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
150
Come!
151
This is Amelia; she is a friend.
152
Jean?
153
A French comrade.
154
Pleased to meet you.
155
Enchanté
156
And this bearded man is Albert, who is a contortionist in the circus.
157
He is an English journalist.
158
Enchanté, delighted.
159
Likewise.
160
Well no... we didn't see each other yesterday.
161
Is it true?
162
No.
163
How did you like my speech?
164
Well, a little too much...
165
I want too much!
166
Especially at the end. But...
167
Well... she wrote it.
168
Sorry, no, joke!
169
It was great! Really!
170
It's great!
171
First results.
172
Yes, let's go, let's go!
173
Are you coming?
174
Yes.
175
It looks like we are going to win.
176
Yes?
177
Yes.
178
What's the matter, aren't you happy?
179
When are you leaving?
Next