |
size 1260672
|
model_fp32.tflite
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a79fbdb2a02afe726e7275d9dffa66c76c112205b99d5288bb268e4b311a4eb
|
| 3 |
+
size 1112896
|
model_int8.tflite
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7623ee6cb622bc70c7bc827a3a7023d8f7b8649de5535b44e51e74dde77b8e5d
|
| 3 |
+
size 237136
|
model_q4_k_m.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 371584
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f611740aeb4b603eefb7763a0feea6894e5bce964f541e31fb6b414d454ac519
|
| 3 |
size 371584
|
saved_model/fingerprint.pb
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12f5df13aafbe65429b8b49c98003285fcc3b578ada1df935b0a80972249a1e8
|
| 3 |
+
size 98
|
saved_model/saved_model.pb
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1450963
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c24b7122448079d3968acdb73613b43c29825da6b657a0dba0c17b7459653514
|
| 3 |
size 1450963
|
saved_model/variables/variables.data-00000-of-00001
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2560444
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b32871a91558014e68459f612d5ce0b710c8570212bc90aa10bdc0ecbac394d
|
| 3 |
size 2560444
|
saved_model/variables/variables.index
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 10404
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d20da283441124896266171725d4b403c85fe7127884e789cbcfb4663986d42
|
| 3 |
size 10404
|
training_metrics.json
CHANGED
|
@@ -3,73 +3,74 @@
|
|
| 3 |
"mode": "waveform",
|
| 4 |
"speaker_overlapping": false,
|
| 5 |
"num_train_utterances": 160,
|
|
|
|
| 6 |
"num_train_windows": 64,
|
| 7 |
"num_eval_windows": 16,
|
| 8 |
-
"training_time_sec":
|
| 9 |
"initial_eval_bits_per_sample": 7.0186,
|
| 10 |
"initial_eval_nll_nats_per_sample": 4.8649,
|
| 11 |
-
"final_eval_bits_per_sample": 4.
|
| 12 |
-
"final_eval_nll_nats_per_sample": 2.
|
| 13 |
"history": [
|
| 14 |
{
|
| 15 |
"epoch": 1,
|
| 16 |
-
"train_loss_bits_per_sample": 6.
|
| 17 |
-
"eval_cross_entropy_bits_per_sample": 6.
|
| 18 |
-
"eval_nll_nats_per_sample": 4.
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 2,
|
| 22 |
-
"train_loss_bits_per_sample": 6.
|
| 23 |
-
"eval_cross_entropy_bits_per_sample":
|
| 24 |
-
"eval_nll_nats_per_sample":
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 3,
|
| 28 |
-
"train_loss_bits_per_sample":
|
| 29 |
-
"eval_cross_entropy_bits_per_sample":
|
| 30 |
-
"eval_nll_nats_per_sample":
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 4,
|
| 34 |
-
"train_loss_bits_per_sample":
|
| 35 |
-
"eval_cross_entropy_bits_per_sample":
|
| 36 |
-
"eval_nll_nats_per_sample":
|
| 37 |
},
|
| 38 |
{
|
| 39 |
"epoch": 5,
|
| 40 |
-
"train_loss_bits_per_sample":
|
| 41 |
-
"eval_cross_entropy_bits_per_sample":
|
| 42 |
-
"eval_nll_nats_per_sample":
|
| 43 |
},
|
| 44 |
{
|
| 45 |
"epoch": 6,
|
| 46 |
-
"train_loss_bits_per_sample":
|
| 47 |
-
"eval_cross_entropy_bits_per_sample":
|
| 48 |
-
"eval_nll_nats_per_sample": 3.
|
| 49 |
},
|
| 50 |
{
|
| 51 |
"epoch": 7,
|
| 52 |
-
"train_loss_bits_per_sample": 5.
|
| 53 |
-
"eval_cross_entropy_bits_per_sample":
|
| 54 |
-
"eval_nll_nats_per_sample": 3.
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 8,
|
| 58 |
-
"train_loss_bits_per_sample": 5.
|
| 59 |
-
"eval_cross_entropy_bits_per_sample":
|
| 60 |
-
"eval_nll_nats_per_sample": 3.
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"epoch": 9,
|
| 64 |
-
"train_loss_bits_per_sample": 5.
|
| 65 |
-
"eval_cross_entropy_bits_per_sample": 4.
|
| 66 |
-
"eval_nll_nats_per_sample":
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 10,
|
| 70 |
-
"train_loss_bits_per_sample": 5.
|
| 71 |
-
"eval_cross_entropy_bits_per_sample": 4.
|
| 72 |
-
"eval_nll_nats_per_sample": 2.
|
| 73 |
}
|
| 74 |
],
|
| 75 |
"exported_files": {
|
|
@@ -78,10 +79,10 @@
|
|
| 78 |
"saved_model_dir": "pretrained_models/wavenet_waveform_coder/saved_model",
|
| 79 |
"tflite": {
|
| 80 |
"fp32_path": "pretrained_models/wavenet_waveform_coder/model_fp32.tflite",
|
| 81 |
-
"fp32_size_bytes":
|
| 82 |
"fp32_verified": true,
|
| 83 |
"int8_path": "pretrained_models/wavenet_waveform_coder/model_int8.tflite",
|
| 84 |
-
"int8_size_bytes":
|
| 85 |
"int8_verified": true,
|
| 86 |
"output_shape": [
|
| 87 |
1,
|
|
|
|
| 3 |
"mode": "waveform",
|
| 4 |
"speaker_overlapping": false,
|
| 5 |
"num_train_utterances": 160,
|
| 6 |
+
"training_audio_duration_sec": 325.12,
|
| 7 |
"num_train_windows": 64,
|
| 8 |
"num_eval_windows": 16,
|
| 9 |
+
"training_time_sec": 25.66,
|
| 10 |
"initial_eval_bits_per_sample": 7.0186,
|
| 11 |
"initial_eval_nll_nats_per_sample": 4.8649,
|
| 12 |
+
"final_eval_bits_per_sample": 4.2561,
|
| 13 |
+
"final_eval_nll_nats_per_sample": 2.9501,
|
| 14 |
"history": [
|
| 15 |
{
|
| 16 |
"epoch": 1,
|
| 17 |
+
"train_loss_bits_per_sample": 6.9295,
|
| 18 |
+
"eval_cross_entropy_bits_per_sample": 6.7548,
|
| 19 |
+
"eval_nll_nats_per_sample": 4.682
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"epoch": 2,
|
| 23 |
+
"train_loss_bits_per_sample": 6.1033,
|
| 24 |
+
"eval_cross_entropy_bits_per_sample": 5.7377,
|
| 25 |
+
"eval_nll_nats_per_sample": 3.9771
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 3,
|
| 29 |
+
"train_loss_bits_per_sample": 5.7087,
|
| 30 |
+
"eval_cross_entropy_bits_per_sample": 5.1983,
|
| 31 |
+
"eval_nll_nats_per_sample": 3.6032
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 4,
|
| 35 |
+
"train_loss_bits_per_sample": 5.5141,
|
| 36 |
+
"eval_cross_entropy_bits_per_sample": 4.8799,
|
| 37 |
+
"eval_nll_nats_per_sample": 3.3825
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 5,
|
| 41 |
+
"train_loss_bits_per_sample": 5.4215,
|
| 42 |
+
"eval_cross_entropy_bits_per_sample": 4.673,
|
| 43 |
+
"eval_nll_nats_per_sample": 3.2391
|
| 44 |
},
|
| 45 |
{
|
| 46 |
"epoch": 6,
|
| 47 |
+
"train_loss_bits_per_sample": 5.3881,
|
| 48 |
+
"eval_cross_entropy_bits_per_sample": 4.5296,
|
| 49 |
+
"eval_nll_nats_per_sample": 3.1397
|
| 50 |
},
|
| 51 |
{
|
| 52 |
"epoch": 7,
|
| 53 |
+
"train_loss_bits_per_sample": 5.3927,
|
| 54 |
+
"eval_cross_entropy_bits_per_sample": 4.4264,
|
| 55 |
+
"eval_nll_nats_per_sample": 3.0681
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 8,
|
| 59 |
+
"train_loss_bits_per_sample": 5.4239,
|
| 60 |
+
"eval_cross_entropy_bits_per_sample": 4.3507,
|
| 61 |
+
"eval_nll_nats_per_sample": 3.0157
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 9,
|
| 65 |
+
"train_loss_bits_per_sample": 5.4747,
|
| 66 |
+
"eval_cross_entropy_bits_per_sample": 4.2949,
|
| 67 |
+
"eval_nll_nats_per_sample": 2.977
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 10,
|
| 71 |
+
"train_loss_bits_per_sample": 5.5406,
|
| 72 |
+
"eval_cross_entropy_bits_per_sample": 4.2561,
|
| 73 |
+
"eval_nll_nats_per_sample": 2.9501
|
| 74 |
}
|
| 75 |
],
|
| 76 |
"exported_files": {
|
|
|
|
| 79 |
"saved_model_dir": "pretrained_models/wavenet_waveform_coder/saved_model",
|
| 80 |
"tflite": {
|
| 81 |
"fp32_path": "pretrained_models/wavenet_waveform_coder/model_fp32.tflite",
|
| 82 |
+
"fp32_size_bytes": 1112896,
|
| 83 |
"fp32_verified": true,
|
| 84 |
"int8_path": "pretrained_models/wavenet_waveform_coder/model_int8.tflite",
|
| 85 |
+
"int8_size_bytes": 237136,
|
| 86 |
"int8_verified": true,
|
| 87 |
"output_shape": [
|
| 88 |
1,
|