使用 rmvpe 推理至 5s 多时显示 Error

系统环境:

❯ fastfetch
                  -`                     lovemilk@***
                 .o+`                    ---------------------
                `ooo/                    OS: Arch Linux x86_64
              `+oooooo:                  Kernel: Linux 6.17.7-zen1-1-zen
              -+oooooo+:                 Uptime: 7 hours, 33 mins
            `/:-:++oooo+:                Packages: 2539 (pacman), 6 (flatpak)
           `/++++/+++++++:               Shell: zsh 5.9
          `/++++++++++++++:              Display (Type_C): 2560x1600 @ 1.25x in 27", 144 Hz [External] *
         `/+++ooooooooooooo/`            Display (FKS-C215J): 1920x1080 in 22", 75 Hz [External]
        ./ooosssso++osssssso+`           DE: KDE Plasma 6.5.1
       .oossssso-````/ossssss+`          WM: KWin (Wayland)
      -osssssso.      :ssssssso.         WM Theme: Breeze
     :osssssss/        osssso+++.        Theme: Breeze (Dark) [Qt], Material-DeepOcean-B-LB [GTK2/3/4]
    /ossssssss/        +ssssooo/-        Icons: breeze-dark [Qt], breeze-dark [GTK2/3/4]
  `/ossssso+/:-        -:/+osssso+-      Font: HarmonyOS Sans (11pt, Medium) [Qt], HarmonyOS Sans Medium (11p]
 `+sso+:-`                 `.-/+oso:     Cursor: Afterglow (24px)
`++:.                           `-/+/    Terminal: konsole 25.8.2
.`                                 `/    Terminal Font: JetBrainsMonoNL Nerd Font Mono (11pt)
                                         CPU: Intel(R) Core(TM) i9-10900 (20) @ 5.20 GHz
                                         GPU 1: NVIDIA GeForce RTX 3070 [Discrete]
                                         GPU 2: Intel UHD Graphics 630 @ 1.20 GHz [Integrated]
                                         Memory: 21.76 GiB / 31.08 GiB (70%)
                                         Swap: 9.18 GiB / 15.54 GiB (59%)
                                         Disk (/): 355.48 GiB / 450.00 GiB (79%) - btrfs
                                         Disk (/mnt/btrfs/512ssd): 437.66 GiB / 476.94 GiB (92%) - btrfs
                                         Disk (/mnt/btrfs/data): 810.94 GiB / 1024.00 GiB (79%) - btrfs
                                         Disk (/mnt/btrfs/data0): 33.09 GiB / 839.02 GiB (4%) - btrfs
                                         Disk (/mnt/ntfs/nvme500gib): 402.76 GiB / 500.01 GiB (81%) - ntfs3
                                         Local IP (enp0s31f6): 192.168.1.100/16
                                         Locale: zh_CN.UTF-8

Python 环境:

❯ python --version
Python 3.8.20

日志输出:

> @ start /home/lovemilk/文档/projects/RVC/Retrieval-based-Voice-Conversion-WebUI
> source .venv/bin/activate && python infer-web.py --pycmd python3.8 --port 9991

2025-11-14 22:30:39 | INFO | configs.config | Found GPU NVIDIA GeForce RTX 3070
2025-11-14 22:30:39 | INFO | configs.config | Half-precision floating-point: True, device: cuda:0
/home/lovemilk/文档/projects/RVC/Retrieval-based-Voice-Conversion-WebUI/.venv/lib/python3.8/site-packages/gradio_client/documentation.py:106: UserWarning: Could not get documentation group for <class 'gradio.mix.Parallel'>: No known documentation group for module 'gradio.mix'
  warnings.warn(f"Could not get documentation group for {cls}: {exc}")
/home/lovemilk/文档/projects/RVC/Retrieval-based-Voice-Conversion-WebUI/.venv/lib/python3.8/site-packages/gradio_client/documentation.py:106: UserWarning: Could not get documentation group for <class 'gradio.mix.Series'>: No known documentation group for module 'gradio.mix'
  warnings.warn(f"Could not get documentation group for {cls}: {exc}")
2025-11-14 22:30:40 | INFO | __main__ | Use Language: zh_CN
Running on local URL:  http://0.0.0.0:9991
2025-11-14 22:30:47 | INFO | infer.modules.vc.modules | Get sid: lovemilk-V20241029.01.pth
2025-11-14 22:30:47 | INFO | infer.modules.vc.modules | Loading: assets/weights/lovemilk-V20241029.01.pth
2025-11-14 22:30:48 | INFO | infer.modules.vc.modules | Select index: 
2025-11-14 22:30:48 | INFO | httpx | HTTP Request: POST http://localhost:9991/api/predict "HTTP/1.1 200 OK"
2025-11-14 22:30:48 | INFO | httpx | HTTP Request: POST http://localhost:9991/reset "HTTP/1.1 200 OK"
2025-11-14 22:30:51 | INFO | fairseq.tasks.hubert_pretraining | current directory is /home/lovemilk/文档/projects/RVC/Retrieval-based-Voice-Conversion-WebUI
2025-11-14 22:30:51 | INFO | fairseq.tasks.hubert_pretraining | HubertPretrainingTask Config {'_name': 'hubert_pretraining', 'data': 'metadata', 'fine_tuning': False, 'labels': ['km'], 'label_dir': 'label', 'label_rate': 50.0, 'sample_rate': 16000, 'normalize': False, 'enable_padding': False, 'max_keep_size': None, 'max_sample_size': 250000, 'min_sample_size': 32000, 'single_target': False, 'random_crop': True, 'pad_audio': False}
2025-11-14 22:30:51 | INFO | fairseq.models.hubert.hubert | HubertModel Config: {'_name': 'hubert', 'label_rate': 50.0, 'extractor_mode': default, 'encoder_layers': 12, 'encoder_embed_dim': 768, 'encoder_ffn_embed_dim': 3072, 'encoder_attention_heads': 12, 'activation_fn': gelu, 'layer_type': transformer, 'dropout': 0.1, 'attention_dropout': 0.1, 'activation_dropout': 0.0, 'encoder_layerdrop': 0.05, 'dropout_input': 0.1, 'dropout_features': 0.1, 'final_dim': 256, 'untie_final_proj': True, 'layer_norm_first': False, 'conv_feature_layers': '[(512,10,5)] + [(512,3,2)] * 4 + [(512,2,2)] * 2', 'conv_bias': False, 'logit_temp': 0.1, 'target_glu': False, 'feature_grad_mult': 0.1, 'mask_length': 10, 'mask_prob': 0.8, 'mask_selection': static, 'mask_other': 0.0, 'no_mask_overlap': False, 'mask_min_space': 1, 'mask_channel_length': 10, 'mask_channel_prob': 0.0, 'mask_channel_selection': static, 'mask_channel_other': 0.0, 'no_mask_channel_overlap': False, 'mask_channel_min_space': 1, 'conv_pos': 128, 'conv_pos_groups': 16, 'latent_temp': [2.0, 0.5, 0.999995], 'skip_masked': False, 'skip_nomask': False, 'checkpoint_activations': False, 'required_seq_len_multiple': 2, 'depthwise_conv_kernel_size': 31, 'attn_type': '', 'pos_enc_type': 'abs', 'fp16': False}
2025-11-14 22:30:55 | INFO | infer.modules.vc.pipeline | Loading rmvpe model,assets/rmvpe/rmvpe.pt
2025-11-14 22:30:56 | INFO | httpx | HTTP Request: POST http://localhost:9991/reset "HTTP/1.1 200 OK"

join socks 输出:

{"msg": "process_completed", "output": {"error": ""}, "success": false}	

BTW, GPT-SoVITS 可以正常运行

Reproduce the failure from infer-web.py with the reported Python 3.8 environment and an RMVPE inference longer than five seconds. Start by tracing the RMVPE loading and inference path in infer/modules/vc/pipeline.py and inspect the process_completed output; done means the inference completes or exposes a specific actionable error.