Prev
# text baseline
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_baseline_lats
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: /home/polovick/v2v_diff/ml/projects/ai-voice-cloning/projects-bucket-pack.json
    batch_size: 1
    gpus: 2
    inference:
      diff_steps: 400
      exp: yt4_baseline_lats/
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
      override_conditioning_features:
        c50: 0.0
        pitch_std: 100.0
        snr: 100.0
      reranking_options:
        mode: MBR
        top_k: 1
      target_len_rate: 0.75
      vocoder: univnet
    num_workers: 1
    output_dir: projects_pack_1/yt4_baseline_lats__2024-06-27_12-43-45
    ticket: TTS-388
  basket_generation_git_hash: 3ad347e2f731481af82dafbe72aa4daa136f907b
model_data_type: tts-cloning
ticket: TTS-388
version: 2024-06-27_12-43-45
encodec_mbr
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_langbycond_revgrad1_encodec-opt-bigbatch
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
    batch_size: 1
    gpus: 2
    inference:
      condition_sample_rate: 24000
      diff_on_codes: false
      diff_steps: 400
      exp: /mount/s3/tts-binary-data-nb/dimdi-y/yt4_langbycond_revgrad1_encodec-opt-bigbatch
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
        prefix_allowed_tokens_fn: encodec_interleaved_layers
        repetition_penalty_span: 50.0
        use_cache: true
      override_conditioning_features:
        c50: 0.0
        snr: 100.0
      reranking_options:
        cdist_time_downsampling_factor: 6
        mode: MBR
        top_k: 1
      vocoder: none
    num_workers: 1
    output_dir: es_en_dsat_mapping_encodec_mbrlat_override_optimizeloudness/yt4_langbycond_revgrad1_encodec-opt-bigbatch__2024-07-09_23-56-02
    ticket: TTS-393
  basket_generation_git_hash: c8fbcc9a0f18330789b7a4441a1efbeea1019935
model_data_type: tts-cloning
ticket: TTS-393
version: 2024-07-09_23-56-02
encodec_mbr_24kfix
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_langbycond_revgrad1_encodec-opt-bigbatch
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_on_codes: false
      diff_steps: 400
      exp: /mount/s3/tts-binary-data-nb/dimdi-y/yt4_langbycond_revgrad1_encodec-opt-bigbatch/
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
        prefix_allowed_tokens_fn: encodec_interleaved_layers
        repetition_penalty_span: 50.0
        use_cache: true
      out_sample_rate: 48000
      override_conditioning_features:
        snr: 100.0
      reranking_options:
        cdist_time_downsampling_factor: 6
        mode: MBR
        top_k: 1
      vocoder: none
    num_workers: 1
    output_dir: es_en_dsat_mapping_encodec_mbrlat_override_optimizeloudness/yt4_langbycond_revgrad1_encodec-opt-bigbatch__2024-07-22_14-21-18
    ticket: TTS-393
  basket_generation_git_hash: 37cb3542e38cc2685aebaf1d162ad5a0b7140f8e
model_data_type: tts-cloning
ticket: TTS-393
version: 2024-07-22_14-21-18
encodec_mbr_24kfix-inhousediff-overrides
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_langbycond_revgrad1_encodec-opt-bigbatch-diffcodes
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_on_codes: false
      diff_steps: 400
      exp: /mount/s3/tts-binary-data-nb/dimdi-y/yt4_langbycond_revgrad1_encodec-opt-bigbatch-diffcodes
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
        prefix_allowed_tokens_fn: encodec_interleaved_layers
        repetition_penalty_span: 50.0
        use_cache: true
      out_sample_rate: 24000
      override_conditioning_features:
        c50: 0.0
        pitch_std: 100.0
        snr: 100.0
      reranking_options:
        cdist_time_downsampling_factor: 2
        mode: MBR
        sakoe_chiba_radius: 24
        top_k: 1
      vocoder: bigvgan
    num_workers: 1
    output_dir: es_en_dsat_mapping_encodec_mbrlat_override_optimizeloudness_inhdiff/yt4_langbycond_revgrad1_encodec-opt-bigbatch-diffcodes__2024-08-02_03-20-48
    ticket: TTS-393
  basket_generation_git_hash: 5707d4c683a4b30e1858324363726a1e6d05bdd8
model_data_type: tts-cloning
ticket: TTS-393
version: 2024-08-02_03-20-48
encodec_mbr_24kfix-normref
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_langbycond_revgrad1_encodec-opt-bigbatch
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_on_codes: false
      diff_steps: 400
      exp: /mount/s3/tts-binary-data-nb/dimdi-y/yt4_langbycond_revgrad1_encodec-opt-bigbatch/
      force_reference_std: -23
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
        prefix_allowed_tokens_fn: encodec_interleaved_layers
        repetition_penalty_span: 50.0
        use_cache: true
      out_sample_rate: 48000
      override_conditioning_features:
        snr: 100.0
      reranking_options:
        cdist_time_downsampling_factor: 6
        mode: MBR
        top_k: 1
      vocoder: none
    num_workers: 1
    output_dir: es_en_dsat_mapping_encodec_mbrlat_override_optimizeloudness/yt4_langbycond_revgrad1_encodec-opt-bigbatch__2024-07-22_16-15-37
    ticket: TTS-393
  basket_generation_git_hash: 37cb3542e38cc2685aebaf1d162ad5a0b7140f8e
model_data_type: tts-cloning
ticket: TTS-393
version: 2024-07-22_16-15-37
encodec_mbr_24kfix-inhousediff-overrides-normloud
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_langbycond_revgrad1_encodec-opt-bigbatch-diffcodes
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_on_codes: false
      diff_steps: 400
      exp: /mount/s3/tts-binary-data-nb/dimdi-y/yt4_langbycond_revgrad1_encodec-opt-bigbatch-diffcodes
      force_reference_std: -23
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
        prefix_allowed_tokens_fn: encodec_interleaved_layers
        repetition_penalty_span: 50.0
        use_cache: true
      out_sample_rate: 24000
      override_conditioning_features:
        c50: 0.0
        pitch_std: 100.0
        snr: 100.0
      reranking_options:
        cdist_time_downsampling_factor: 2
        mode: MBR
        sakoe_chiba_radius: 24
        top_k: 1
      vocoder: bigvgan
    num_workers: 1
    output_dir: es_en_dsat_mapping_encodec_mbrlat_override_optimizeloudness_inhdiff/yt4_langbycond_revgrad1_encodec-opt-bigbatch-diffcodes__2024-08-02_10-29-24
    ticket: TTS-393
  basket_generation_git_hash: 5707d4c683a4b30e1858324363726a1e6d05bdd8
model_data_type: tts-cloning
ticket: TTS-393
version: 2024-08-02_10-29-24
encodec_mbr_normloud-contr-oversnr-dfnet
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_langbycond_revgrad1-contrastive_encodec-opt-bigbatch_normloud
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_on_codes: false
      diff_steps: 400
      exp: /mount/s3/tts-binary-data-nb/dimdi-y/yt4_langbycond_revgrad1-contrastive_encodec-opt-bigbatch_normloud/
      force_reference_std: -21.5
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
        prefix_allowed_tokens_fn: encodec_interleaved_layers
        repetition_penalty_span: 50.0
        use_cache: true
      out_sample_rate: 48000
      override_conditioning_features:
        snr: 100.0
      reranking_options:
        cdist_time_downsampling_factor: 6
        mode: MBR
        top_k: 1
      vocoder: none
    num_workers: 1
    output_dir: es_en_dsat_mapping_encodec_mbrlat_override_optimizeloudness/yt4_langbycond_revgrad1-contrastive_encodec-opt-bigbatch_normloud__2024-07-14_18-13-45
    ticket: TTS-393
  basket_generation_git_hash: c8fbcc9a0f18330789b7a4441a1efbeea1019935
model_data_type: tts-cloning
ticket: TTS-393
version: 2024-07-14_18-13-45
encodec_mbr_normloud-contr-oversnr-dfnet-24kfix
basket_config_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
data_meta: null
exp_name: yt4_langbycond_revgrad1-contrastive_encodec-opt-bigbatch_normloud
lang: en
meta:
  basket_generation_config:
    basket_lang: en
    basket_path: quality/tts/tortoise-baskets/dsat_to_en_5projects_704_fixedref.json
    batch_size: 1
    gpus: 1
    inference:
      condition_sample_rate: 24000
      diff_on_codes: false
      diff_steps: 400
      exp: /mount/s3/tts-binary-data-nb/dimdi-y/yt4_langbycond_revgrad1-contrastive_encodec-opt-bigbatch_normloud/
      force_reference_std: -21.5
      gpt_generate_args:
        do_sample: true
        num_return_sequences: 50
        prefix_allowed_tokens_fn: encodec_interleaved_layers
        repetition_penalty_span: 50.0
        use_cache: true
      out_sample_rate: 48000
      override_conditioning_features:
        snr: 100.0
      reranking_options:
        cdist_time_downsampling_factor: 6
        mode: MBR
        top_k: 1
      vocoder: none
    num_workers: 1
    output_dir: es_en_dsat_mapping_encodec_mbrlat_override_optimizeloudness/yt4_langbycond_revgrad1-contrastive_encodec-opt-bigbatch_normloud__2024-07-22_14-43-07
    ticket: TTS-393
  basket_generation_git_hash: 37cb3542e38cc2685aebaf1d162ad5a0b7140f8e
model_data_type: tts-cloning
ticket: TTS-393
version: 2024-07-22_14-43-07
450
Don't worry.
451
Look, Kat, listen to me...
452
No, please, Mika.
453
Just leave me alone, okay?
454
Involving you now would do no good.
455
I'll call you, okay?
456
Promise.
457
Your promises are worth very little.
458
Hi guys!
459
Sorry for the absence.
460
Tonight I will show you something special.
461
Stay connected.
462
Does it have to be strange for you?
463
No.
464
It is like a blind date.
465
Don't you think so?
466
Pitch darkness!
467
I like it, though.
468
Me too.
469
To this new friendship.
470
Eva. Nice name for the woman of mystery.
471
Fits like a glove.
472
I'm sorry, I didn't... I didn't tell you my name.
473
I wanted to be mysterious, too.
474
My name is.
475
Katherine...which is a good name for the girl who follows me around town. It fits like a glove.
476
I...Um...
477
Hi guys!
478
Sorry for the absence.
479
Tonight I will show you something special.
Next