samikhan121 commited on
Commit
5691c0d
1 Parent(s): 2ea3826

finetune on nctb-f1 35 epochs

Browse files
best_model.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e90037e3c753f0e07a215d7dbd3478fd97d1a27cfaddadacc5dfa5a251cbd1d
3
+ size 997942774
best_model_826339.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e90037e3c753f0e07a215d7dbd3478fd97d1a27cfaddadacc5dfa5a251cbd1d
3
+ size 997942774
config.json ADDED
@@ -0,0 +1,267 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_path": "/home/azureuser/BanglaTTS/nctb-vits-single-male-9",
3
+ "logger_uri": null,
4
+ "run_name": "vits-ft-nctb",
5
+ "project_name": null,
6
+ "run_description": "\ud83d\udc38Coqui trainer run.",
7
+ "print_step": 500,
8
+ "plot_step": 100,
9
+ "model_param_stats": false,
10
+ "wandb_entity": null,
11
+ "dashboard_logger": "tensorboard",
12
+ "save_on_interrupt": true,
13
+ "log_model_step": null,
14
+ "save_step": 1000,
15
+ "save_n_checkpoints": 5,
16
+ "save_checkpoints": true,
17
+ "save_all_best": false,
18
+ "save_best_after": 0,
19
+ "target_loss": null,
20
+ "print_eval": false,
21
+ "test_delay_epochs": -1,
22
+ "run_eval": true,
23
+ "run_eval_steps": null,
24
+ "distributed_backend": "nccl",
25
+ "distributed_url": "tcp://localhost:54321",
26
+ "mixed_precision": true,
27
+ "precision": "fp16",
28
+ "epochs": 35,
29
+ "batch_size": 48,
30
+ "eval_batch_size": 8,
31
+ "grad_clip": [
32
+ 1000,
33
+ 1000
34
+ ],
35
+ "scheduler_after_epoch": true,
36
+ "lr": 0.001,
37
+ "optimizer": "AdamW",
38
+ "optimizer_params": {
39
+ "betas": [
40
+ 0.8,
41
+ 0.99
42
+ ],
43
+ "eps": 1e-09,
44
+ "weight_decay": 0.01
45
+ },
46
+ "lr_scheduler": null,
47
+ "lr_scheduler_params": {},
48
+ "use_grad_scaler": false,
49
+ "allow_tf32": false,
50
+ "cudnn_enable": true,
51
+ "cudnn_deterministic": false,
52
+ "cudnn_benchmark": true,
53
+ "training_seed": 54321,
54
+ "model": "vits",
55
+ "num_loader_workers": 8,
56
+ "num_eval_loader_workers": 4,
57
+ "use_noise_augment": false,
58
+ "audio": {
59
+ "fft_size": 1024,
60
+ "sample_rate": 16000,
61
+ "win_length": 1024,
62
+ "hop_length": 256,
63
+ "num_mels": 80,
64
+ "mel_fmin": 0,
65
+ "mel_fmax": null
66
+ },
67
+ "use_phonemes": false,
68
+ "phonemizer": null,
69
+ "phoneme_language": null,
70
+ "compute_input_seq_cache": true,
71
+ "text_cleaner": "multilingual_cleaners",
72
+ "enable_eos_bos_chars": false,
73
+ "test_sentences_file": "",
74
+ "phoneme_cache_path": null,
75
+ "characters": {
76
+ "characters_class": "TTS.tts.utils.text.characters.Graphemes",
77
+ "vocab_dict": null,
78
+ "pad": "<PAD>",
79
+ "eos": "<EOS>",
80
+ "bos": "<BOS>",
81
+ "blank": "<BLNK>",
82
+ "characters": "abcdefghijklmnopqrstuvwxyz0123456789+=/*\u221a\u09a4\u099f\u09eb\u09ad\u09bf\u0990\u098b\u0996\u098a\u09dc\u0987\u099c\u09ae\u098f\u09c7\u0998\u0999\u09b8\u09c0\u09dd\u09b9\u099e\u2018\u0988\u0995\u09a3\u09ec\u0981\u09d7\u09b6\u09a2\u09a0\u200c\u09e7\u09cd\u09e8\u09ee\u09a6\u09c3\u0994\u0997\u0993\u2014\u099b\u0989\u0982\u09ac\u09c8\u099d\u09be\u09af\u09ab\u200d\u099a\u09b0\u09b7\u0985\u09cc\u09ce\u09a5\u09a1\u09bc\u09ea\u09a7\u09e6\u09c1\u09c2\u09e9\u0986\u0983\u09aa\u09df\u2019'\u201d^\u09a8\u09b2\u09cb_\u2026\u09f0",
83
+ "punctuations": "-\u2013:;!,|.?\u0965\u0964 \u201c",
84
+ "phonemes": null,
85
+ "is_unique": true,
86
+ "is_sorted": true
87
+ },
88
+ "add_blank": true,
89
+ "batch_group_size": 5,
90
+ "loss_masking": null,
91
+ "min_audio_len": 1,
92
+ "max_audio_len": Infinity,
93
+ "min_text_len": 1,
94
+ "max_text_len": Infinity,
95
+ "compute_f0": false,
96
+ "compute_energy": false,
97
+ "compute_linear_spec": true,
98
+ "precompute_num_workers": 0,
99
+ "start_by_longest": false,
100
+ "shuffle": false,
101
+ "drop_last": false,
102
+ "datasets": [
103
+ {
104
+ "formatter": "",
105
+ "dataset_name": "",
106
+ "path": "/home/azureuser/BanglaTTS/nctb-audiobook-no-numbers/male/SP_9/",
107
+ "meta_file_train": "/home/azureuser/BanglaTTS/nctb-audiobook-no-numbers/male/SP_9/metadata.txt",
108
+ "ignored_speakers": null,
109
+ "language": "",
110
+ "phonemizer": "",
111
+ "meta_file_val": "",
112
+ "meta_file_attn_mask": ""
113
+ }
114
+ ],
115
+ "test_sentences": [
116
+ [
117
+ "\u0986\u09ae\u09b0\u09be \u09ac\u09be\u0982\u09b2\u09be\u09af\u09bc \u0993\u09af\u09bc\u09c7\u09ac \u09a1\u09c7\u09ad\u09c7\u09b2\u09aa\u09ae\u09c7\u09a8\u09cd\u099f \u09a8\u09bf\u09af\u09bc\u09c7 \u0995\u09be\u099c \u0995\u09b0\u09a4\u09c7 \u0997\u09bf\u09af\u09bc\u09c7 \u09aa\u09cd\u09b0\u09a5\u09ae \u09af\u09c7 \u09b8\u09ae\u09b8\u09cd\u09af\u09be\u099f\u09be\u09b0 \u09ae\u09c1\u0996\u09cb\u09ae\u09c1\u0996\u09bf \u09b9\u0987, \u09b8\u09c7\u099f\u09be \u09b9\u09b2\u09cb, \u09ac\u09be\u0982\u09b2\u09be \u09a1\u09c7\u09ae\u09cb \u099f\u09c7\u0995\u09cd\u09b8\u099f"
118
+ ],
119
+ [
120
+ "\u0986\u09ae\u09bf \u09ac\u09be\u0999\u09be\u09b2\u09bf \u09ad\u09be\u09b7\u09be\u09df \u0995\u09a5\u09be \u09ac\u09b2\u09a4\u09c7 \u09aa\u09be\u09b0\u09bf\u0964"
121
+ ],
122
+ [
123
+ "\u0986\u09ae\u09b0\u09be \u09aa\u09cd\u09b0\u0995\u09c3\u09a4\u09bf \u0995\u09c7 \u09ad\u09be\u09b2\u09cb\u09ac\u09be\u09b8\u09bf\u0964"
124
+ ],
125
+ [
126
+ "\u0986\u09aa\u09a8\u09bf \u0995\u09c7\u09ae\u09a8 \u0986\u099b\u09c7\u09a8?"
127
+ ]
128
+ ],
129
+ "eval_split_max_size": null,
130
+ "eval_split_size": 0.01,
131
+ "use_speaker_weighted_sampler": false,
132
+ "speaker_weighted_sampler_alpha": 1.0,
133
+ "use_language_weighted_sampler": true,
134
+ "language_weighted_sampler_alpha": 1.0,
135
+ "use_length_weighted_sampler": false,
136
+ "length_weighted_sampler_alpha": 1.0,
137
+ "model_args": {
138
+ "num_chars": 140,
139
+ "out_channels": 513,
140
+ "spec_segment_size": 32,
141
+ "hidden_channels": 192,
142
+ "hidden_channels_ffn_text_encoder": 768,
143
+ "num_heads_text_encoder": 2,
144
+ "num_layers_text_encoder": 6,
145
+ "kernel_size_text_encoder": 3,
146
+ "dropout_p_text_encoder": 0.1,
147
+ "dropout_p_duration_predictor": 0.5,
148
+ "kernel_size_posterior_encoder": 5,
149
+ "dilation_rate_posterior_encoder": 1,
150
+ "num_layers_posterior_encoder": 16,
151
+ "kernel_size_flow": 5,
152
+ "dilation_rate_flow": 1,
153
+ "num_layers_flow": 4,
154
+ "resblock_type_decoder": "1",
155
+ "resblock_kernel_sizes_decoder": [
156
+ 3,
157
+ 7,
158
+ 11
159
+ ],
160
+ "resblock_dilation_sizes_decoder": [
161
+ [
162
+ 1,
163
+ 3,
164
+ 5
165
+ ],
166
+ [
167
+ 1,
168
+ 3,
169
+ 5
170
+ ],
171
+ [
172
+ 1,
173
+ 3,
174
+ 5
175
+ ]
176
+ ],
177
+ "upsample_rates_decoder": [
178
+ 8,
179
+ 8,
180
+ 2,
181
+ 2
182
+ ],
183
+ "upsample_initial_channel_decoder": 512,
184
+ "upsample_kernel_sizes_decoder": [
185
+ 16,
186
+ 16,
187
+ 4,
188
+ 4
189
+ ],
190
+ "periods_multi_period_discriminator": [
191
+ 2,
192
+ 3,
193
+ 5,
194
+ 7,
195
+ 11
196
+ ],
197
+ "use_sdp": true,
198
+ "noise_scale": 1.0,
199
+ "inference_noise_scale": 0.667,
200
+ "length_scale": 1,
201
+ "noise_scale_dp": 1.0,
202
+ "inference_noise_scale_dp": 1.0,
203
+ "max_inference_len": null,
204
+ "init_discriminator": true,
205
+ "use_spectral_norm_disriminator": false,
206
+ "use_speaker_embedding": false,
207
+ "num_speakers": 0,
208
+ "speakers_file": null,
209
+ "d_vector_file": null,
210
+ "speaker_embedding_channels": 256,
211
+ "use_d_vector_file": false,
212
+ "d_vector_dim": 0,
213
+ "detach_dp_input": true,
214
+ "use_language_embedding": false,
215
+ "embedded_language_dim": 4,
216
+ "num_languages": 0,
217
+ "language_ids_file": null,
218
+ "use_speaker_encoder_as_loss": false,
219
+ "speaker_encoder_config_path": "",
220
+ "speaker_encoder_model_path": "",
221
+ "condition_dp_on_speaker": true,
222
+ "freeze_encoder": false,
223
+ "freeze_DP": false,
224
+ "freeze_PE": false,
225
+ "freeze_flow_decoder": false,
226
+ "freeze_waveform_decoder": false,
227
+ "encoder_sample_rate": null,
228
+ "interpolate_z": true,
229
+ "reinit_DP": false,
230
+ "reinit_text_encoder": false
231
+ },
232
+ "lr_gen": 0.0002,
233
+ "lr_disc": 0.0002,
234
+ "lr_scheduler_gen": "ExponentialLR",
235
+ "lr_scheduler_gen_params": {
236
+ "gamma": 0.999875,
237
+ "last_epoch": -1
238
+ },
239
+ "lr_scheduler_disc": "ExponentialLR",
240
+ "lr_scheduler_disc_params": {
241
+ "gamma": 0.999875,
242
+ "last_epoch": -1
243
+ },
244
+ "kl_loss_alpha": 1.0,
245
+ "disc_loss_alpha": 1.0,
246
+ "gen_loss_alpha": 1.0,
247
+ "feat_loss_alpha": 1.0,
248
+ "mel_loss_alpha": 45.0,
249
+ "dur_loss_alpha": 1.0,
250
+ "speaker_encoder_loss_alpha": 1.0,
251
+ "return_wav": true,
252
+ "use_weighted_sampler": false,
253
+ "weighted_sampler_attrs": {},
254
+ "weighted_sampler_multipliers": {},
255
+ "r": 1,
256
+ "num_speakers": 0,
257
+ "use_speaker_embedding": false,
258
+ "speakers_file": null,
259
+ "speaker_embedding_channels": 256,
260
+ "language_ids_file": null,
261
+ "use_language_embedding": false,
262
+ "use_d_vector_file": false,
263
+ "d_vector_file": null,
264
+ "d_vector_dim": 0,
265
+ "restore_path": "/home/azureuser/BanglaTTS/nctb-vits-single-female-1/checkpoint.pth",
266
+ "github_branch": "inside_docker"
267
+ }
events.out.tfevents.1711879521.poc-deploy.842678.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4b469df86600d5be2b2470001e12959fd7f913ade96aa301b1444f4604e87ba
3
+ size 38435335
train_vits.py ADDED
@@ -0,0 +1,134 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+
3
+ from trainer import Trainer, TrainerArgs
4
+
5
+ from TTS.tts.configs.shared_configs import BaseDatasetConfig,BaseAudioConfig,CharactersConfig
6
+ from TTS.tts.configs.vits_config import VitsConfig
7
+ from TTS.tts.datasets import load_tts_samples
8
+ from TTS.tts.models.vits import Vits, VitsAudioConfig
9
+ from TTS.tts.utils.text.tokenizer import TTSTokenizer
10
+ from TTS.utils.audio import AudioProcessor
11
+
12
+ output_path = os.path.dirname(os.path.abspath(__file__))
13
+ RESTORE_PATH = '/home/azureuser/BanglaTTS/nctb-vits-single-female-1/checkpoint.pth'
14
+ SPEAKER_ID = 9
15
+ SPEAKER_GENDER = 'male'
16
+ meta_file = f"/home/azureuser/BanglaTTS/nctb-audiobook-no-numbers/{SPEAKER_GENDER}/SP_{SPEAKER_ID}/metadata.txt"
17
+ root_path = f"/home/azureuser/BanglaTTS/nctb-audiobook-no-numbers/{SPEAKER_GENDER}/SP_{SPEAKER_ID}"
18
+
19
+ def formatter(root_path, meta_file, **kwargs): # pylint: disable=unused-argument
20
+ """Normalizes the LJSpeech meta data file to TTS format
21
+ https://keithito.com/LJ-Speech-Dataset/"""
22
+ txt_file = meta_file
23
+ items = []
24
+ speaker_name = f"nctb_{SPEAKER_GENDER}_{SPEAKER_ID}"
25
+ with open(txt_file, "r", encoding="utf-8") as ttf:
26
+ for line in ttf:
27
+ cols = line.split("|")
28
+ wav_file = os.path.join(root_path,'audio', cols[0])
29
+ try:
30
+ text = cols[1]
31
+ except:
32
+ print("not found")
33
+
34
+ items.append({"text": text, "audio_file": wav_file, "speaker_name": speaker_name, "root_path": root_path})
35
+ return items
36
+
37
+
38
+ dataset_config = BaseDatasetConfig(
39
+ meta_file_train=meta_file, path=os.path.join(root_path, "")
40
+ )
41
+
42
+ characters_config = CharactersConfig(
43
+ pad = '<PAD>',
44
+ eos = '<EOS>', #'<EOS>', #'।',
45
+ bos = '<BOS>',# None,
46
+ blank = '<BLNK>',
47
+ phonemes = None,
48
+ characters = "abcdefghijklmnopqrstuvwxyz0123456789+=/*√তট৫ভিঐঋখঊড়ইজমএেঘঙসীঢ়হঞ‘ঈকণ৬ঁৗশঢঠ\u200c১্২৮দৃঔগও—ছউংবৈঝাযফ\u200dচরষঅৌৎথড়৪ধ০ুূ৩আঃপয়’'”^নলো_…ৰ",
49
+ #characters = "তট৫ভিঐঋখঊড়ইজমএেঘঙসীঢ়হঞ‘ঈকণ৬ঁৗশঢঠ\u200c১্২৮দৃঔগও—ছউংবৈঝাযফ\u200dচরষঅৌৎথড়৪ধ০ুূ৩আঃপয়’নলোˌamɾʃˈonbŋlitjʰɔdkpeɟːfɡuhrʈæsʒɖwəc",
50
+ punctuations = "-–:;!,|.?॥। “",
51
+ )
52
+
53
+ #ণ´0ুয)wCছ=ক'স_{rMথd“ো+W।চঋ৷ঔ…’Eৰওঢxoঝূৎ5iটআইSyAc—ড√ল8ঁিk়াYVz‍ফLbD-শlপ য়–গ(রঐ্ঊ‘অ‌Gঈষgভ!:n;ীO?vড়aq/tRঘবএঠpধ
54
+ #ংখJঙঢ়]ৃউNহত,”নৗIfBৈmP॥sueঃৌhFমজদঞT.*েHj[
55
+
56
+ audio_config = VitsAudioConfig(
57
+ sample_rate=16000, win_length=1024, hop_length=256, num_mels=80, mel_fmin=0, mel_fmax=None
58
+ )
59
+
60
+ # VitsConfig: all model related values for training, validating and testing.
61
+
62
+ config = VitsConfig(
63
+ audio=audio_config,
64
+ run_name="vits-ft-nctb",
65
+ batch_size=48,
66
+ eval_batch_size=8,
67
+ batch_group_size=5,
68
+ num_loader_workers=8,
69
+ num_eval_loader_workers=4,
70
+ run_eval=True,
71
+ test_delay_epochs=-1,
72
+ epochs=35, # testing
73
+ # phonemizer="bn_phonemizer",# multi_phonemizer
74
+ text_cleaner='multilingual_cleaners',#'multilingual_cleaners', #"collapse_whitespace" phoneme_cleaners multilingual_cleaners
75
+ use_phonemes=False,
76
+ # phoneme_language="bn",
77
+
78
+ # phoneme_cache_path=os.path.join(output_path, "phoneme_cache"),
79
+ compute_input_seq_cache=True,
80
+ add_blank=True,
81
+ use_language_weighted_sampler = True,
82
+ print_step=500,
83
+ print_eval=False,
84
+ mixed_precision=True,
85
+ output_path=output_path,
86
+ datasets=[dataset_config],
87
+ characters = characters_config,
88
+ save_step=1000,
89
+ cudnn_benchmark=True,
90
+ # dashboard_logger = 'wandb',
91
+ test_sentences = [
92
+ ["আমরা বাংলায় ওয়েব ডেভেলপমেন্ট নিয়ে কাজ করতে গিয়ে প্রথম যে সমস্যাটার মুখোমুখি হই, সেটা হলো, বাংলা ডেমো টেক্সট"],
93
+ ["আমি বাঙালি ভাষায় কথা বলতে পারি।"],
94
+ ["আমরা প্রকৃতি কে ভালোবাসি।"],
95
+ ["আপনি কেমন আছেন?"],
96
+ ]
97
+ )
98
+
99
+ # INITIALIZE THE AUDIO PROCESSOR
100
+ # Audio processor is used for feature extraction and audio I/O.
101
+ # It mainly serves to the dataloader and the training loggers.
102
+ ap = AudioProcessor.init_from_config(config)
103
+
104
+ # INITIALIZE THE TOKENIZER
105
+ # Tokenizer is used to convert text to sequences of token IDs.
106
+ # config is updated with the default characters if not defined in the config.
107
+ tokenizer, config = TTSTokenizer.init_from_config(config)
108
+
109
+ # LOAD DATA SAMPLES
110
+ # Each sample is a list of ```[text, audio_file_path, speaker_name]```
111
+ # You can define your custom sample loader returning the list of samples.
112
+ # Or define your custom formatter and pass it to the `load_tts_samples`.
113
+ # Check `TTS.tts.datasets.load_tts_samples` for more details.
114
+ train_samples, eval_samples = load_tts_samples(
115
+ dataset_config,
116
+ formatter=formatter,
117
+ eval_split=True,
118
+ eval_split_max_size=config.eval_split_max_size,
119
+ eval_split_size=config.eval_split_size,
120
+ )
121
+
122
+ # init model
123
+ model = Vits(config, ap, tokenizer, speaker_manager=None)
124
+
125
+ # init the trainer and 🚀
126
+ trainer = Trainer(
127
+ TrainerArgs(restore_path = RESTORE_PATH),
128
+ config,
129
+ output_path,
130
+ model=model,
131
+ train_samples=train_samples,
132
+ eval_samples=eval_samples,
133
+ )
134
+ trainer.fit()
trainer_0_log.txt ADDED
@@ -0,0 +1,958 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ > Training Environment:
2
+ | > Backend: Torch
3
+ | > Mixed precision: True
4
+ | > Precision: fp16
5
+ | > Current device: 0
6
+ | > Num. of GPUs: 1
7
+ | > Num. of CPUs: 24
8
+ | > Num. of Torch Threads: 24
9
+ | > Torch seed: 54321
10
+ | > Torch CUDNN: True
11
+ | > Torch CUDNN deterministic: False
12
+ | > Torch CUDNN benchmark: True
13
+ | > Torch TF32 MatMul: False
14
+ > Start Tensorboard: tensorboard --logdir=/home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
15
+ > Restoring from checkpoint.pth ...
16
+ > Restoring Model...
17
+ > Partial model initialization...
18
+ | > Layer dimention missmatch between model definition and checkpoint: text_encoder.emb.weight
19
+ | > 948 / 949 layers are restored.
20
+ > Model restored from step 826000
21
+
22
+ > Model has 83060908 parameters
23
+
24
+  > EPOCH: 0/35
25
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
26
+
27
+  > TRAINING (2024-03-31 10:05:23) 
28
+
29
+  > EVALUATION 
30
+
31
+
32
+ --> EVAL PERFORMANCE
33
+ | > avg_loader_time: 0.0025544166564941406 (+0)
34
+ | > avg_loss_disc: 2.579561471939087 (+0)
35
+ | > avg_loss_disc_real_0: 0.22504794597625732 (+0)
36
+ | > avg_loss_disc_real_1: 0.23633995652198792 (+0)
37
+ | > avg_loss_disc_real_2: 0.23231595754623413 (+0)
38
+ | > avg_loss_disc_real_3: 0.24025149643421173 (+0)
39
+ | > avg_loss_disc_real_4: 0.18514925241470337 (+0)
40
+ | > avg_loss_disc_real_5: 0.22825410962104797 (+0)
41
+ | > avg_loss_0: 2.579561471939087 (+0)
42
+ | > avg_loss_gen: 2.095607280731201 (+0)
43
+ | > avg_loss_kl: 2.954455852508545 (+0)
44
+ | > avg_loss_feat: 6.803194046020508 (+0)
45
+ | > avg_loss_mel: 16.120973587036133 (+0)
46
+ | > avg_loss_duration: 1.8933916091918945 (+0)
47
+ | > avg_loss_1: 29.86762237548828 (+0)
48
+
49
+ > BEST MODEL : /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000/best_model_826027.pth
50
+
51
+  > EPOCH: 1/35
52
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
53
+
54
+  > TRAINING (2024-03-31 10:06:29) 
55
+
56
+  > EVALUATION 
57
+
58
+
59
+ --> EVAL PERFORMANCE
60
+ | > avg_loader_time: 0.0036211013793945312 (+0.0010666847229003906)
61
+ | > avg_loss_disc: 2.360954523086548 (-0.21860694885253906)
62
+ | > avg_loss_disc_real_0: 0.11433742940425873 (-0.1107105165719986)
63
+ | > avg_loss_disc_real_1: 0.11346497386693954 (-0.12287498265504837)
64
+ | > avg_loss_disc_real_2: 0.14866462349891663 (-0.0836513340473175)
65
+ | > avg_loss_disc_real_3: 0.2403942048549652 (+0.000142708420753479)
66
+ | > avg_loss_disc_real_4: 0.24388839304447174 (+0.05873914062976837)
67
+ | > avg_loss_disc_real_5: 0.19841952621936798 (-0.029834583401679993)
68
+ | > avg_loss_0: 2.360954523086548 (-0.21860694885253906)
69
+ | > avg_loss_gen: 2.211665153503418 (+0.1160578727722168)
70
+ | > avg_loss_kl: 2.9490206241607666 (-0.00543522834777832)
71
+ | > avg_loss_feat: 8.537314414978027 (+1.7341203689575195)
72
+ | > avg_loss_mel: 22.604536056518555 (+6.483562469482422)
73
+ | > avg_loss_duration: 1.9063528776168823 (+0.012961268424987793)
74
+ | > avg_loss_1: 38.20888900756836 (+8.341266632080078)
75
+
76
+
77
+  > EPOCH: 2/35
78
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
79
+
80
+  > TRAINING (2024-03-31 10:06:51) 
81
+
82
+  > EVALUATION 
83
+
84
+
85
+ --> EVAL PERFORMANCE
86
+ | > avg_loader_time: 0.0029449462890625 (-0.0006761550903320312)
87
+ | > avg_loss_disc: 2.836214780807495 (+0.47526025772094727)
88
+ | > avg_loss_disc_real_0: 0.18877620995044708 (+0.07443878054618835)
89
+ | > avg_loss_disc_real_1: 0.23450841009616852 (+0.12104343622922897)
90
+ | > avg_loss_disc_real_2: 0.17941533029079437 (+0.030750706791877747)
91
+ | > avg_loss_disc_real_3: 0.21590179204940796 (-0.02449241280555725)
92
+ | > avg_loss_disc_real_4: 0.2847995460033417 (+0.040911152958869934)
93
+ | > avg_loss_disc_real_5: 0.20625822246074677 (+0.007838696241378784)
94
+ | > avg_loss_0: 2.836214780807495 (+0.47526025772094727)
95
+ | > avg_loss_gen: 1.9285584688186646 (-0.2831066846847534)
96
+ | > avg_loss_kl: 2.9806227684020996 (+0.03160214424133301)
97
+ | > avg_loss_feat: 7.743529319763184 (-0.7937850952148438)
98
+ | > avg_loss_mel: 17.637630462646484 (-4.96690559387207)
99
+ | > avg_loss_duration: 1.8946624994277954 (-0.011690378189086914)
100
+ | > avg_loss_1: 32.185001373291016 (-6.023887634277344)
101
+
102
+
103
+  > EPOCH: 3/35
104
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
105
+
106
+  > TRAINING (2024-03-31 10:07:10) 
107
+
108
+  > EVALUATION 
109
+
110
+
111
+ --> EVAL PERFORMANCE
112
+ | > avg_loader_time: 0.0026216506958007812 (-0.00032329559326171875)
113
+ | > avg_loss_disc: 2.4784274101257324 (-0.3577873706817627)
114
+ | > avg_loss_disc_real_0: 0.14986148476600647 (-0.03891472518444061)
115
+ | > avg_loss_disc_real_1: 0.1897958517074585 (-0.04471255838871002)
116
+ | > avg_loss_disc_real_2: 0.16911624372005463 (-0.010299086570739746)
117
+ | > avg_loss_disc_real_3: 0.25678882002830505 (+0.040887027978897095)
118
+ | > avg_loss_disc_real_4: 0.16667822003364563 (-0.11812132596969604)
119
+ | > avg_loss_disc_real_5: 0.11786417663097382 (-0.08839404582977295)
120
+ | > avg_loss_0: 2.4784274101257324 (-0.3577873706817627)
121
+ | > avg_loss_gen: 2.258392333984375 (+0.32983386516571045)
122
+ | > avg_loss_kl: 2.243771553039551 (-0.7368512153625488)
123
+ | > avg_loss_feat: 10.215510368347168 (+2.4719810485839844)
124
+ | > avg_loss_mel: 20.45659065246582 (+2.818960189819336)
125
+ | > avg_loss_duration: 1.8232922554016113 (-0.07137024402618408)
126
+ | > avg_loss_1: 36.997554779052734 (+4.812553405761719)
127
+
128
+
129
+  > EPOCH: 4/35
130
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
131
+
132
+  > TRAINING (2024-03-31 10:07:30) 
133
+
134
+  > EVALUATION 
135
+
136
+
137
+ --> EVAL PERFORMANCE
138
+ | > avg_loader_time: 0.002149343490600586 (-0.0004723072052001953)
139
+ | > avg_loss_disc: 2.8351030349731445 (+0.3566756248474121)
140
+ | > avg_loss_disc_real_0: 0.202328160405159 (+0.05246667563915253)
141
+ | > avg_loss_disc_real_1: 0.19721390306949615 (+0.007418051362037659)
142
+ | > avg_loss_disc_real_2: 0.19663220643997192 (+0.027515962719917297)
143
+ | > avg_loss_disc_real_3: 0.19390971958637238 (-0.06287910044193268)
144
+ | > avg_loss_disc_real_4: 0.2577243447303772 (+0.09104612469673157)
145
+ | > avg_loss_disc_real_5: 0.3611280024051666 (+0.2432638257741928)
146
+ | > avg_loss_0: 2.8351030349731445 (+0.3566756248474121)
147
+ | > avg_loss_gen: 2.100310802459717 (-0.1580815315246582)
148
+ | > avg_loss_kl: 2.5330567359924316 (+0.28928518295288086)
149
+ | > avg_loss_feat: 7.338407039642334 (-2.877103328704834)
150
+ | > avg_loss_mel: 19.65290641784668 (-0.8036842346191406)
151
+ | > avg_loss_duration: 1.7398550510406494 (-0.08343720436096191)
152
+ | > avg_loss_1: 33.36453628540039 (-3.6330184936523438)
153
+
154
+
155
+  > EPOCH: 5/35
156
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
157
+
158
+  > TRAINING (2024-03-31 10:07:48) 
159
+
160
+  > EVALUATION 
161
+
162
+
163
+ --> EVAL PERFORMANCE
164
+ | > avg_loader_time: 0.003258228302001953 (+0.0011088848114013672)
165
+ | > avg_loss_disc: 2.8697474002838135 (+0.034644365310668945)
166
+ | > avg_loss_disc_real_0: 0.1347222477197647 (-0.06760591268539429)
167
+ | > avg_loss_disc_real_1: 0.15822717547416687 (-0.038986727595329285)
168
+ | > avg_loss_disc_real_2: 0.1845826804637909 (-0.01204952597618103)
169
+ | > avg_loss_disc_real_3: 0.2523535192012787 (+0.05844379961490631)
170
+ | > avg_loss_disc_real_4: 0.14204007387161255 (-0.11568427085876465)
171
+ | > avg_loss_disc_real_5: 0.1628914326429367 (-0.19823656976222992)
172
+ | > avg_loss_0: 2.8697474002838135 (+0.034644365310668945)
173
+ | > avg_loss_gen: 1.76955246925354 (-0.33075833320617676)
174
+ | > avg_loss_kl: 1.7157461643218994 (-0.8173105716705322)
175
+ | > avg_loss_feat: 9.666239738464355 (+2.3278326988220215)
176
+ | > avg_loss_mel: 21.284114837646484 (+1.6312084197998047)
177
+ | > avg_loss_duration: 1.681728720664978 (-0.05812633037567139)
178
+ | > avg_loss_1: 36.11738204956055 (+2.7528457641601562)
179
+
180
+
181
+  > EPOCH: 6/35
182
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
183
+
184
+  > TRAINING (2024-03-31 10:08:08) 
185
+
186
+  > EVALUATION 
187
+
188
+
189
+ --> EVAL PERFORMANCE
190
+ | > avg_loader_time: 0.0019330978393554688 (-0.0013251304626464844)
191
+ | > avg_loss_disc: 2.5998082160949707 (-0.2699391841888428)
192
+ | > avg_loss_disc_real_0: 0.10893561691045761 (-0.0257866308093071)
193
+ | > avg_loss_disc_real_1: 0.21296951174736023 (+0.05474233627319336)
194
+ | > avg_loss_disc_real_2: 0.22080537676811218 (+0.03622269630432129)
195
+ | > avg_loss_disc_real_3: 0.1968725025653839 (-0.055481016635894775)
196
+ | > avg_loss_disc_real_4: 0.3077707290649414 (+0.16573065519332886)
197
+ | > avg_loss_disc_real_5: 0.11337371170520782 (-0.04951772093772888)
198
+ | > avg_loss_0: 2.5998082160949707 (-0.2699391841888428)
199
+ | > avg_loss_gen: 2.1091349124908447 (+0.3395824432373047)
200
+ | > avg_loss_kl: 2.2263669967651367 (+0.5106208324432373)
201
+ | > avg_loss_feat: 11.159883499145508 (+1.4936437606811523)
202
+ | > avg_loss_mel: 21.78208351135254 (+0.4979686737060547)
203
+ | > avg_loss_duration: 1.6071841716766357 (-0.07454454898834229)
204
+ | > avg_loss_1: 38.8846549987793 (+2.76727294921875)
205
+
206
+
207
+  > EPOCH: 7/35
208
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
209
+
210
+  > TRAINING (2024-03-31 10:08:29) 
211
+
212
+  > EVALUATION 
213
+
214
+
215
+ --> EVAL PERFORMANCE
216
+ | > avg_loader_time: 0.0019304752349853516 (-2.6226043701171875e-06)
217
+ | > avg_loss_disc: 2.4500019550323486 (-0.14980626106262207)
218
+ | > avg_loss_disc_real_0: 0.09126907587051392 (-0.017666541039943695)
219
+ | > avg_loss_disc_real_1: 0.13244855403900146 (-0.08052095770835876)
220
+ | > avg_loss_disc_real_2: 0.116974376142025 (-0.10383100062608719)
221
+ | > avg_loss_disc_real_3: 0.13961941003799438 (-0.057253092527389526)
222
+ | > avg_loss_disc_real_4: 0.1451183557510376 (-0.1626523733139038)
223
+ | > avg_loss_disc_real_5: 0.11358281224966049 (+0.00020910054445266724)
224
+ | > avg_loss_0: 2.4500019550323486 (-0.14980626106262207)
225
+ | > avg_loss_gen: 1.6871346235275269 (-0.42200028896331787)
226
+ | > avg_loss_kl: 2.055893659591675 (-0.17047333717346191)
227
+ | > avg_loss_feat: 12.322132110595703 (+1.1622486114501953)
228
+ | > avg_loss_mel: 22.194091796875 (+0.41200828552246094)
229
+ | > avg_loss_duration: 1.6218042373657227 (+0.014620065689086914)
230
+ | > avg_loss_1: 39.88105392456055 (+0.99639892578125)
231
+
232
+
233
+  > EPOCH: 8/35
234
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
235
+
236
+  > TRAINING (2024-03-31 10:08:49) 
237
+
238
+  > EVALUATION 
239
+
240
+
241
+ --> EVAL PERFORMANCE
242
+ | > avg_loader_time: 0.0020303726196289062 (+9.989738464355469e-05)
243
+ | > avg_loss_disc: 2.458998680114746 (+0.008996725082397461)
244
+ | > avg_loss_disc_real_0: 0.1939530372619629 (+0.10268396139144897)
245
+ | > avg_loss_disc_real_1: 0.18585415184497833 (+0.05340559780597687)
246
+ | > avg_loss_disc_real_2: 0.22415761649608612 (+0.10718324035406113)
247
+ | > avg_loss_disc_real_3: 0.23032917082309723 (+0.09070976078510284)
248
+ | > avg_loss_disc_real_4: 0.256417453289032 (+0.11129909753799438)
249
+ | > avg_loss_disc_real_5: 0.22092194855213165 (+0.10733913630247116)
250
+ | > avg_loss_0: 2.458998680114746 (+0.008996725082397461)
251
+ | > avg_loss_gen: 2.281503200531006 (+0.594368577003479)
252
+ | > avg_loss_kl: 2.1694588661193848 (+0.11356520652770996)
253
+ | > avg_loss_feat: 6.771675109863281 (-5.550457000732422)
254
+ | > avg_loss_mel: 19.732463836669922 (-2.461627960205078)
255
+ | > avg_loss_duration: 1.5910258293151855 (-0.03077840805053711)
256
+ | > avg_loss_1: 32.54612731933594 (-7.334926605224609)
257
+
258
+
259
+  > EPOCH: 9/35
260
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
261
+
262
+  > TRAINING (2024-03-31 10:09:12) 
263
+
264
+  > EVALUATION 
265
+
266
+
267
+ --> EVAL PERFORMANCE
268
+ | > avg_loader_time: 0.002134561538696289 (+0.00010418891906738281)
269
+ | > avg_loss_disc: 2.3202075958251953 (-0.13879108428955078)
270
+ | > avg_loss_disc_real_0: 0.1641819328069687 (-0.0297711044549942)
271
+ | > avg_loss_disc_real_1: 0.2069983035326004 (+0.02114415168762207)
272
+ | > avg_loss_disc_real_2: 0.19238199293613434 (-0.03177562355995178)
273
+ | > avg_loss_disc_real_3: 0.21948115527629852 (-0.010848015546798706)
274
+ | > avg_loss_disc_real_4: 0.20658878982067108 (-0.0498286634683609)
275
+ | > avg_loss_disc_real_5: 0.21016131341457367 (-0.010760635137557983)
276
+ | > avg_loss_0: 2.3202075958251953 (-0.13879108428955078)
277
+ | > avg_loss_gen: 2.370643138885498 (+0.08913993835449219)
278
+ | > avg_loss_kl: 1.8640230894088745 (-0.30543577671051025)
279
+ | > avg_loss_feat: 7.0724196434021 (+0.30074453353881836)
280
+ | > avg_loss_mel: 20.123937606811523 (+0.39147377014160156)
281
+ | > avg_loss_duration: 1.5861397981643677 (-0.004886031150817871)
282
+ | > avg_loss_1: 33.01716232299805 (+0.4710350036621094)
283
+
284
+
285
+  > EPOCH: 10/35
286
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
287
+
288
+  > TRAINING (2024-03-31 10:09:34) 
289
+
290
+  > EVALUATION 
291
+
292
+
293
+ --> EVAL PERFORMANCE
294
+ | > avg_loader_time: 0.0019383430480957031 (-0.00019621849060058594)
295
+ | > avg_loss_disc: 2.6848320960998535 (+0.3646245002746582)
296
+ | > avg_loss_disc_real_0: 0.1019434928894043 (-0.06223843991756439)
297
+ | > avg_loss_disc_real_1: 0.22225375473499298 (+0.015255451202392578)
298
+ | > avg_loss_disc_real_2: 0.29437896609306335 (+0.10199697315692902)
299
+ | > avg_loss_disc_real_3: 0.369914710521698 (+0.15043355524539948)
300
+ | > avg_loss_disc_real_4: 0.2501886785030365 (+0.04359988868236542)
301
+ | > avg_loss_disc_real_5: 0.2736251652240753 (+0.06346385180950165)
302
+ | > avg_loss_0: 2.6848320960998535 (+0.3646245002746582)
303
+ | > avg_loss_gen: 2.2344932556152344 (-0.13614988327026367)
304
+ | > avg_loss_kl: 2.1780905723571777 (+0.3140674829483032)
305
+ | > avg_loss_feat: 8.309244155883789 (+1.2368245124816895)
306
+ | > avg_loss_mel: 18.31509780883789 (-1.8088397979736328)
307
+ | > avg_loss_duration: 1.5895346403121948 (+0.0033948421478271484)
308
+ | > avg_loss_1: 32.626461029052734 (-0.3907012939453125)
309
+
310
+
311
+  > EPOCH: 11/35
312
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
313
+
314
+  > TRAINING (2024-03-31 10:09:59) 
315
+
316
+  > EVALUATION 
317
+
318
+
319
+ --> EVAL PERFORMANCE
320
+ | > avg_loader_time: 0.0018219947814941406 (-0.0001163482666015625)
321
+ | > avg_loss_disc: 2.6257660388946533 (-0.059066057205200195)
322
+ | > avg_loss_disc_real_0: 0.21502818167209625 (+0.11308468878269196)
323
+ | > avg_loss_disc_real_1: 0.17747360467910767 (-0.044780150055885315)
324
+ | > avg_loss_disc_real_2: 0.17541557550430298 (-0.11896339058876038)
325
+ | > avg_loss_disc_real_3: 0.18490779399871826 (-0.18500691652297974)
326
+ | > avg_loss_disc_real_4: 0.17014773190021515 (-0.08004094660282135)
327
+ | > avg_loss_disc_real_5: 0.16564661264419556 (-0.10797855257987976)
328
+ | > avg_loss_0: 2.6257660388946533 (-0.059066057205200195)
329
+ | > avg_loss_gen: 1.820864200592041 (-0.41362905502319336)
330
+ | > avg_loss_kl: 2.026675224304199 (-0.15141534805297852)
331
+ | > avg_loss_feat: 10.131511688232422 (+1.8222675323486328)
332
+ | > avg_loss_mel: 20.05714988708496 (+1.7420520782470703)
333
+ | > avg_loss_duration: 1.570576548576355 (-0.018958091735839844)
334
+ | > avg_loss_1: 35.60677719116211 (+2.980316162109375)
335
+
336
+
337
+  > EPOCH: 12/35
338
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
339
+
340
+  > TRAINING (2024-03-31 10:10:21) 
341
+
342
+  > EVALUATION 
343
+
344
+
345
+ --> EVAL PERFORMANCE
346
+ | > avg_loader_time: 0.0019230842590332031 (+0.0001010894775390625)
347
+ | > avg_loss_disc: 2.7327022552490234 (+0.10693621635437012)
348
+ | > avg_loss_disc_real_0: 0.23379340767860413 (+0.018765226006507874)
349
+ | > avg_loss_disc_real_1: 0.24418646097183228 (+0.06671285629272461)
350
+ | > avg_loss_disc_real_2: 0.2957330346107483 (+0.12031745910644531)
351
+ | > avg_loss_disc_real_3: 0.2668507397174835 (+0.08194294571876526)
352
+ | > avg_loss_disc_real_4: 0.30550694465637207 (+0.13535921275615692)
353
+ | > avg_loss_disc_real_5: 0.28343188762664795 (+0.11778527498245239)
354
+ | > avg_loss_0: 2.7327022552490234 (+0.10693621635437012)
355
+ | > avg_loss_gen: 2.1545863151550293 (+0.3337221145629883)
356
+ | > avg_loss_kl: 1.9101665019989014 (-0.11650872230529785)
357
+ | > avg_loss_feat: 3.6649274826049805 (-6.466584205627441)
358
+ | > avg_loss_mel: 9.204965591430664 (-10.852184295654297)
359
+ | > avg_loss_duration: 1.5653738975524902 (-0.005202651023864746)
360
+ | > avg_loss_1: 18.50002098083496 (-17.10675621032715)
361
+
362
+ > BEST MODEL : /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000/best_model_826339.pth
363
+
364
+  > EPOCH: 13/35
365
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
366
+
367
+  > TRAINING (2024-03-31 10:10:45) 
368
+
369
+  > EVALUATION 
370
+
371
+
372
+ --> EVAL PERFORMANCE
373
+ | > avg_loader_time: 0.002056598663330078 (+0.000133514404296875)
374
+ | > avg_loss_disc: 2.495265245437622 (-0.23743700981140137)
375
+ | > avg_loss_disc_real_0: 0.1456478089094162 (-0.08814559876918793)
376
+ | > avg_loss_disc_real_1: 0.29628562927246094 (+0.05209916830062866)
377
+ | > avg_loss_disc_real_2: 0.23488187789916992 (-0.06085115671157837)
378
+ | > avg_loss_disc_real_3: 0.24811452627182007 (-0.018736213445663452)
379
+ | > avg_loss_disc_real_4: 0.2415287345647812 (-0.06397821009159088)
380
+ | > avg_loss_disc_real_5: 0.23950818181037903 (-0.04392370581626892)
381
+ | > avg_loss_0: 2.495265245437622 (-0.23743700981140137)
382
+ | > avg_loss_gen: 2.332970380783081 (+0.17838406562805176)
383
+ | > avg_loss_kl: 2.0096042156219482 (+0.09943771362304688)
384
+ | > avg_loss_feat: 4.751073837280273 (+1.086146354675293)
385
+ | > avg_loss_mel: 14.550800323486328 (+5.345834732055664)
386
+ | > avg_loss_duration: 1.5579960346221924 (-0.0073778629302978516)
387
+ | > avg_loss_1: 25.202444076538086 (+6.702423095703125)
388
+
389
+
390
+  > EPOCH: 14/35
391
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
392
+
393
+  > TRAINING (2024-03-31 10:11:08) 
394
+
395
+  > EVALUATION 
396
+
397
+
398
+ --> EVAL PERFORMANCE
399
+ | > avg_loader_time: 0.001909017562866211 (-0.0001475811004638672)
400
+ | > avg_loss_disc: 2.571105480194092 (+0.07584023475646973)
401
+ | > avg_loss_disc_real_0: 0.14245551824569702 (-0.0031922906637191772)
402
+ | > avg_loss_disc_real_1: 0.16970688104629517 (-0.12657874822616577)
403
+ | > avg_loss_disc_real_2: 0.2034100443124771 (-0.03147183358669281)
404
+ | > avg_loss_disc_real_3: 0.216097891330719 (-0.032016634941101074)
405
+ | > avg_loss_disc_real_4: 0.28562402725219727 (+0.04409529268741608)
406
+ | > avg_loss_disc_real_5: 0.16510680317878723 (-0.0744013786315918)
407
+ | > avg_loss_0: 2.571105480194092 (+0.07584023475646973)
408
+ | > avg_loss_gen: 1.963249683380127 (-0.3697206974029541)
409
+ | > avg_loss_kl: 2.300732135772705 (+0.29112792015075684)
410
+ | > avg_loss_feat: 7.8136515617370605 (+3.062577724456787)
411
+ | > avg_loss_mel: 19.084020614624023 (+4.533220291137695)
412
+ | > avg_loss_duration: 1.556723952293396 (-0.0012720823287963867)
413
+ | > avg_loss_1: 32.71837615966797 (+7.515932083129883)
414
+
415
+
416
+  > EPOCH: 15/35
417
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
418
+
419
+  > TRAINING (2024-03-31 10:11:32) 
420
+
421
+  > EVALUATION 
422
+
423
+
424
+ --> EVAL PERFORMANCE
425
+ | > avg_loader_time: 0.0020017623901367188 (+9.274482727050781e-05)
426
+ | > avg_loss_disc: 2.7033262252807617 (+0.13222074508666992)
427
+ | > avg_loss_disc_real_0: 0.1850876808166504 (+0.04263216257095337)
428
+ | > avg_loss_disc_real_1: 0.2569637894630432 (+0.08725690841674805)
429
+ | > avg_loss_disc_real_2: 0.24873216450214386 (+0.04532212018966675)
430
+ | > avg_loss_disc_real_3: 0.2321649193763733 (+0.016067028045654297)
431
+ | > avg_loss_disc_real_4: 0.25387001037597656 (-0.0317540168762207)
432
+ | > avg_loss_disc_real_5: 0.3422059416770935 (+0.17709913849830627)
433
+ | > avg_loss_0: 2.7033262252807617 (+0.13222074508666992)
434
+ | > avg_loss_gen: 2.466217517852783 (+0.5029678344726562)
435
+ | > avg_loss_kl: 2.473677158355713 (+0.1729450225830078)
436
+ | > avg_loss_feat: 8.864312171936035 (+1.0506606101989746)
437
+ | > avg_loss_mel: 20.450681686401367 (+1.3666610717773438)
438
+ | > avg_loss_duration: 1.5559325218200684 (-0.0007914304733276367)
439
+ | > avg_loss_1: 35.810821533203125 (+3.0924453735351562)
440
+
441
+
442
+  > EPOCH: 16/35
443
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
444
+
445
+  > TRAINING (2024-03-31 10:11:55) 
446
+
447
+  > EVALUATION 
448
+
449
+
450
+ --> EVAL PERFORMANCE
451
+ | > avg_loader_time: 0.0018572807312011719 (-0.00014448165893554688)
452
+ | > avg_loss_disc: 2.526150703430176 (-0.17717552185058594)
453
+ | > avg_loss_disc_real_0: 0.22012364864349365 (+0.03503596782684326)
454
+ | > avg_loss_disc_real_1: 0.20115916430950165 (-0.055804625153541565)
455
+ | > avg_loss_disc_real_2: 0.17505493760108948 (-0.07367722690105438)
456
+ | > avg_loss_disc_real_3: 0.17389781773090363 (-0.058267101645469666)
457
+ | > avg_loss_disc_real_4: 0.2565661072731018 (+0.002696096897125244)
458
+ | > avg_loss_disc_real_5: 0.1770845651626587 (-0.16512137651443481)
459
+ | > avg_loss_0: 2.526150703430176 (-0.17717552185058594)
460
+ | > avg_loss_gen: 2.2745003700256348 (-0.19171714782714844)
461
+ | > avg_loss_kl: 1.9097598791122437 (-0.5639172792434692)
462
+ | > avg_loss_feat: 7.223072052001953 (-1.641240119934082)
463
+ | > avg_loss_mel: 18.74079704284668 (-1.7098846435546875)
464
+ | > avg_loss_duration: 1.5443365573883057 (-0.011595964431762695)
465
+ | > avg_loss_1: 31.69246482849121 (-4.118356704711914)
466
+
467
+
468
+  > EPOCH: 17/35
469
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
470
+
471
+  > TRAINING (2024-03-31 10:12:18) 
472
+
473
+  > EVALUATION 
474
+
475
+
476
+ --> EVAL PERFORMANCE
477
+ | > avg_loader_time: 0.0019030570983886719 (+4.57763671875e-05)
478
+ | > avg_loss_disc: 2.5047130584716797 (-0.021437644958496094)
479
+ | > avg_loss_disc_real_0: 0.2906636595726013 (+0.07054001092910767)
480
+ | > avg_loss_disc_real_1: 0.20504474639892578 (+0.0038855820894241333)
481
+ | > avg_loss_disc_real_2: 0.18310394883155823 (+0.00804901123046875)
482
+ | > avg_loss_disc_real_3: 0.25080615282058716 (+0.07690833508968353)
483
+ | > avg_loss_disc_real_4: 0.3209937810897827 (+0.06442767381668091)
484
+ | > avg_loss_disc_real_5: 0.24696961045265198 (+0.06988504528999329)
485
+ | > avg_loss_0: 2.5047130584716797 (-0.021437644958496094)
486
+ | > avg_loss_gen: 2.532362699508667 (+0.2578623294830322)
487
+ | > avg_loss_kl: 2.36430025100708 (+0.4545403718948364)
488
+ | > avg_loss_feat: 6.0262274742126465 (-1.1968445777893066)
489
+ | > avg_loss_mel: 19.45924186706543 (+0.71844482421875)
490
+ | > avg_loss_duration: 1.5199471712112427 (-0.02438938617706299)
491
+ | > avg_loss_1: 31.90207862854004 (+0.20961380004882812)
492
+
493
+
494
+  > EPOCH: 18/35
495
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
496
+
497
+  > TRAINING (2024-03-31 10:12:43) 
498
+
499
+  > EVALUATION 
500
+
501
+
502
+ --> EVAL PERFORMANCE
503
+ | > avg_loader_time: 0.001995563507080078 (+9.250640869140625e-05)
504
+ | > avg_loss_disc: 2.2198214530944824 (-0.28489160537719727)
505
+ | > avg_loss_disc_real_0: 0.13773468136787415 (-0.15292897820472717)
506
+ | > avg_loss_disc_real_1: 0.1682843267917633 (-0.036760419607162476)
507
+ | > avg_loss_disc_real_2: 0.19840839505195618 (+0.01530444622039795)
508
+ | > avg_loss_disc_real_3: 0.21382148563861847 (-0.03698466718196869)
509
+ | > avg_loss_disc_real_4: 0.24576376378536224 (-0.07523001730442047)
510
+ | > avg_loss_disc_real_5: 0.21759817004203796 (-0.029371440410614014)
511
+ | > avg_loss_0: 2.2198214530944824 (-0.28489160537719727)
512
+ | > avg_loss_gen: 2.7460777759552 (+0.2137150764465332)
513
+ | > avg_loss_kl: 1.839993953704834 (-0.5243062973022461)
514
+ | > avg_loss_feat: 7.987682819366455 (+1.9614553451538086)
515
+ | > avg_loss_mel: 19.310388565063477 (-0.14885330200195312)
516
+ | > avg_loss_duration: 1.5255579948425293 (+0.005610823631286621)
517
+ | > avg_loss_1: 33.40970230102539 (+1.5076236724853516)
518
+
519
+
520
+  > EPOCH: 19/35
521
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
522
+
523
+  > TRAINING (2024-03-31 10:13:09) 
524
+
525
+  --> TIME: 2024-03-31 10:13:13 -- STEP: 5/26 -- GLOBAL_STEP: 826500
526
+ | > loss_disc: 2.4146957397460938 (2.450313854217529)
527
+ | > loss_disc_real_0: 0.14674012362957 (0.159730064868927)
528
+ | > loss_disc_real_1: 0.1975105255842209 (0.21452410221099855)
529
+ | > loss_disc_real_2: 0.1922743171453476 (0.22672945261001587)
530
+ | > loss_disc_real_3: 0.1833949238061905 (0.20404482185840606)
531
+ | > loss_disc_real_4: 0.15661051869392395 (0.20521345734596252)
532
+ | > loss_disc_real_5: 0.1906307488679886 (0.2050049215555191)
533
+ | > loss_0: 2.4146957397460938 (2.450313854217529)
534
+ | > grad_norm_0: tensor(16.6246, device='cuda:0') (tensor(13.1455, device='cuda:0'))
535
+ | > loss_gen: 2.416147470474243 (2.4157761573791503)
536
+ | > loss_kl: 2.1608104705810547 (2.2007261753082275)
537
+ | > loss_feat: 8.497369766235352 (8.68954849243164)
538
+ | > loss_mel: 18.566688537597656 (18.833867263793945)
539
+ | > loss_duration: 1.5417736768722534 (1.5525463581085206)
540
+ | > amp_scaler: 256.0 (256.0)
541
+ | > loss_1: 33.18278884887695 (33.692465209960936)
542
+ | > grad_norm_1: tensor(125.5654, device='cuda:0') (tensor(142.0199, device='cuda:0'))
543
+ | > current_lr_0: 0.00019952553399667344
544
+ | > current_lr_1: 0.00019952553399667344
545
+ | > step_time: 0.6896 (0.6795442581176758)
546
+ | > loader_time: 0.0076 (0.0070612430572509766)
547
+
548
+
549
+  > EVALUATION 
550
+
551
+
552
+ --> EVAL PERFORMANCE
553
+ | > avg_loader_time: 0.0018498897552490234 (-0.0001456737518310547)
554
+ | > avg_loss_disc: 2.860830783843994 (+0.6410093307495117)
555
+ | > avg_loss_disc_real_0: 0.07924967259168625 (-0.0584850087761879)
556
+ | > avg_loss_disc_real_1: 0.16997618973255157 (+0.001691862940788269)
557
+ | > avg_loss_disc_real_2: 0.19842827320098877 (+1.9878149032592773e-05)
558
+ | > avg_loss_disc_real_3: 0.2056673765182495 (-0.008154109120368958)
559
+ | > avg_loss_disc_real_4: 0.25505146384239197 (+0.009287700057029724)
560
+ | > avg_loss_disc_real_5: 0.3589268624782562 (+0.14132869243621826)
561
+ | > avg_loss_0: 2.860830783843994 (+0.6410093307495117)
562
+ | > avg_loss_gen: 1.702101707458496 (-1.043976068496704)
563
+ | > avg_loss_kl: 1.9191093444824219 (+0.07911539077758789)
564
+ | > avg_loss_feat: 5.535099029541016 (-2.4525837898254395)
565
+ | > avg_loss_mel: 17.000019073486328 (-2.3103694915771484)
566
+ | > avg_loss_duration: 1.5454752445220947 (+0.01991724967956543)
567
+ | > avg_loss_1: 27.701805114746094 (-5.707897186279297)
568
+
569
+
570
+  > EPOCH: 20/35
571
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
572
+
573
+  > TRAINING (2024-03-31 10:13:34) 
574
+
575
+  > EVALUATION 
576
+
577
+
578
+ --> EVAL PERFORMANCE
579
+ | > avg_loader_time: 0.0021533966064453125 (+0.00030350685119628906)
580
+ | > avg_loss_disc: 2.2710814476013184 (-0.5897493362426758)
581
+ | > avg_loss_disc_real_0: 0.09960463643074036 (+0.020354963839054108)
582
+ | > avg_loss_disc_real_1: 0.19256524741649628 (+0.022589057683944702)
583
+ | > avg_loss_disc_real_2: 0.20135581493377686 (+0.002927541732788086)
584
+ | > avg_loss_disc_real_3: 0.18886281549930573 (-0.016804561018943787)
585
+ | > avg_loss_disc_real_4: 0.23164379596710205 (-0.023407667875289917)
586
+ | > avg_loss_disc_real_5: 0.2516683042049408 (-0.10725855827331543)
587
+ | > avg_loss_0: 2.2710814476013184 (-0.5897493362426758)
588
+ | > avg_loss_gen: 2.2265045642852783 (+0.5244028568267822)
589
+ | > avg_loss_kl: 2.0785317420959473 (+0.1594223976135254)
590
+ | > avg_loss_feat: 7.550324440002441 (+2.015225410461426)
591
+ | > avg_loss_mel: 17.956676483154297 (+0.9566574096679688)
592
+ | > avg_loss_duration: 1.5389719009399414 (-0.00650334358215332)
593
+ | > avg_loss_1: 31.351009368896484 (+3.6492042541503906)
594
+
595
+
596
+  > EPOCH: 21/35
597
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
598
+
599
+  > TRAINING (2024-03-31 10:13:56) 
600
+
601
+  > EVALUATION 
602
+
603
+
604
+ --> EVAL PERFORMANCE
605
+ | > avg_loader_time: 0.0020716190338134766 (-8.177757263183594e-05)
606
+ | > avg_loss_disc: 2.556429147720337 (+0.28534770011901855)
607
+ | > avg_loss_disc_real_0: 0.12546581029891968 (+0.02586117386817932)
608
+ | > avg_loss_disc_real_1: 0.2367255538702011 (+0.044160306453704834)
609
+ | > avg_loss_disc_real_2: 0.17709685862064362 (-0.02425895631313324)
610
+ | > avg_loss_disc_real_3: 0.2612573504447937 (+0.07239453494548798)
611
+ | > avg_loss_disc_real_4: 0.2858666181564331 (+0.054222822189331055)
612
+ | > avg_loss_disc_real_5: 0.16043631732463837 (-0.09123198688030243)
613
+ | > avg_loss_0: 2.556429147720337 (+0.28534770011901855)
614
+ | > avg_loss_gen: 1.996422290802002 (-0.23008227348327637)
615
+ | > avg_loss_kl: 2.1214284896850586 (+0.04289674758911133)
616
+ | > avg_loss_feat: 10.281383514404297 (+2.7310590744018555)
617
+ | > avg_loss_mel: 19.507619857788086 (+1.550943374633789)
618
+ | > avg_loss_duration: 1.5056617259979248 (-0.0333101749420166)
619
+ | > avg_loss_1: 35.412513732910156 (+4.061504364013672)
620
+
621
+
622
+  > EPOCH: 22/35
623
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
624
+
625
+  > TRAINING (2024-03-31 10:14:22) 
626
+
627
+  > EVALUATION 
628
+
629
+
630
+ --> EVAL PERFORMANCE
631
+ | > avg_loader_time: 0.0019686222076416016 (-0.000102996826171875)
632
+ | > avg_loss_disc: 2.432293176651001 (-0.12413597106933594)
633
+ | > avg_loss_disc_real_0: 0.1689225435256958 (+0.04345673322677612)
634
+ | > avg_loss_disc_real_1: 0.20391099154949188 (-0.03281456232070923)
635
+ | > avg_loss_disc_real_2: 0.21518933773040771 (+0.0380924791097641)
636
+ | > avg_loss_disc_real_3: 0.1842881739139557 (-0.07696917653083801)
637
+ | > avg_loss_disc_real_4: 0.27542081475257874 (-0.01044580340385437)
638
+ | > avg_loss_disc_real_5: 0.16442622244358063 (+0.003989905118942261)
639
+ | > avg_loss_0: 2.432293176651001 (-0.12413597106933594)
640
+ | > avg_loss_gen: 2.2132091522216797 (+0.21678686141967773)
641
+ | > avg_loss_kl: 1.8254882097244263 (-0.2959402799606323)
642
+ | > avg_loss_feat: 6.03596830368042 (-4.245415210723877)
643
+ | > avg_loss_mel: 19.248849868774414 (-0.2587699890136719)
644
+ | > avg_loss_duration: 1.5278148651123047 (+0.022153139114379883)
645
+ | > avg_loss_1: 30.851329803466797 (-4.561183929443359)
646
+
647
+
648
+  > EPOCH: 23/35
649
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
650
+
651
+  > TRAINING (2024-03-31 10:14:46) 
652
+
653
+  > EVALUATION 
654
+
655
+
656
+ --> EVAL PERFORMANCE
657
+ | > avg_loader_time: 0.0018432140350341797 (-0.00012540817260742188)
658
+ | > avg_loss_disc: 2.521531105041504 (+0.08923792839050293)
659
+ | > avg_loss_disc_real_0: 0.24499277770519257 (+0.07607023417949677)
660
+ | > avg_loss_disc_real_1: 0.26089686155319214 (+0.056985870003700256)
661
+ | > avg_loss_disc_real_2: 0.17814958095550537 (-0.037039756774902344)
662
+ | > avg_loss_disc_real_3: 0.22235319018363953 (+0.03806501626968384)
663
+ | > avg_loss_disc_real_4: 0.2719460129737854 (-0.003474801778793335)
664
+ | > avg_loss_disc_real_5: 0.2647368311882019 (+0.10031060874462128)
665
+ | > avg_loss_0: 2.521531105041504 (+0.08923792839050293)
666
+ | > avg_loss_gen: 2.6347410678863525 (+0.42153191566467285)
667
+ | > avg_loss_kl: 2.1445794105529785 (+0.31909120082855225)
668
+ | > avg_loss_feat: 9.143684387207031 (+3.1077160835266113)
669
+ | > avg_loss_mel: 18.559925079345703 (-0.6889247894287109)
670
+ | > avg_loss_duration: 1.5272908210754395 (-0.0005240440368652344)
671
+ | > avg_loss_1: 34.01021957397461 (+3.1588897705078125)
672
+
673
+
674
+  > EPOCH: 24/35
675
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
676
+
677
+  > TRAINING (2024-03-31 10:15:10) 
678
+
679
+  > EVALUATION 
680
+
681
+
682
+ --> EVAL PERFORMANCE
683
+ | > avg_loader_time: 0.0021209716796875 (+0.0002777576446533203)
684
+ | > avg_loss_disc: 2.410466194152832 (-0.11106491088867188)
685
+ | > avg_loss_disc_real_0: 0.08595563471317291 (-0.15903714299201965)
686
+ | > avg_loss_disc_real_1: 0.17231160402297974 (-0.0885852575302124)
687
+ | > avg_loss_disc_real_2: 0.2597108781337738 (+0.08156129717826843)
688
+ | > avg_loss_disc_real_3: 0.2549273371696472 (+0.03257414698600769)
689
+ | > avg_loss_disc_real_4: 0.19602034986019135 (-0.07592566311359406)
690
+ | > avg_loss_disc_real_5: 0.26538583636283875 (+0.0006490051746368408)
691
+ | > avg_loss_0: 2.410466194152832 (-0.11106491088867188)
692
+ | > avg_loss_gen: 2.4343385696411133 (-0.20040249824523926)
693
+ | > avg_loss_kl: 2.1930034160614014 (+0.04842400550842285)
694
+ | > avg_loss_feat: 8.565384864807129 (-0.5782995223999023)
695
+ | > avg_loss_mel: 19.013505935668945 (+0.4535808563232422)
696
+ | > avg_loss_duration: 1.5361652374267578 (+0.00887441635131836)
697
+ | > avg_loss_1: 33.742401123046875 (-0.2678184509277344)
698
+
699
+
700
+  > EPOCH: 25/35
701
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
702
+
703
+  > TRAINING (2024-03-31 10:15:38) 
704
+
705
+  > EVALUATION 
706
+
707
+
708
+ --> EVAL PERFORMANCE
709
+ | > avg_loader_time: 0.0021474361419677734 (+2.6464462280273438e-05)
710
+ | > avg_loss_disc: 2.44284725189209 (+0.03238105773925781)
711
+ | > avg_loss_disc_real_0: 0.11237086355686188 (+0.026415228843688965)
712
+ | > avg_loss_disc_real_1: 0.1936785876750946 (+0.021366983652114868)
713
+ | > avg_loss_disc_real_2: 0.17568925023078918 (-0.08402162790298462)
714
+ | > avg_loss_disc_real_3: 0.1501607447862625 (-0.1047665923833847)
715
+ | > avg_loss_disc_real_4: 0.254810631275177 (+0.05879028141498566)
716
+ | > avg_loss_disc_real_5: 0.18697001039981842 (-0.07841582596302032)
717
+ | > avg_loss_0: 2.44284725189209 (+0.03238105773925781)
718
+ | > avg_loss_gen: 1.9306268692016602 (-0.5037117004394531)
719
+ | > avg_loss_kl: 2.0787177085876465 (-0.11428570747375488)
720
+ | > avg_loss_feat: 6.487472057342529 (-2.0779128074645996)
721
+ | > avg_loss_mel: 17.353788375854492 (-1.6597175598144531)
722
+ | > avg_loss_duration: 1.5673844814300537 (+0.0312192440032959)
723
+ | > avg_loss_1: 29.41798973083496 (-4.324411392211914)
724
+
725
+
726
+  > EPOCH: 26/35
727
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
728
+
729
+  > TRAINING (2024-03-31 10:16:03) 
730
+
731
+  > EVALUATION 
732
+
733
+
734
+ --> EVAL PERFORMANCE
735
+ | > avg_loader_time: 0.0018668174743652344 (-0.00028061866760253906)
736
+ | > avg_loss_disc: 2.4484174251556396 (+0.005570173263549805)
737
+ | > avg_loss_disc_real_0: 0.2349560707807541 (+0.12258520722389221)
738
+ | > avg_loss_disc_real_1: 0.19571739435195923 (+0.002038806676864624)
739
+ | > avg_loss_disc_real_2: 0.1958414912223816 (+0.020152240991592407)
740
+ | > avg_loss_disc_real_3: 0.24930566549301147 (+0.09914492070674896)
741
+ | > avg_loss_disc_real_4: 0.24968351423740387 (-0.005127117037773132)
742
+ | > avg_loss_disc_real_5: 0.2158995419740677 (+0.028929531574249268)
743
+ | > avg_loss_0: 2.4484174251556396 (+0.005570173263549805)
744
+ | > avg_loss_gen: 2.308260917663574 (+0.37763404846191406)
745
+ | > avg_loss_kl: 2.1599037647247314 (+0.08118605613708496)
746
+ | > avg_loss_feat: 6.9100341796875 (+0.4225621223449707)
747
+ | > avg_loss_mel: 16.561206817626953 (-0.7925815582275391)
748
+ | > avg_loss_duration: 1.5258960723876953 (-0.0414884090423584)
749
+ | > avg_loss_1: 29.465303421020508 (+0.047313690185546875)
750
+
751
+
752
+  > EPOCH: 27/35
753
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
754
+
755
+  > TRAINING (2024-03-31 10:16:31) 
756
+
757
+  > EVALUATION 
758
+
759
+
760
+ --> EVAL PERFORMANCE
761
+ | > avg_loader_time: 0.0028181076049804688 (+0.0009512901306152344)
762
+ | > avg_loss_disc: 2.3155648708343506 (-0.13285255432128906)
763
+ | > avg_loss_disc_real_0: 0.14989317953586578 (-0.0850628912448883)
764
+ | > avg_loss_disc_real_1: 0.23751142621040344 (+0.041794031858444214)
765
+ | > avg_loss_disc_real_2: 0.23824138939380646 (+0.042399898171424866)
766
+ | > avg_loss_disc_real_3: 0.25855839252471924 (+0.009252727031707764)
767
+ | > avg_loss_disc_real_4: 0.23557884991168976 (-0.014104664325714111)
768
+ | > avg_loss_disc_real_5: 0.18954725563526154 (-0.026352286338806152)
769
+ | > avg_loss_0: 2.3155648708343506 (-0.13285255432128906)
770
+ | > avg_loss_gen: 2.498218059539795 (+0.1899571418762207)
771
+ | > avg_loss_kl: 2.103855609893799 (-0.05604815483093262)
772
+ | > avg_loss_feat: 7.433771133422852 (+0.5237369537353516)
773
+ | > avg_loss_mel: 18.378904342651367 (+1.817697525024414)
774
+ | > avg_loss_duration: 1.541215419769287 (+0.015319347381591797)
775
+ | > avg_loss_1: 31.955965042114258 (+2.49066162109375)
776
+
777
+
778
+  > EPOCH: 28/35
779
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
780
+
781
+  > TRAINING (2024-03-31 10:16:59) 
782
+
783
+  > EVALUATION 
784
+
785
+
786
+ --> EVAL PERFORMANCE
787
+ | > avg_loader_time: 0.0021948814392089844 (-0.0006232261657714844)
788
+ | > avg_loss_disc: 2.719372272491455 (+0.4038074016571045)
789
+ | > avg_loss_disc_real_0: 0.211879700422287 (+0.061986520886421204)
790
+ | > avg_loss_disc_real_1: 0.27071627974510193 (+0.033204853534698486)
791
+ | > avg_loss_disc_real_2: 0.28000608086586 (+0.04176469147205353)
792
+ | > avg_loss_disc_real_3: 0.3032557964324951 (+0.04469740390777588)
793
+ | > avg_loss_disc_real_4: 0.2568355202674866 (+0.021256670355796814)
794
+ | > avg_loss_disc_real_5: 0.25513720512390137 (+0.06558994948863983)
795
+ | > avg_loss_0: 2.719372272491455 (+0.4038074016571045)
796
+ | > avg_loss_gen: 2.1684322357177734 (-0.3297858238220215)
797
+ | > avg_loss_kl: 1.8129353523254395 (-0.2909202575683594)
798
+ | > avg_loss_feat: 5.296431541442871 (-2.1373395919799805)
799
+ | > avg_loss_mel: 16.067489624023438 (-2.3114147186279297)
800
+ | > avg_loss_duration: 1.5332417488098145 (-0.007973670959472656)
801
+ | > avg_loss_1: 26.878530502319336 (-5.077434539794922)
802
+
803
+
804
+  > EPOCH: 29/35
805
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
806
+
807
+  > TRAINING (2024-03-31 10:17:29) 
808
+
809
+  > EVALUATION 
810
+
811
+
812
+ --> EVAL PERFORMANCE
813
+ | > avg_loader_time: 0.0018680095672607422 (-0.0003268718719482422)
814
+ | > avg_loss_disc: 2.799192428588867 (+0.07982015609741211)
815
+ | > avg_loss_disc_real_0: 0.18425635993480682 (-0.027623340487480164)
816
+ | > avg_loss_disc_real_1: 0.2576143741607666 (-0.013101905584335327)
817
+ | > avg_loss_disc_real_2: 0.24912326037883759 (-0.0308828204870224)
818
+ | > avg_loss_disc_real_3: 0.26739048957824707 (-0.03586530685424805)
819
+ | > avg_loss_disc_real_4: 0.20493660867214203 (-0.05189891159534454)
820
+ | > avg_loss_disc_real_5: 0.17635811865329742 (-0.07877908647060394)
821
+ | > avg_loss_0: 2.799192428588867 (+0.07982015609741211)
822
+ | > avg_loss_gen: 1.9312785863876343 (-0.23715364933013916)
823
+ | > avg_loss_kl: 1.9877099990844727 (+0.1747746467590332)
824
+ | > avg_loss_feat: 6.73224401473999 (+1.4358124732971191)
825
+ | > avg_loss_mel: 13.986883163452148 (-2.080606460571289)
826
+ | > avg_loss_duration: 1.5008467435836792 (-0.032395005226135254)
827
+ | > avg_loss_1: 26.138961791992188 (-0.7395687103271484)
828
+
829
+
830
+  > EPOCH: 30/35
831
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
832
+
833
+  > TRAINING (2024-03-31 10:17:56) 
834
+
835
+  > EVALUATION 
836
+
837
+
838
+ --> EVAL PERFORMANCE
839
+ | > avg_loader_time: 0.0018367767333984375 (-3.123283386230469e-05)
840
+ | > avg_loss_disc: 2.38435697555542 (-0.41483545303344727)
841
+ | > avg_loss_disc_real_0: 0.12377185374498367 (-0.06048450618982315)
842
+ | > avg_loss_disc_real_1: 0.18688198924064636 (-0.07073238492012024)
843
+ | > avg_loss_disc_real_2: 0.23647478222846985 (-0.012648478150367737)
844
+ | > avg_loss_disc_real_3: 0.18817488849163055 (-0.07921560108661652)
845
+ | > avg_loss_disc_real_4: 0.17610274255275726 (-0.028833866119384766)
846
+ | > avg_loss_disc_real_5: 0.09066959470510483 (-0.0856885239481926)
847
+ | > avg_loss_0: 2.38435697555542 (-0.41483545303344727)
848
+ | > avg_loss_gen: 2.238271713256836 (+0.30699312686920166)
849
+ | > avg_loss_kl: 1.7533683776855469 (-0.23434162139892578)
850
+ | > avg_loss_feat: 9.994836807250977 (+3.2625927925109863)
851
+ | > avg_loss_mel: 19.69933319091797 (+5.71245002746582)
852
+ | > avg_loss_duration: 1.5312981605529785 (+0.030451416969299316)
853
+ | > avg_loss_1: 35.21710968017578 (+9.078147888183594)
854
+
855
+
856
+  > EPOCH: 31/35
857
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
858
+
859
+  > TRAINING (2024-03-31 10:18:22) 
860
+
861
+  > EVALUATION 
862
+
863
+
864
+ --> EVAL PERFORMANCE
865
+ | > avg_loader_time: 0.0018820762634277344 (+4.5299530029296875e-05)
866
+ | > avg_loss_disc: 2.741990566253662 (+0.3576335906982422)
867
+ | > avg_loss_disc_real_0: 0.1434936225414276 (+0.01972176879644394)
868
+ | > avg_loss_disc_real_1: 0.1855868548154831 (-0.001295134425163269)
869
+ | > avg_loss_disc_real_2: 0.24143913388252258 (+0.004964351654052734)
870
+ | > avg_loss_disc_real_3: 0.20179814100265503 (+0.013623252511024475)
871
+ | > avg_loss_disc_real_4: 0.17437708377838135 (-0.0017256587743759155)
872
+ | > avg_loss_disc_real_5: 0.2018098384141922 (+0.11114024370908737)
873
+ | > avg_loss_0: 2.741990566253662 (+0.3576335906982422)
874
+ | > avg_loss_gen: 1.7299083471298218 (-0.5083633661270142)
875
+ | > avg_loss_kl: 1.9740378856658936 (+0.22066950798034668)
876
+ | > avg_loss_feat: 7.29171895980835 (-2.703117847442627)
877
+ | > avg_loss_mel: 15.08013916015625 (-4.619194030761719)
878
+ | > avg_loss_duration: 1.5062572956085205 (-0.025040864944458008)
879
+ | > avg_loss_1: 27.582061767578125 (-7.635047912597656)
880
+
881
+
882
+  > EPOCH: 32/35
883
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
884
+
885
+  > TRAINING (2024-03-31 10:18:50) 
886
+
887
+  > EVALUATION 
888
+
889
+
890
+ --> EVAL PERFORMANCE
891
+ | > avg_loader_time: 0.002040386199951172 (+0.0001583099365234375)
892
+ | > avg_loss_disc: 2.652589797973633 (-0.0894007682800293)
893
+ | > avg_loss_disc_real_0: 0.13831813633441925 (-0.005175486207008362)
894
+ | > avg_loss_disc_real_1: 0.2390671968460083 (+0.05348034203052521)
895
+ | > avg_loss_disc_real_2: 0.3314564526081085 (+0.09001731872558594)
896
+ | > avg_loss_disc_real_3: 0.3158210217952728 (+0.1140228807926178)
897
+ | > avg_loss_disc_real_4: 0.21650941669940948 (+0.04213233292102814)
898
+ | > avg_loss_disc_real_5: 0.15263138711452484 (-0.04917845129966736)
899
+ | > avg_loss_0: 2.652589797973633 (-0.0894007682800293)
900
+ | > avg_loss_gen: 2.1515212059020996 (+0.42161285877227783)
901
+ | > avg_loss_kl: 1.8530592918395996 (-0.12097859382629395)
902
+ | > avg_loss_feat: 8.823948860168457 (+1.5322299003601074)
903
+ | > avg_loss_mel: 19.046369552612305 (+3.9662303924560547)
904
+ | > avg_loss_duration: 1.550673007965088 (+0.04441571235656738)
905
+ | > avg_loss_1: 33.42557144165039 (+5.843509674072266)
906
+
907
+
908
+  > EPOCH: 33/35
909
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
910
+
911
+  > TRAINING (2024-03-31 10:19:18) 
912
+
913
+  > EVALUATION 
914
+
915
+
916
+ --> EVAL PERFORMANCE
917
+ | > avg_loader_time: 0.0018901824951171875 (-0.00015020370483398438)
918
+ | > avg_loss_disc: 2.5330142974853516 (-0.11957550048828125)
919
+ | > avg_loss_disc_real_0: 0.1612425148487091 (+0.022924378514289856)
920
+ | > avg_loss_disc_real_1: 0.15939201414585114 (-0.07967518270015717)
921
+ | > avg_loss_disc_real_2: 0.27959954738616943 (-0.05185690522193909)
922
+ | > avg_loss_disc_real_3: 0.22299066185951233 (-0.0928303599357605)
923
+ | > avg_loss_disc_real_4: 0.2748161256313324 (+0.05830670893192291)
924
+ | > avg_loss_disc_real_5: 0.20549322664737701 (+0.05286183953285217)
925
+ | > avg_loss_0: 2.5330142974853516 (-0.11957550048828125)
926
+ | > avg_loss_gen: 2.130063533782959 (-0.021457672119140625)
927
+ | > avg_loss_kl: 1.8332613706588745 (-0.019797921180725098)
928
+ | > avg_loss_feat: 6.8745574951171875 (-1.9493913650512695)
929
+ | > avg_loss_mel: 18.154199600219727 (-0.8921699523925781)
930
+ | > avg_loss_duration: 1.5514920949935913 (+0.000819087028503418)
931
+ | > avg_loss_1: 30.543575286865234 (-2.8819961547851562)
932
+
933
+
934
+  > EPOCH: 34/35
935
+ --> /home/azureuser/BanglaTTS/nctb-vits-single-male-9/vits-ft-nctb-March-31-2024_10+05AM-0000000
936
+
937
+  > TRAINING (2024-03-31 10:19:48) 
938
+
939
+  > EVALUATION 
940
+
941
+
942
+ --> EVAL PERFORMANCE
943
+ | > avg_loader_time: 0.0021767616271972656 (+0.0002865791320800781)
944
+ | > avg_loss_disc: 2.4458301067352295 (-0.08718419075012207)
945
+ | > avg_loss_disc_real_0: 0.18880796432495117 (+0.027565449476242065)
946
+ | > avg_loss_disc_real_1: 0.15713262557983398 (-0.002259388566017151)
947
+ | > avg_loss_disc_real_2: 0.2482978254556656 (-0.031301721930503845)
948
+ | > avg_loss_disc_real_3: 0.21974673867225647 (-0.0032439231872558594)
949
+ | > avg_loss_disc_real_4: 0.2329111099243164 (-0.04190501570701599)
950
+ | > avg_loss_disc_real_5: 0.2461576908826828 (+0.040664464235305786)
951
+ | > avg_loss_0: 2.4458301067352295 (-0.08718419075012207)
952
+ | > avg_loss_gen: 2.317957878112793 (+0.18789434432983398)
953
+ | > avg_loss_kl: 2.012809991836548 (+0.17954862117767334)
954
+ | > avg_loss_feat: 6.7331624031066895 (-0.14139509201049805)
955
+ | > avg_loss_mel: 16.866455078125 (-1.2877445220947266)
956
+ | > avg_loss_duration: 1.5175280570983887 (-0.03396403789520264)
957
+ | > avg_loss_1: 29.447914123535156 (-1.0956611633300781)
958
+