GPT-SoVITS模型|男声2口语声音模型|男声磁性配音模型

声音特点适用领域：磁性温和，峥气十足，角色，娱乐，影视，资讯，百科

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。

参考音频样音 peiyin_140

模型下载

此处内容已隐藏，请付费后查看

训练日志

2025-07-11 08:06:30,536	男声2 口语	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/gsv-v2final-pretrained/s2G2333k.pth', 'pretrained_s2D': 'GPT_SoVITS/pretrained_models/gsv-v2final-pretrained/s2D2333k.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/男声2 口语'}, '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, 'version': 'v2'}, 's2_ckpt_dir': 'logs/男声2 口语', 'content_module': 'cnhubert', 'save_weight_dir': 'SoVITS_weights_v2', 'name': '男声2 口语', 'version': 'v2', 'pretrain': None, 'resume_step': None}
2025-07-11 08:06:32,436	男声2 口语	INFO	loaded pretrained GPT_SoVITS/pretrained_models/gsv-v2final-pretrained/s2G2333k.pth
2025-07-11 08:06:32,765	男声2 口语	INFO	loaded pretrained GPT_SoVITS/pretrained_models/gsv-v2final-pretrained/s2D2333k.pth
2025-07-11 08:07:11,664	男声2 口语	INFO	Train Epoch: 1 [0%]
2025-07-11 08:07:11,664	男声2 口语	INFO	[2.9515578746795654, 1.5290586948394775, 5.988468170166016, 23.8193416595459, 0.0, 1.6040661334991455, 0, 9.99875e-05]
2025-07-11 08:07:40,666	男声2 口语	INFO	====> Epoch: 1
2025-07-11 08:08:09,109	男声2 口语	INFO	====> Epoch: 2
2025-07-11 08:08:36,961	男声2 口语	INFO	====> Epoch: 3
2025-07-11 08:09:04,235	男声2 口语	INFO	====> Epoch: 4
2025-07-11 08:09:21,278	男声2 口语	INFO	Train Epoch: 5 [55%]
2025-07-11 08:09:21,280	男声2 口语	INFO	[2.6409244537353516, 2.6877622604370117, 8.784132957458496, 23.79684066772461, 0.0, 1.8762003183364868, 100, 9.993751562304699e-05]
2025-07-11 08:09:32,172	男声2 口语	INFO	====> Epoch: 5
2025-07-11 08:10:00,084	男声2 口语	INFO	====> Epoch: 6
2025-07-11 08:10:27,641	男声2 口语	INFO	====> Epoch: 7
2025-07-11 08:10:55,101	男声2 口语	INFO	====> Epoch: 8
2025-07-11 08:11:24,517	男声2 口语	INFO	====> Epoch: 9
2025-07-11 08:11:28,490	男声2 口语	INFO	Train Epoch: 10 [9%]
2025-07-11 08:11:28,491	男声2 口语	INFO	[2.5608654022216797, 2.5396413803100586, 6.612776756286621, 21.601232528686523, 0.0, 1.5293047428131104, 200, 9.987507028906759e-05]
2025-07-11 08:11:40,623	男声2 口语	INFO	====> Epoch: 10
2025-07-11 08:11:55,657	男声2 口语	INFO	====> Epoch: 11
2025-07-11 08:12:10,863	男声2 口语	INFO	====> Epoch: 12
2025-07-11 08:12:25,856	男声2 口语	INFO	====> Epoch: 13
2025-07-11 08:12:36,763	男声2 口语	INFO	Train Epoch: 14 [64%]
2025-07-11 08:12:36,763	男声2 口语	INFO	[2.621643543243408, 1.939083218574524, 6.820486068725586, 20.677940368652344, 0.0, 1.6165881156921387, 300, 9.982514211643064e-05]
2025-07-11 08:12:41,390	男声2 口语	INFO	====> Epoch: 14
2025-07-11 08:12:56,394	男声2 口语	INFO	====> Epoch: 15
2025-07-11 08:13:11,446	男声2 口语	INFO	====> Epoch: 16
2025-07-11 08:13:26,768	男声2 口语	INFO	====> Epoch: 17
2025-07-11 08:13:41,824	男声2 口语	INFO	====> Epoch: 18
2025-07-11 08:13:46,749	男声2 口语	INFO	Train Epoch: 19 [18%]
2025-07-11 08:13:46,750	男声2 口语	INFO	[2.737532138824463, 2.1285741329193115, 7.10671329498291, 21.86750030517578, 0.0, 1.5401709079742432, 400, 9.976276699833672e-05]
2025-07-11 08:13:57,342	男声2 口语	INFO	====> Epoch: 19
2025-07-11 08:14:12,552	男声2 口语	INFO	Saving model and optimizer state at iteration 20 to logs/男声2 口语/logs_s2\G_233333333333.pth
2025-07-11 08:14:13,612	男声2 口语	INFO	Saving model and optimizer state at iteration 20 to logs/男声2 口语/logs_s2\D_233333333333.pth
2025-07-11 08:14:14,969	男声2 口语	INFO	saving ckpt 男声2 口语_e20:Success.
2025-07-11 08:14:14,969	男声2 口语	INFO	====> Epoch: 20