Add model

README.md

@@ -1,3 +1,24 @@
 ---
 license: apache-2.0
 ---
+
+# KhanomTan TTS v1.0
+
+KhnomTan TTS (ขนมตาล) is a open-source Thai text-to-speech model that supports multilingual speakers. It supports Thai, English, and others.
+
+KhnomTan TTS is a YourTTS model that trained with supports Thai. We add the Thai speech corpus from TSync 1* and TSync 2* to [mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS](https://huggingface.co/datasets/mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS) that train the model with YourTTS model.
+
+## Config
+We have Thai characters to the graphemes config for training the model and use the Speaker Encoder model from the speaker encoder model from [🐸 Coqui-TTS](https://github.com/coqui-ai/TTS/releases/tag/speaker_encoder_model).
+
+## Dataset
+We add Tsync 1 corpus and Tsync 2 corpus, which are not complete datasets, and then add those to [mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS](https://huggingface.co/datasets/mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS) dataset.
+
+## Trained the model
+We use the 🐸 Coqui-TTS multilingual VITS-model recipe (version 0.7.1 or the commit id is d46fbc240ccf21797d42ac26cb27eb0b9f8d31c4) for training the model, and we use the speaker encoder model from [🐸 Coqui-TTS](https://github.com/coqui-ai/TTS/releases/tag/speaker_encoder_model) then we release the best model to public access.
+
+Model cards: ____
+Dataset (Tsync 1 and Tsync 2 only): ____
+GitHub: ____
+
+*Note: Those are not complete corpus. We can access the public corpus only.

best_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc1edb9b669cf2b40c24bb2fabd9cfb904fb6692a4569c5402fb1ea650ce3160
+size 1028070250

checkpoint_440000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12aaa5ec82aea04ad88bc1d78b7f89f0a1fd842cab58bf8073529b7e5095f8af
+size 1028071658

config.json

@@ -0,0 +1,299 @@
+{
+    "output_path": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts",
+    "logger_uri": null,
+    "run_name": "vits_vctk",
+    "project_name": null,
+    "run_description": "\ud83d\udc38Coqui trainer run.",
+    "print_step": 25,
+    "plot_step": 100,
+    "model_param_stats": false,
+    "wandb_entity": null,
+    "dashboard_logger": "tensorboard",
+    "log_model_step": 10000,
+    "save_step": 10000,
+    "save_n_checkpoints": 5,
+    "save_checkpoints": true,
+    "save_all_best": false,
+    "save_best_after": 10000,
+    "target_loss": null,
+    "print_eval": false,
+    "test_delay_epochs": -1,
+    "run_eval": true,
+    "run_eval_steps": null,
+    "distributed_backend": "nccl",
+    "distributed_url": "tcp://localhost:54321",
+    "mixed_precision": false,
+    "epochs": 1000,
+    "batch_size": 32,
+    "eval_batch_size": 16,
+    "grad_clip": [
+        1000.0,
+        1000.0
+    ],
+    "scheduler_after_epoch": true,
+    "lr": 0.001,
+    "optimizer": "AdamW",
+    "optimizer_params": {
+        "betas": [
+            0.8,
+            0.99
+        ],
+        "eps": 1e-09,
+        "weight_decay": 0.01
+    },
+    "lr_scheduler": "",
+    "lr_scheduler_params": {},
+    "use_grad_scaler": false,
+    "cudnn_enable": true,
+    "cudnn_deterministic": false,
+    "cudnn_benchmark": false,
+    "training_seed": 54321,
+    "model": "vits",
+    "num_loader_workers": 4,
+    "num_eval_loader_workers": 4,
+    "use_noise_augment": false,
+    "audio": {
+        "fft_size": 1024,
+        "sample_rate": 16000,
+        "win_length": 1024,
+        "hop_length": 256,
+        "num_mels": 80,
+        "mel_fmin": 0,
+        "mel_fmax": null
+    },
+    "use_phonemes": false,
+    "phonemizer": null,
+    "phoneme_language": "en-us",
+    "compute_input_seq_cache": true,
+    "text_cleaner": "multilingual_cleaners",
+    "enable_eos_bos_chars": false,
+    "test_sentences_file": "",
+    "phoneme_cache_path": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts/phoneme_cache",
+    "characters": {
+        "characters_class": "TTS.tts.models.vits.VitsCharacters",
+        "vocab_dict": null,
+        "pad": "<PAD>",
+        "eos": "<EOS>",
+        "bos": "<BOS>",
+        "blank": "<BLNK>",
+        "characters": "!\u00a1'(),-.:;\u00bf?abcdefghijklmnopqrstuvwxyz\u00b5\u00df\u00e0\u00e1\u00e2\u00e3\u00e4\u00e5\u00e6\u00e7\u00e8\u00e9\u00ea\u00eb\u00ec\u00ed\u00ee\u00ef\u00f1\u00f2\u00f3\u00f4\u00f5\u00f6\u00f9\u00fa\u00fb\u00fc\u0105\u0107\u0119\u0142\u0144\u0153\u015b\u015f\u017a\u017c\u0192\u0430\u0431\u0432\u0433\u0434\u0435\u0436\u0437\u0438\u0439\u043a\u043b\u043c\u043d\u043e\u043f\u0440\u0441\u0442\u0443\u0444\u0445\u0446\u0447\u0448\u0449\u044a\u044b\u044c\u044d\u044e\u044f\u0451\u0454\u0456\u0457\u0491\u04e7 \u00ab\u00b0\u00b1\u00b5\u00bb$%&\u2018\u2019\u201a\u201c`\u201d\u201e\u0e01\u0e02\u0e03\u0e04\u0e05\u0e06\u0e07\u0e08\u0e09\u0e0a\u0e0b\u0e0c\u0e0d\u0e0e\u0e0f\u0e10\u0e11\u0e12\u0e13\u0e14\u0e15\u0e16\u0e17\u0e18\u0e19\u0e1a\u0e1b\u0e1c\u0e1d\u0e1e\u0e1f\u0e20\u0e21\u0e22\u0e23\u0e25\u0e27\u0e28\u0e29\u0e2a\u0e2b\u0e2c\u0e2d\u0e2e\u0e24\u0e26\u0e30\u0e31\u0e32\u0e33\u0e34\u0e35\u0e36\u0e37\u0e38\u0e39\u0e40\u0e41\u0e42\u0e43\u0e44\u0e45\u0e4d\u0e47\u0e48\u0e49\u0e4a\u0e4b\u0e2f\u0e3a\u0e46\u0e4c\u0e4d\u0e4e",
+        "punctuations": "!\u00a1'(),-.:;\u00bf? ",
+        "phonemes": null,
+        "is_unique": true,
+        "is_sorted": true
+    },
+    "add_blank": true,
+    "batch_group_size": 0,
+    "loss_masking": null,
+    "min_audio_len": 32768,
+    "max_audio_len": 160000,
+    "min_text_len": 1,
+    "max_text_len": Infinity,
+    "compute_f0": false,
+    "compute_linear_spec": true,
+    "precompute_num_workers": 0,
+    "start_by_longest": false,
+    "datasets": [
+        {
+            "name": "mailabs",
+            "path": "/workspace/wt/tts/mailabs2/x-lb",
+            "meta_file_train": null,
+            "ignored_speakers": null,
+            "language": "x-lb",
+            "meta_file_val": "",
+            "meta_file_attn_mask": ""
+        },
+        {
+            "name": "mailabs",
+            "path": "/workspace/wt/tts/mailabs2/fr-fr",
+            "meta_file_train": null,
+            "ignored_speakers": null,
+            "language": "fr-fr",
+            "meta_file_val": "",
+            "meta_file_attn_mask": ""
+        },
+        {
+            "name": "mailabs",
+            "path": "/workspace/wt/tts/mailabs2/th-th",
+            "meta_file_train": null,
+            "ignored_speakers": null,
+            "language": "th-th",
+            "meta_file_val": "",
+            "meta_file_attn_mask": ""
+        },
+        {
+            "name": "mailabs",
+            "path": "/workspace/wt/tts/mailabs2/pt-br",
+            "meta_file_train": null,
+            "ignored_speakers": null,
+            "language": "pt-br",
+            "meta_file_val": "",
+            "meta_file_attn_mask": ""
+        },
+        {
+            "name": "mailabs",
+            "path": "/workspace/wt/tts/mailabs2/en",
+            "meta_file_train": null,
+            "ignored_speakers": null,
+            "language": "en",
+            "meta_file_val": "",
+            "meta_file_attn_mask": ""
+        },
+        {
+            "name": "mailabs",
+            "path": "/workspace/wt/tts/mailabs2/x-de",
+            "meta_file_train": null,
+            "ignored_speakers": null,
+            "language": "x-de",
+            "meta_file_val": "",
+            "meta_file_attn_mask": ""
+        }
+    ],
+    "test_sentences": [
+        [
+            "It took me quite a long time to develop a voice, and now that I have it I'm not going to be silent.",
+            "Linda",
+            null,
+            "en"
+        ],
+        [
+            "\u0e17\u0e14 \u0e2a\u0e2d\u0e1a \u0e01\u0e32\u0e23 \u0e2d\u0e48\u0e32\u0e19 \u0e20\u0e32\u0e29\u0e32 \u0e44\u0e17\u0e22",
+            "Tsynctwo",
+            null,
+            "th-th"
+        ]
+    ],
+    "eval_split_max_size": null,
+    "eval_split_size": 0.01,
+    "use_speaker_weighted_sampler": false,
+    "speaker_weighted_sampler_alpha": 1.0,
+    "use_language_weighted_sampler": true,
+    "language_weighted_sampler_alpha": 1.0,
+    "use_length_weighted_sampler": false,
+    "length_weighted_sampler_alpha": 1.0,
+    "model_args": {
+        "num_chars": 220,
+        "out_channels": 513,
+        "spec_segment_size": 32,
+        "hidden_channels": 192,
+        "hidden_channels_ffn_text_encoder": 768,
+        "num_heads_text_encoder": 2,
+        "num_layers_text_encoder": 6,
+        "kernel_size_text_encoder": 3,
+        "dropout_p_text_encoder": 0.1,
+        "dropout_p_duration_predictor": 0.5,
+        "kernel_size_posterior_encoder": 5,
+        "dilation_rate_posterior_encoder": 1,
+        "num_layers_posterior_encoder": 16,
+        "kernel_size_flow": 5,
+        "dilation_rate_flow": 1,
+        "num_layers_flow": 4,
+        "resblock_type_decoder": "1",
+        "resblock_kernel_sizes_decoder": [
+            3,
+            7,
+            11
+        ],
+        "resblock_dilation_sizes_decoder": [
+            [
+                1,
+                3,
+                5
+            ],
+            [
+                1,
+                3,
+                5
+            ],
+            [
+                1,
+                3,
+                5
+            ]
+        ],
+        "upsample_rates_decoder": [
+            8,
+            8,
+            2,
+            2
+        ],
+        "upsample_initial_channel_decoder": 512,
+        "upsample_kernel_sizes_decoder": [
+            16,
+            16,
+            4,
+            4
+        ],
+        "periods_multi_period_discriminator": [
+            2,
+            3,
+            5,
+            7,
+            11
+        ],
+        "use_sdp": false,
+        "noise_scale": 1.0,
+        "inference_noise_scale": 0.667,
+        "length_scale": 1.0,
+        "noise_scale_dp": 1.0,
+        "inference_noise_scale_dp": 1.0,
+        "max_inference_len": null,
+        "init_discriminator": true,
+        "use_spectral_norm_disriminator": false,
+        "use_speaker_embedding": true,
+        "num_speakers": 20,
+        "speakers_file": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts/vits_vctk-August-13-2022_06+53AM-d46fbc24/speakers.pth",
+        "d_vector_file": null,
+        "speaker_embedding_channels": 256,
+        "use_d_vector_file": false,
+        "d_vector_dim": 0,
+        "detach_dp_input": true,
+        "use_language_embedding": true,
+        "embedded_language_dim": 4,
+        "num_languages": 6,
+        "language_ids_file": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts/vits_vctk-August-13-2022_06+53AM-d46fbc24/language_ids.json",
+        "use_speaker_encoder_as_loss": false,
+        "speaker_encoder_config_path": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts/config_se.json",
+        "speaker_encoder_model_path": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts/model_se.pth",
+        "condition_dp_on_speaker": true,
+        "freeze_encoder": false,
+        "freeze_DP": false,
+        "freeze_PE": false,
+        "freeze_flow_decoder": false,
+        "freeze_waveform_decoder": false,
+        "encoder_sample_rate": null,
+        "interpolate_z": true,
+        "reinit_DP": false,
+        "reinit_text_encoder": false
+    },
+    "lr_gen": 0.0002,
+    "lr_disc": 0.0002,
+    "lr_scheduler_gen": "ExponentialLR",
+    "lr_scheduler_gen_params": {
+        "gamma": 0.999875,
+        "last_epoch": -1
+    },
+    "lr_scheduler_disc": "ExponentialLR",
+    "lr_scheduler_disc_params": {
+        "gamma": 0.999875,
+        "last_epoch": -1
+    },
+    "kl_loss_alpha": 1.0,
+    "disc_loss_alpha": 1.0,
+    "gen_loss_alpha": 1.0,
+    "feat_loss_alpha": 1.0,
+    "mel_loss_alpha": 45.0,
+    "dur_loss_alpha": 1.0,
+    "speaker_encoder_loss_alpha": 1.0,
+    "return_wav": true,
+    "r": 1,
+    "num_speakers": 0,
+    "use_speaker_embedding": true,
+    "speakers_file": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts/vits_vctk-August-13-2022_06+53AM-d46fbc24/speakers.pth",
+    "speaker_embedding_channels": 256,
+    "language_ids_file": "/workspace/wt/tts/TTS-1/recipes/multilingual/vits_tts/vits_vctk-August-13-2022_06+53AM-d46fbc24/language_ids.json",
+    "use_language_embedding": true,
+    "use_d_vector_file": false,
+    "d_vector_file": null,
+    "d_vector_dim": 0
+}

config_se.json

@@ -0,0 +1,121 @@
+{
+    "model": "speaker_encoder",
+    "run_name": "speaker_encoder",
+    "run_description": "resnet speaker encoder trained with commonvoice all languages dev and train, Voxceleb 1 dev and Voxceleb 2 dev",
+    "epochs": 100000,
+    "batch_size": null,
+    "eval_batch_size": null,
+    "mixed_precision": false,
+    "run_eval": true,
+    "test_delay_epochs": 0,
+    "print_eval": false,
+    "print_step": 50,
+    "tb_plot_step": 100,
+    "tb_model_param_stats": false,
+    "save_step": 1000,
+    "checkpoint": true,
+    "keep_all_best": false,
+    "keep_after": 10000,
+    "num_loader_workers": 8,
+    "num_val_loader_workers": 0,
+    "use_noise_augment": false,
+    "output_path": "../checkpoints/speaker_encoder/language_balanced/normalized/angleproto-4-samples-by-speakers/",
+    "distributed_backend": "nccl",
+    "distributed_url": "tcp://localhost:54321",
+    "audio": {
+        "fft_size": 512,
+        "win_length": 400,
+        "hop_length": 160,
+        "frame_shift_ms": null,
+        "frame_length_ms": null,
+        "stft_pad_mode": "reflect",
+        "sample_rate": 16000,
+        "resample": false,
+        "preemphasis": 0.97,
+        "ref_level_db": 20,
+        "do_sound_norm": false,
+        "do_trim_silence": false,
+        "trim_db": 60,
+        "power": 1.5,
+        "griffin_lim_iters": 60,
+        "num_mels": 64,
+        "mel_fmin": 0.0,
+        "mel_fmax": 8000.0,
+        "spec_gain": 20,
+        "signal_norm": false,
+        "min_level_db": -100,
+        "symmetric_norm": false,
+        "max_norm": 4.0,
+        "clip_norm": false,
+        "stats_path": null,
+        "do_rms_norm": true,
+        "db_level": -27.0
+    },
+    "datasets": [
+        {
+            "name": "voxceleb2",
+            "path": "/workspace/scratch/ecasanova/datasets/VoxCeleb/vox2_dev_aac/",
+            "meta_file_train": null,
+            "ununsed_speakers": null,
+            "meta_file_val": null,
+            "meta_file_attn_mask": "",
+            "language": "voxceleb"
+        }
+    ],
+    "model_params": {
+        "model_name": "resnet",
+        "input_dim": 64,
+        "use_torch_spec": true,
+        "log_input": true,
+        "proj_dim": 512
+    },
+    "audio_augmentation": {
+        "p": 0.5,
+        "rir": {
+            "rir_path": "/workspace/store/ecasanova/ComParE/RIRS_NOISES/simulated_rirs/",
+            "conv_mode": "full"
+        },
+        "additive": {
+            "sounds_path": "/workspace/store/ecasanova/ComParE/musan/",
+            "speech": {
+                "min_snr_in_db": 13,
+                "max_snr_in_db": 20,
+                "min_num_noises": 1,
+                "max_num_noises": 1
+            },
+            "noise": {
+                "min_snr_in_db": 0,
+                "max_snr_in_db": 15,
+                "min_num_noises": 1,
+                "max_num_noises": 1
+            },
+            "music": {
+                "min_snr_in_db": 5,
+                "max_snr_in_db": 15,
+                "min_num_noises": 1,
+                "max_num_noises": 1
+            }
+        },
+        "gaussian": {
+            "p": 0.0,
+            "min_amplitude": 0.0,
+            "max_amplitude": 1e-05
+        }
+    },
+    "storage": {
+        "sample_from_storage_p": 0.5,
+        "storage_size": 40
+    },
+    "max_train_step": 1000000,
+    "loss": "angleproto",
+    "grad_clip": 3.0,
+    "lr": 0.0001,
+    "lr_decay": false,
+    "warmup_steps": 4000,
+    "wd": 1e-06,
+    "steps_plot_stats": 100,
+    "num_speakers_in_batch": 100,
+    "num_utters_per_speaker": 4,
+    "skip_speakers": true,
+    "voice_len": 2.0
+}

language_ids.json

@@ -0,0 +1,8 @@
+{
+    "en": 0,
+    "fr-fr": 1,
+    "pt-br": 2,
+    "th-th": 3,
+    "x-de": 4,
+    "x-lb": 5
+}

model-training/events.out.tfevents.1660373622.29615ef310df

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f296dcd5b7a02a7cde7813cc362814ee19d170fb4aca22a4a1361142e83867b2
+size 592190465

model_se.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f96efb20cbeeefd81fd8336d7f0155bf8902f82f9474e58ccb19d9e12345172
+size 44610930

speakers.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f76ff12945402e4be4a1c3c0396399a60fb08a14cf2c43b9814088a6a81ee0fc
+size 687