GPT-SoVITS模型|魔婉宁声音模型魔婉宁配音模型

擅长领域：朗诵、情感、影视、温柔细腻，自然动听

GPT-SoVITS模型配音效果

鉴于GPT-SoVITS模型自回归特性，即其配音情绪高度依赖于所提供的参考音频，特此说明：本视频所展示的配音情绪仅为采用某一特定参考音频时的效果示例，并不全面反映GPT-SoVITS模型能够生成的全部情绪范围及最终配音质量的上限。模型的最终表现将随着不同参考音频的输入而展现出多样化。

训练日志

2025-07-31 13:57:51,911	魔婉宁	INFO	{'train': {'log_interval': 100, 'eval_interval': 500, 'seed': 1234, 'epochs': 20, 'learning_rate': 0.0001, 'betas': [0.8, 0.99], 'eps': 1e-09, 'batch_size': 11, 'fp16_run': True, 'lr_decay': 0.999875, 'segment_size': 20480, 'init_lr_ratio': 1, 'warmup_epochs': 0, 'c_mel': 45, 'c_kl': 1.0, 'text_low_lr_rate': 0.4, 'pretrained_s2G': 'GPT_SoVITS/pretrained_models/s2G488k.pth', 'pretrained_s2D': 'GPT_SoVITS/pretrained_models/s2D488k.pth', 'if_save_latest': True, 'if_save_every_weights': True, 'save_every_epoch': 20, 'gpu_numbers': '0'}, 'data': {'max_wav_value': 32768.0, 'sampling_rate': 32000, 'filter_length': 2048, 'hop_length': 640, 'win_length': 2048, 'n_mel_channels': 128, 'mel_fmin': 0.0, 'mel_fmax': None, 'add_blank': True, 'n_speakers': 300, 'cleaned_text': True, 'exp_dir': 'logs/魔婉宁'}, 'model': {'inter_channels': 192, 'hidden_channels': 192, 'filter_channels': 768, 'n_heads': 2, 'n_layers': 6, 'kernel_size': 3, 'p_dropout': 0.1, 'resblock': '1', 'resblock_kernel_sizes': [3, 7, 11], 'resblock_dilation_sizes': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'upsample_rates': [10, 8, 2, 2, 2], 'upsample_initial_channel': 512, 'upsample_kernel_sizes': [16, 16, 8, 2, 2], 'n_layers_q': 3, 'use_spectral_norm': False, 'gin_channels': 512, 'semantic_frame_rate': '25hz', 'freeze_quantizer': True}, 's2_ckpt_dir': 'logs/魔婉宁', 'content_module': 'cnhubert', 'save_weight_dir': 'SoVITS_weights', 'name': '魔婉宁', 'pretrain': None, 'resume_step': None}
2025-07-31 13:57:53,340	魔婉宁	INFO	loaded pretrained GPT_SoVITS/pretrained_models/s2G488k.pth
2025-07-31 13:57:53,531	魔婉宁	INFO	loaded pretrained GPT_SoVITS/pretrained_models/s2D488k.pth
2025-07-31 13:58:25,828	魔婉宁	INFO	Train Epoch: 1 [0%]
2025-07-31 13:58:25,828	魔婉宁	INFO	[2.435288906097412, 2.62093448638916, 12.226040840148926, 21.775989532470703, 0.0, 2.524244785308838, 0, 9.99875e-05]
2025-07-31 13:58:39,596	魔婉宁	INFO	====> Epoch: 1
2025-07-31 13:58:53,718	魔婉宁	INFO	====> Epoch: 2
2025-07-31 13:59:07,381	魔婉宁	INFO	====> Epoch: 3
2025-07-31 13:59:21,455	魔婉宁	INFO	====> Epoch: 4
2025-07-31 13:59:35,357	魔婉宁	INFO	====> Epoch: 5
2025-07-31 13:59:48,922	魔婉宁	INFO	Train Epoch: 6 [88%]
2025-07-31 13:59:48,922	魔婉宁	INFO	[2.3553202152252197, 2.5672004222869873, 10.036107063293457, 17.688003540039062, 0.0, 1.458526849746704, 100, 9.99250234335941e-05]
2025-07-31 13:59:49,937	魔婉宁	INFO	====> Epoch: 6
2025-07-31 14:00:03,714	魔婉宁	INFO	====> Epoch: 7
2025-07-31 14:00:17,402	魔婉宁	INFO	====> Epoch: 8
2025-07-31 14:00:30,931	魔婉宁	INFO	====> Epoch: 9
2025-07-31 14:00:45,581	魔婉宁	INFO	====> Epoch: 10
2025-07-31 14:00:59,270	魔婉宁	INFO	====> Epoch: 11
2025-07-31 14:01:10,025	魔婉宁	INFO	Train Epoch: 12 [76%]
2025-07-31 14:01:10,025	魔婉宁	INFO	[2.2855141162872314, 2.675412178039551, 10.600923538208008, 17.342241287231445, 0.0, 1.5843344926834106, 200, 9.98501030820433e-05]
2025-07-31 14:01:12,939	魔婉宁	INFO	====> Epoch: 12
2025-07-31 14:01:26,636	魔婉宁	INFO	====> Epoch: 13
2025-07-31 14:01:42,520	魔婉宁	INFO	====> Epoch: 14
2025-07-31 14:01:56,232	魔婉宁	INFO	====> Epoch: 15
2025-07-31 14:02:09,877	魔婉宁	INFO	====> Epoch: 16
2025-07-31 14:02:23,881	魔婉宁	INFO	====> Epoch: 17
2025-07-31 14:02:34,280	魔婉宁	INFO	Train Epoch: 18 [65%]
2025-07-31 14:02:34,281	魔婉宁	INFO	[2.4412951469421387, 2.585385322570801, 11.181450843811035, 17.255224227905273, 0.0, 1.1558743715286255, 300, 9.977523890319963e-05]
2025-07-31 14:02:38,426	魔婉宁	INFO	====> Epoch: 18
2025-07-31 14:02:51,926	魔婉宁	INFO	====> Epoch: 19
2025-07-31 14:03:06,158	魔婉宁	INFO	Saving model and optimizer state at iteration 20 to logs/魔婉宁/logs_s2\G_233333333333.pth
2025-07-31 14:03:06,935	魔婉宁	INFO	Saving model and optimizer state at iteration 20 to logs/魔婉宁/logs_s2\D_233333333333.pth
2025-07-31 14:03:08,631	魔婉宁	INFO	saving ckpt 魔婉宁_e20:Success.
2025-07-31 14:03:08,632	魔婉宁	INFO	====> Epoch: 20