Prev
# uttid text baseline
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_mini_100.json
data_meta: null
exp_name: gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice
lang: en-us
meta:
  basket_generation_config:
    basket_lang: en-us
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_mini_100.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_k: 3
      diff_steps: 100
      disable_optimized_diffusion: true
      duplicate_reference: true
      exp: /mount/s3/tts-binary-data-nb/dchebakov/models/gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice/
      gpt_generate_args:
        do_sample: true
        enforce_silent_start: wavtokenizer
        num_return_sequences: 50
        use_cache: true
      out_sample_rate: 24000
      override_conditioning_features:
        bad_text_proba: 0.0
        c50: 0.0
        dmcs_flatness: 100500.0
        dmcs_roll_off_0.995: 100500.0
        pitch_std: 100.0
        snr: 100.0
      reranking_options:
        mode: MBR
        top_k: 1
      target_len_rate: 1.0
      vocoder: bigvgan
      voice_samples_preprocessing:
      - demucs
      - deepfilternet
    num_workers: 1
    output_dir: dsat_mini/gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice__2025-07-07_09-37-06
    ref_dir: dsat_mini/ref
    ticket: QUALITY-41
  basket_generation_git_hash: e961ea383299ff424f66fd9505a4c7b12ba65799
model_data_type: tts-cloning
ticket: QUALITY-41
version: 2025-07-07_09-37-06
freq_feats
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_mini_100.json
data_meta: null
exp_name: gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice__diffusion_yt4_wavtokenizer_freq_feats_noref
lang: en-us
meta:
  basket_generation_config:
    basket_lang: en-us
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_mini_100.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_k: 3
      diff_steps: 100
      diffusion_exp: /mount/s3/tts-binary-data-nb/eg/exp/diffusion_yt4_wavtokenizer_freq_feats_noref/
      duplicate_reference: true
      exp: /mount/s3/tts-binary-data-nb/dchebakov/models/gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice/
      gpt_generate_args:
        do_sample: true
        enforce_silent_start: wavtokenizer
        num_return_sequences: 50
        use_cache: true
      out_sample_rate: 24000
      override_conditioning_features:
        bad_text_proba: 0.0
        c50: 0.0
        dmcs_flatness: 100500.0
        dmcs_roll_off_0.995: 100500.0
        pitch_std: 100.0
        snr: 100.0
      reranking_options:
        mode: MBR
        top_k: 1
      target_len_rate: 1.0
      vocoder: bigvgan
      voice_samples_preprocessing:
      - demucs
      - deepfilternet
    num_workers: 1
    output_dir: dsat_mini/gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice__diffusion_yt4_wavtokenizer_freq_feats_noref__2025-07-07_09-14-34
    ref_dir: dsat_mini/ref
    ticket: QUALITY-41
  basket_generation_git_hash: e961ea383299ff424f66fd9505a4c7b12ba65799
model_data_type: tts-cloning
ticket: QUALITY-41
version: 2025-07-07_09-14-34
freq_feats_cv
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_mini_100.json
data_meta: null
exp_name: gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice__diffusion_yt4_wavtokenizer_freq_feats_noref_condcv
lang: en-us
meta:
  basket_generation_config:
    basket_lang: en-us
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_mini_100.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_k: 3
      diff_steps: 100
      diffusion_exp: /mount/s3/tts-binary-data-nb/eg/exp/diffusion_yt4_wavtokenizer_freq_feats_noref_condcv/
      duplicate_reference: true
      exp: /mount/s3/tts-binary-data-nb/dchebakov/models/gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice/
      gpt_generate_args:
        do_sample: true
        enforce_silent_start: wavtokenizer
        num_return_sequences: 50
        use_cache: true
      out_sample_rate: 24000
      override_conditioning_features:
        bad_text_proba: 0.0
        c50: 0.0
        dmcs_flatness: 100500.0
        dmcs_roll_off_0.995: 100500.0
        pitch_std: 100.0
        snr: 100.0
      reranking_options:
        mode: MBR
        top_k: 1
      target_len_rate: 1.0
      vocoder: bigvgan
      voice_samples_preprocessing:
      - demucs
      - deepfilternet
    num_workers: 1
    output_dir: dsat_mini/gpt_yt4_langbycond_25freq_fsq_entropy__noclearvoice__diffusion_yt4_wavtokenizer_freq_feats_noref_condcv__2025-07-07_09-14-34
    ref_dir: dsat_mini/ref
    ticket: QUALITY-41
  basket_generation_git_hash: e961ea383299ff424f66fd9505a4c7b12ba65799
model_data_type: tts-cloning
ticket: QUALITY-41
version: 2025-07-07_09-14-34
dvae_freq_feats
basket_config_path: quality/tts/tortoise-baskets/dsat_042025_to_en_mini_100.json
data_meta: null
exp_name: yt4_en_accent_clf_entropy_t5__diffusion_yt4_en_accent_clf_entropy_t5_freq_feats_EN_noref
lang: en-us
meta:
  basket_generation_config:
    basket_lang: en-us
    basket_path: quality/tts/tortoise-baskets/dsat_042025_to_en_mini_100.json
    batch_size: 1
    gpus: 1
    inference:
      diff_k: 3
      diff_steps: 100
      diffusion_exp: /mount/s3/tts-binary-data-nb/eg/exp/diffusion_yt4_en_accent_clf_entropy_t5_freq_feats_EN_noref/
      disable_optimized_diffusion: true
      exp: /mount/s3/tts-binary-data-nb/eg/exp/yt4_en_accent_clf_entropy_t5/
      gpt_generate_args:
        do_sample: true
        enforce_silent_start: true
        num_return_sequences: 50
        use_cache: true
      override_conditioning_features:
        bad_text_proba: 0.0
        c50: 0.0
        dmcs_flatness: 100500.0
        dmcs_roll_off_0.995: 100500.0
        pitch_std: 100500.0
        snr: 100500.0
      reranking_options:
        mode: MBR
        top_k: 1
      target_len_rate: 1.0
      vocoder: bigvgan
      voice_samples_preprocessing:
      - deepfilternet
    num_workers: 1
    output_dir: dsat_042025_to_en_mini/yt4_en_accent_clf_entropy_t5__diffusion_yt4_en_accent_clf_entropy_t5_freq_feats_EN_noref__2025-06-02_13-57-41
    ref_dir: dsat_042025_to_en_mini/ref
    ticket: QUALITY-41
  basket_generation_git_hash: c185fc1cf93b68974cf48ed189a1c77d4535cc35
model_data_type: tts-cloning
ticket: QUALITY-41
version: 2025-06-02_13-57-41
DC-Profession-nwje__Folge-1-Astar-yejb_de/F0__35.297-38.083
And that kind of stuff just didn't fit into their worldview at all.
Cloned_test_Die_Biene_Maja_de_F1/Cloned_test_Die_Biene_Maja_de_F1_0018
IYUNOKonrev_Bla-5xua__Blah-Season-1_250310-hdwk_ko/M2__500.656-501.472
she would be it.
DC-Profession-nwje__Folge-5-Vanessa-djwo_de/F0__88.349-93.100
When the valves open up and you get that engine sound and you have loud music blaring on the stereo...
Cloned_test_Die_Biene_Maja_de_F2/Cloned_test_Die_Biene_Maja_de_F2_0043
Cloned_test_Die_Biene_Maja_de_F0/Cloned_test_Die_Biene_Maja_de_F0_0023
IYUNOKonrev_Bla-5xua__Blah-Season-1_250310-hdwk_ko/F1__86.167-88.550
I can't get fired again, but things aren't looking great.
Cloned_test_Die_Biene_Maja_de_F0/Cloned_test_Die_Biene_Maja_de_F0_0038
Dubbing_AD_TEST_dubf-cloned_es_M1/Dubbing_AD_TEST_dubf-cloned_es_M1_0040
Dubbing_AD_TEST_dubf-cloned_es_F0/Dubbing_AD_TEST_dubf-cloned_es_F0_0005
Rosa_de_Guadalupe_-_-56007d_es_F0/Rosa_de_Guadalupe_-_-56007d_es_F0_0046
Dubbing_AD_TEST_dubf-cloned_es_M0/Dubbing_AD_TEST_dubf-cloned_es_M0_0026
Rosa_de_Guadalupe_-_-56007d_es_M2/Rosa_de_Guadalupe_-_-56007d_es_M2_0114
CB-WINGED-TES-1jcc__vlc-record-2025-03-31-09h49m41-06pp_pt/F3__144.965-151.035
Exactly! In our families, everyone is connected to each other by a cord of love.
CB-4-min-test-onmv__H_MAGISSA_S01_E004-part1-lowre-d632_el/M1__130.863-132.713
You want us to slaughter each other over a stranger?
Cloned_test_Die_Biene_Maja_de_F0/Cloned_test_Die_Biene_Maja_de_F0_0040
Rosa_de_Guadalupe_-_-56007d_es_F0/Rosa_de_Guadalupe_-_-56007d_es_F0_0034
bardot_fr_F1/bardot_fr_F1_0017
IYUNOKonrev_Bla-5xua__Blah-Season-1_250310-hdwk_ko/F1__85.423-85.670
Yup.
CB-Chinese-mi-idgm__CHINESE_DUBBING_VIDEO-kni9_zh/F1__1.826-2.451
Enough!!
Next