tuandaodev commited on
Commit
fc5ca60
·
verified ·
1 Parent(s): 6555f96

Init model

Browse files
best_model_16928.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aceb6c537ec1f17c9391197e080faef855c5c7dc4246cba371e5cd587cd7a2d5
3
+ size 5652603561
config.json ADDED
@@ -0,0 +1,193 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_path": "/kaggle/temp/checkpoints/",
3
+ "logger_uri": null,
4
+ "run_name": "GPT_XTTS_FT",
5
+ "project_name": "XTTS_trainer",
6
+ "run_description": [
7
+ "\n GPT XTTS training\n "
8
+ ],
9
+ "print_step": 50,
10
+ "plot_step": 100,
11
+ "model_param_stats": false,
12
+ "wandb_entity": null,
13
+ "dashboard_logger": "tensorboard",
14
+ "save_on_interrupt": true,
15
+ "log_model_step": 100,
16
+ "save_step": 1000,
17
+ "save_n_checkpoints": 1,
18
+ "save_checkpoints": true,
19
+ "save_all_best": false,
20
+ "save_best_after": 10000,
21
+ "target_loss": null,
22
+ "print_eval": false,
23
+ "test_delay_epochs": 0,
24
+ "run_eval": true,
25
+ "run_eval_steps": null,
26
+ "distributed_backend": "nccl",
27
+ "distributed_url": "tcp://localhost:54321",
28
+ "mixed_precision": false,
29
+ "precision": "fp16",
30
+ "epochs": 5,
31
+ "batch_size": 3,
32
+ "eval_batch_size": 16,
33
+ "grad_clip": 0.0,
34
+ "scheduler_after_epoch": true,
35
+ "lr": 5e-06,
36
+ "optimizer": "AdamW",
37
+ "optimizer_params": {
38
+ "betas": [
39
+ 0.9,
40
+ 0.96
41
+ ],
42
+ "eps": 1e-08,
43
+ "weight_decay": 0.01
44
+ },
45
+ "lr_scheduler": "MultiStepLR",
46
+ "lr_scheduler_params": {
47
+ "milestones": [
48
+ 900000,
49
+ 2700000,
50
+ 5400000
51
+ ],
52
+ "gamma": 0.5,
53
+ "last_epoch": -1
54
+ },
55
+ "use_grad_scaler": false,
56
+ "allow_tf32": false,
57
+ "cudnn_enable": true,
58
+ "cudnn_deterministic": false,
59
+ "cudnn_benchmark": false,
60
+ "training_seed": 54321,
61
+ "model": "xtts",
62
+ "num_loader_workers": 8,
63
+ "num_eval_loader_workers": 0,
64
+ "use_noise_augment": false,
65
+ "audio": {
66
+ "sample_rate": 22050,
67
+ "output_sample_rate": 24000,
68
+ "dvae_sample_rate": 22050
69
+ },
70
+ "use_phonemes": false,
71
+ "phonemizer": null,
72
+ "phoneme_language": null,
73
+ "compute_input_seq_cache": false,
74
+ "text_cleaner": null,
75
+ "enable_eos_bos_chars": false,
76
+ "test_sentences_file": "",
77
+ "phoneme_cache_path": null,
78
+ "characters": null,
79
+ "add_blank": false,
80
+ "batch_group_size": 0,
81
+ "loss_masking": null,
82
+ "min_audio_len": 1,
83
+ "max_audio_len": Infinity,
84
+ "min_text_len": 1,
85
+ "max_text_len": Infinity,
86
+ "compute_f0": false,
87
+ "compute_energy": false,
88
+ "compute_linear_spec": false,
89
+ "precompute_num_workers": 0,
90
+ "start_by_longest": false,
91
+ "shuffle": false,
92
+ "drop_last": false,
93
+ "datasets": [
94
+ {
95
+ "formatter": "",
96
+ "dataset_name": "",
97
+ "path": "",
98
+ "meta_file_train": "",
99
+ "ignored_speakers": null,
100
+ "language": "",
101
+ "phonemizer": "",
102
+ "meta_file_val": "",
103
+ "meta_file_attn_mask": ""
104
+ }
105
+ ],
106
+ "test_sentences": [],
107
+ "eval_split_max_size": 256,
108
+ "eval_split_size": 0.01,
109
+ "use_speaker_weighted_sampler": false,
110
+ "speaker_weighted_sampler_alpha": 1.0,
111
+ "use_language_weighted_sampler": false,
112
+ "language_weighted_sampler_alpha": 1.0,
113
+ "use_length_weighted_sampler": false,
114
+ "length_weighted_sampler_alpha": 1.0,
115
+ "model_args": {
116
+ "gpt_batch_size": 1,
117
+ "enable_redaction": false,
118
+ "kv_cache": true,
119
+ "gpt_checkpoint": "",
120
+ "clvp_checkpoint": null,
121
+ "decoder_checkpoint": null,
122
+ "num_chars": 255,
123
+ "tokenizer_file": "/kaggle/temp/checkpoints/XTTS_v2.0_original_model_files/vocab.json",
124
+ "gpt_max_audio_tokens": 605,
125
+ "gpt_max_text_tokens": 402,
126
+ "gpt_max_prompt_tokens": 70,
127
+ "gpt_layers": 30,
128
+ "gpt_n_model_channels": 1024,
129
+ "gpt_n_heads": 16,
130
+ "gpt_number_text_tokens": 8498,
131
+ "gpt_start_text_token": 261,
132
+ "gpt_stop_text_token": 0,
133
+ "gpt_num_audio_tokens": 1026,
134
+ "gpt_start_audio_token": 1024,
135
+ "gpt_stop_audio_token": 1025,
136
+ "gpt_code_stride_len": 1024,
137
+ "gpt_use_masking_gt_prompt_approach": true,
138
+ "gpt_use_perceiver_resampler": true,
139
+ "input_sample_rate": 22050,
140
+ "output_sample_rate": 24000,
141
+ "output_hop_length": 256,
142
+ "decoder_input_dim": 1024,
143
+ "d_vector_dim": 512,
144
+ "cond_d_vector_in_each_upsampling_layer": true,
145
+ "duration_const": 102400,
146
+ "min_conditioning_length": 11025,
147
+ "max_conditioning_length": 132300,
148
+ "gpt_loss_text_ce_weight": 0.01,
149
+ "gpt_loss_mel_ce_weight": 1.0,
150
+ "debug_loading_failures": false,
151
+ "max_wav_length": 330750,
152
+ "max_text_length": 400,
153
+ "mel_norm_file": "/kaggle/temp/checkpoints/XTTS_v2.0_original_model_files/mel_stats.pth",
154
+ "dvae_checkpoint": "/kaggle/temp/checkpoints/XTTS_v2.0_original_model_files/dvae.pth",
155
+ "xtts_checkpoint": "/kaggle/temp/checkpoints/XTTS_v2.0_original_model_files/model.pth",
156
+ "vocoder": ""
157
+ },
158
+ "model_dir": null,
159
+ "languages": [
160
+ "en",
161
+ "es",
162
+ "fr",
163
+ "de",
164
+ "it",
165
+ "pt",
166
+ "pl",
167
+ "tr",
168
+ "ru",
169
+ "nl",
170
+ "cs",
171
+ "ar",
172
+ "zh-cn",
173
+ "hu",
174
+ "ko",
175
+ "ja",
176
+ "hi",
177
+ "vi"
178
+ ],
179
+ "temperature": 0.75,
180
+ "length_penalty": 1.0,
181
+ "repetition_penalty": 5.0,
182
+ "top_k": 50,
183
+ "top_p": 0.85,
184
+ "num_gpt_outputs": 1,
185
+ "gpt_cond_len": 30,
186
+ "gpt_cond_chunk_len": 4,
187
+ "max_ref_len": 30,
188
+ "sound_norm_refs": false,
189
+ "optimizer_wd_only_on_weights": true,
190
+ "weighted_loss_attrs": null,
191
+ "weighted_loss_multipliers": null,
192
+ "github_branch": "* main"
193
+ }
dvae.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5fda6acc73c5077c914d60618b096eff0d547c11b4c8b7ceab7905fb236facce
3
+ size 210893114
events.out.tfevents.1729445198.feef3d2cfaf9.453.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96d5e1472ab60158f7c4e1499c06729125c4af696647e860754e16e8fb4ef68b
3
+ size 87759
trainer_0_log.txt ADDED
The diff for this file is too large to render. See raw diff
 
vocab.json ADDED
The diff for this file is too large to render. See raw diff