datong-new/rvc – Run with an API on Replicate

datong-new / rvc

Cold

Public
195 runs
T4

Run with an API

Playground API Examples README Versions

Input

audio_for_train

file

audio file for train

audio_for_infer

file

Preview

Video Player is loading.

Current Time 00:00:000

Duration 00:00:000

Loaded: 0%

Stream Type LIVE

Remaining Time 00:00:000

audio file for cloning

ckpt

file

your trained checkpoint

f0_up_key

integer

f0_up_key

Default: 0

accompaniment

boolean

if output have accompaniment (for music clone)

Default: true

operation

*string

1. train (audio_for_train is needed); 2 infer (audio_for_infer and ckpt is needed); 3. train_infer (audio_for_train and audio_for_infer is needed)

Run this model in Node.js with one line of code:

npx create-replicate --model=datong-new/rvc

or set up a project from scratch

Install Replicate’s Node.js client library:

npm install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import and set up the client:

import Replicate from "replicate";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,
});

Run datong-new/rvc using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "datong-new/rvc:5da9f66869beacc8f2484215e25c88053acfe24044d64d4b26bbc40f7b5428dc",
  {
    input: {
      f0_up_key: 8,
      operation: "train_infer",
      accompaniment: true,
      audio_for_infer: "https://replicate.delivery/pbxt/KSuE9SkdHEsXPUlZmWj4kMwRNZwJ0CR0EFTzjJAwVqKY8brY/1.wav",
      audio_for_train: "https://replicate.delivery/pbxt/KSuE9M9iVBPPXjGWfkGpiwD9iZOlHSwAwVmX0vHaA2hJ41Ca/wobunanguo.flac"
    }
  }
);

console.log(output);

To learn more, take a look at the guide on getting started with Node.js.

Install Replicate’s Python client library:

pip install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import the client:

import replicate

Run datong-new/rvc using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "datong-new/rvc:5da9f66869beacc8f2484215e25c88053acfe24044d64d4b26bbc40f7b5428dc",
    input={
        "f0_up_key": 8,
        "operation": "train_infer",
        "accompaniment": True,
        "audio_for_infer": "https://replicate.delivery/pbxt/KSuE9SkdHEsXPUlZmWj4kMwRNZwJ0CR0EFTzjJAwVqKY8brY/1.wav",
        "audio_for_train": "https://replicate.delivery/pbxt/KSuE9M9iVBPPXjGWfkGpiwD9iZOlHSwAwVmX0vHaA2hJ41Ca/wobunanguo.flac"
    }
)
print(output)

To learn more, take a look at the guide on getting started with Python.

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Run datong-new/rvc using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

curl -s -X POST \
  -H "Authorization: Bearer $REPLICATE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Prefer: wait" \
  -d $'{
    "version": "datong-new/rvc:5da9f66869beacc8f2484215e25c88053acfe24044d64d4b26bbc40f7b5428dc",
    "input": {
      "f0_up_key": 8,
      "operation": "train_infer",
      "accompaniment": true,
      "audio_for_infer": "https://replicate.delivery/pbxt/KSuE9SkdHEsXPUlZmWj4kMwRNZwJ0CR0EFTzjJAwVqKY8brY/1.wav",
      "audio_for_train": "https://replicate.delivery/pbxt/KSuE9M9iVBPPXjGWfkGpiwD9iZOlHSwAwVmX0vHaA2hJ41Ca/wobunanguo.flac"
    }
  }' \
  https://api.replicate.com/v1/predictions

To learn more, take a look at Replicate’s HTTP API reference docs.

Output

ckpt_path

default.pth

cloned_audio

Video Player is loading.

Current Time 00:00:000

Duration 00:00:000

Loaded: 0%

Stream Type LIVE

Remaining Time 00:00:000

{
  "completed_at": "2024-02-25T12:51:58.322834Z",
  "created_at": "2024-02-25T12:44:50.386033Z",
  "data_removed": false,
  "error": null,
  "id": "kwohmxzbqtrk2ssyxusm2wf4sy",
  "input": {
    "f0_up_key": 8,
    "operation": "train_infer",
    "accompaniment": true,
    "audio_for_infer": "https://replicate.delivery/pbxt/KSuE9SkdHEsXPUlZmWj4kMwRNZwJ0CR0EFTzjJAwVqKY8brY/1.wav",
    "audio_for_train": "https://replicate.delivery/pbxt/KSuE9M9iVBPPXjGWfkGpiwD9iZOlHSwAwVmX0vHaA2hJ41Ca/wobunanguo.flac"
  },
  "logs": "0%|          | 0/116 [00:00<?, ?it/s]\n  1%|          | 1/116 [00:01<02:31,  1.32s/it]\n  3%|▎         | 3/116 [00:01<00:46,  2.44it/s]\n  4%|▍         | 5/116 [00:01<00:26,  4.19it/s]\n  6%|▌         | 7/116 [00:01<00:17,  6.09it/s]\n  8%|▊         | 9/116 [00:01<00:13,  7.95it/s]\n  9%|▉         | 11/116 [00:02<00:10,  9.62it/s]\n 11%|█         | 13/116 [00:02<00:09, 11.13it/s]\n 13%|█▎        | 15/116 [00:02<00:08, 12.23it/s]\n 15%|█▍        | 17/116 [00:02<00:07, 13.11it/s]\n 16%|█▋        | 19/116 [00:02<00:06, 13.90it/s]\n 18%|█▊        | 21/116 [00:02<00:06, 14.54it/s]\n 20%|█▉        | 23/116 [00:02<00:06, 14.98it/s]\n 22%|██▏       | 25/116 [00:02<00:05, 15.21it/s]\n 23%|██▎       | 27/116 [00:03<00:05, 15.31it/s]\n 25%|██▌       | 29/116 [00:03<00:05, 15.47it/s]\n 27%|██▋       | 31/116 [00:03<00:05, 15.77it/s]\n 28%|██▊       | 33/116 [00:03<00:05, 15.90it/s]\n 30%|███       | 35/116 [00:03<00:05, 16.05it/s]\n 32%|███▏      | 37/116 [00:03<00:04, 16.16it/s]\n 34%|███▎      | 39/116 [00:03<00:04, 16.15it/s]\n 35%|███▌      | 41/116 [00:03<00:04, 15.93it/s]\n 37%|███▋      | 43/116 [00:04<00:04, 15.79it/s]\n 39%|███▉      | 45/116 [00:04<00:04, 15.85it/s]\n 41%|████      | 47/116 [00:04<00:04, 15.82it/s]\n 42%|████▏     | 49/116 [00:04<00:04, 15.92it/s]\n 44%|████▍     | 51/116 [00:04<00:04, 15.89it/s]\n 46%|████▌     | 53/116 [00:04<00:03, 15.96it/s]\n 47%|████▋     | 55/116 [00:04<00:03, 15.96it/s]\n 49%|████▉     | 57/116 [00:04<00:03, 15.92it/s]\n 51%|█████     | 59/116 [00:05<00:03, 15.87it/s]\n 53%|█████▎    | 61/116 [00:05<00:03, 15.94it/s]\n 54%|█████▍    | 63/116 [00:05<00:03, 16.00it/s]\n 56%|█████▌    | 65/116 [00:05<00:03, 16.06it/s]\n 58%|█████▊    | 67/116 [00:05<00:03, 16.01it/s]\n 59%|█████▉    | 69/116 [00:05<00:02, 16.05it/s]\n 61%|██████    | 71/116 [00:05<00:02, 16.06it/s]\n 63%|██████▎   | 73/116 [00:05<00:02, 16.09it/s]\n 65%|██████▍   | 75/116 [00:06<00:02, 15.72it/s]\n 66%|██████▋   | 77/116 [00:06<00:02, 15.75it/s]\n 68%|██████▊   | 79/116 [00:06<00:02, 15.66it/s]\n 70%|██████▉   | 81/116 [00:06<00:02, 15.75it/s]\n 72%|███████▏  | 83/116 [00:06<00:02, 15.84it/s]\n 73%|███████▎  | 85/116 [00:06<00:01, 15.84it/s]\n 75%|███████▌  | 87/116 [00:06<00:01, 15.81it/s]\n 77%|███████▋  | 89/116 [00:06<00:01, 15.84it/s]\n 78%|███████▊  | 91/116 [00:07<00:01, 15.84it/s]\n 80%|████████  | 93/116 [00:07<00:01, 15.91it/s]\n 82%|████████▏ | 95/116 [00:07<00:01, 15.86it/s]\n 84%|████████▎ | 97/116 [00:07<00:01, 15.89it/s]\n 85%|████████▌ | 99/116 [00:07<00:01, 15.87it/s]\n 87%|████████▋ | 101/116 [00:07<00:00, 15.92it/s]\n 89%|████████▉ | 103/116 [00:07<00:00, 15.79it/s]\n 91%|█████████ | 105/116 [00:07<00:00, 15.36it/s]\n 92%|█████████▏| 107/116 [00:08<00:00, 15.20it/s]\n 94%|█████████▍| 109/116 [00:08<00:00, 14.89it/s]\n 96%|█████████▌| 111/116 [00:08<00:00, 14.91it/s]\n 97%|█████████▋| 113/116 [00:08<00:00, 14.87it/s]\n 99%|█████████▉| 115/116 [00:08<00:00, 14.91it/s]\n100%|██████████| 116/116 [00:08<00:00, 13.33it/s]\n2024-02-25 12:49:19 | INFO | fairseq.tasks.hubert_pretraining | current directory is /src\n2024-02-25 12:49:19 | 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}\n2024-02-25 12:49:19 | 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}\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/nn/utils/weight_norm.py:28: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.\nwarnings.warn(\"torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.\")\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/nn/utils/weight_norm.py:28: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.\nwarnings.warn(\"torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.\")\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.\nNote: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)\nreturn _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.\nNote: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)\nreturn _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.\nNote: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)\nreturn _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.\nNote: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)\nreturn _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.\nNote: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)\nreturn _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]\n/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/autograd/__init__.py:266: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed.  This is not an error, but may impair performance.\ngrad.sizes() = [64, 1, 4], strides() = [4, 1, 1]\nbucket_view.sizes() = [64, 1, 4], strides() = [4, 4, 1] (Triggered internally at ../torch/csrc/distributed/c10d/reducer.cpp:322.)\nVariable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass\n/root/.pyenv/versions/3.8.10/lib/python3.8/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 20 leaked semaphore objects to clean up at shutdown\nwarnings.warn('resource_tracker: There appear to be %d '\n  0%|          | 0/65 [00:00<?, ?it/s]\n  2%|▏         | 1/65 [00:00<00:06,  9.82it/s]\n  5%|▍         | 3/65 [00:00<00:05, 11.97it/s]\n  8%|▊         | 5/65 [00:00<00:04, 13.09it/s]\n 11%|█         | 7/65 [00:00<00:04, 13.78it/s]\n 14%|█▍        | 9/65 [00:00<00:03, 14.08it/s]\n 17%|█▋        | 11/65 [00:00<00:03, 14.28it/s]\n 20%|██        | 13/65 [00:00<00:03, 14.47it/s]\n 23%|██▎       | 15/65 [00:01<00:03, 14.56it/s]\n 26%|██▌       | 17/65 [00:01<00:03, 14.66it/s]\n 29%|██▉       | 19/65 [00:01<00:03, 14.64it/s]\n 32%|███▏      | 21/65 [00:01<00:02, 14.76it/s]\n 35%|███▌      | 23/65 [00:01<00:02, 14.84it/s]\n 38%|███▊      | 25/65 [00:01<00:02, 14.95it/s]\n 42%|████▏     | 27/65 [00:01<00:02, 15.00it/s]\n 45%|████▍     | 29/65 [00:02<00:02, 15.02it/s]\n 48%|████▊     | 31/65 [00:02<00:02, 14.98it/s]\n 51%|█████     | 33/65 [00:02<00:02, 14.79it/s]\n 54%|█████▍    | 35/65 [00:02<00:02, 14.89it/s]\n 57%|█████▋    | 37/65 [00:02<00:01, 14.84it/s]\n 60%|██████    | 39/65 [00:02<00:01, 14.79it/s]\n 63%|██████▎   | 41/65 [00:02<00:01, 14.79it/s]\n 66%|██████▌   | 43/65 [00:02<00:01, 14.84it/s]\n 69%|██████▉   | 45/65 [00:03<00:01, 14.84it/s]\n 72%|███████▏  | 47/65 [00:03<00:01, 14.78it/s]\n 75%|███████▌  | 49/65 [00:03<00:01, 14.78it/s]\n 78%|███████▊  | 51/65 [00:03<00:00, 14.82it/s]\n 82%|████████▏ | 53/65 [00:03<00:00, 14.93it/s]\n 85%|████████▍ | 55/65 [00:03<00:00, 14.89it/s]\n 88%|████████▊ | 57/65 [00:03<00:00, 14.88it/s]\n 91%|█████████ | 59/65 [00:04<00:00, 14.90it/s]\n 94%|█████████▍| 61/65 [00:04<00:00, 14.90it/s]\n 97%|█████████▋| 63/65 [00:04<00:00, 14.98it/s]\n100%|██████████| 65/65 [00:04<00:00, 14.97it/s]\n100%|██████████| 65/65 [00:04<00:00, 14.67it/s]",
  "metrics": {
    "predict_time": 262.949361,
    "total_time": 427.936801
  },
  "output": {
    "ckpt_path": "https://replicate.delivery/pbxt/V3fTVi4SctXuQSxtwkw57z1fJdHeSzJeUXJl1DvlmG02e0RTC/default.pth",
    "cloned_audio": "https://replicate.delivery/pbxt/dIl0M30PMzaTLpGgF2v4BpZAqhBOMESzTcVTylAFpmW7pjmE/audio_cloned.wav"
  },
  "started_at": "2024-02-25T12:47:35.373473Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/kwohmxzbqtrk2ssyxusm2wf4sy",
    "cancel": "https://api.replicate.com/v1/predictions/kwohmxzbqtrk2ssyxusm2wf4sy/cancel"
  },
  "version": "07402961aee9e589b0d5fd05368158da3f8df772849c95c2a08c2e3d031fd19a"
}

Generated in

4 minutes 23 seconds

Tweak itReport

0%|          | 0/116 [00:00<?, ?it/s]
  1%|          | 1/116 [00:01<02:31,  1.32s/it]
  3%|▎         | 3/116 [00:01<00:46,  2.44it/s]
  4%|▍         | 5/116 [00:01<00:26,  4.19it/s]
  6%|▌         | 7/116 [00:01<00:17,  6.09it/s]
  8%|▊         | 9/116 [00:01<00:13,  7.95it/s]
  9%|▉         | 11/116 [00:02<00:10,  9.62it/s]
 11%|█         | 13/116 [00:02<00:09, 11.13it/s]
 13%|█▎        | 15/116 [00:02<00:08, 12.23it/s]
 15%|█▍        | 17/116 [00:02<00:07, 13.11it/s]
 16%|█▋        | 19/116 [00:02<00:06, 13.90it/s]
 18%|█▊        | 21/116 [00:02<00:06, 14.54it/s]
 20%|█▉        | 23/116 [00:02<00:06, 14.98it/s]
 22%|██▏       | 25/116 [00:02<00:05, 15.21it/s]
 23%|██▎       | 27/116 [00:03<00:05, 15.31it/s]
 25%|██▌       | 29/116 [00:03<00:05, 15.47it/s]
 27%|██▋       | 31/116 [00:03<00:05, 15.77it/s]
 28%|██▊       | 33/116 [00:03<00:05, 15.90it/s]
 30%|███       | 35/116 [00:03<00:05, 16.05it/s]
 32%|███▏      | 37/116 [00:03<00:04, 16.16it/s]
 34%|███▎      | 39/116 [00:03<00:04, 16.15it/s]
 35%|███▌      | 41/116 [00:03<00:04, 15.93it/s]
 37%|███▋      | 43/116 [00:04<00:04, 15.79it/s]
 39%|███▉      | 45/116 [00:04<00:04, 15.85it/s]
 41%|████      | 47/116 [00:04<00:04, 15.82it/s]
 42%|████▏     | 49/116 [00:04<00:04, 15.92it/s]
 44%|████▍     | 51/116 [00:04<00:04, 15.89it/s]
 46%|████▌     | 53/116 [00:04<00:03, 15.96it/s]
 47%|████▋     | 55/116 [00:04<00:03, 15.96it/s]
 49%|████▉     | 57/116 [00:04<00:03, 15.92it/s]
 51%|█████     | 59/116 [00:05<00:03, 15.87it/s]
 53%|█████▎    | 61/116 [00:05<00:03, 15.94it/s]
 54%|█████▍    | 63/116 [00:05<00:03, 16.00it/s]
 56%|█████▌    | 65/116 [00:05<00:03, 16.06it/s]
 58%|█████▊    | 67/116 [00:05<00:03, 16.01it/s]
 59%|█████▉    | 69/116 [00:05<00:02, 16.05it/s]
 61%|██████    | 71/116 [00:05<00:02, 16.06it/s]
 63%|██████▎   | 73/116 [00:05<00:02, 16.09it/s]
 65%|██████▍   | 75/116 [00:06<00:02, 15.72it/s]
 66%|██████▋   | 77/116 [00:06<00:02, 15.75it/s]
 68%|██████▊   | 79/116 [00:06<00:02, 15.66it/s]
 70%|██████▉   | 81/116 [00:06<00:02, 15.75it/s]
 72%|███████▏  | 83/116 [00:06<00:02, 15.84it/s]
 73%|███████▎  | 85/116 [00:06<00:01, 15.84it/s]
 75%|███████▌  | 87/116 [00:06<00:01, 15.81it/s]
 77%|███████▋  | 89/116 [00:06<00:01, 15.84it/s]
 78%|███████▊  | 91/116 [00:07<00:01, 15.84it/s]
 80%|████████  | 93/116 [00:07<00:01, 15.91it/s]
 82%|████████▏ | 95/116 [00:07<00:01, 15.86it/s]
 84%|████████▎ | 97/116 [00:07<00:01, 15.89it/s]
 85%|████████▌ | 99/116 [00:07<00:01, 15.87it/s]
 87%|████████▋ | 101/116 [00:07<00:00, 15.92it/s]
 89%|████████▉ | 103/116 [00:07<00:00, 15.79it/s]
 91%|█████████ | 105/116 [00:07<00:00, 15.36it/s]
 92%|█████████▏| 107/116 [00:08<00:00, 15.20it/s]
 94%|█████████▍| 109/116 [00:08<00:00, 14.89it/s]
 96%|█████████▌| 111/116 [00:08<00:00, 14.91it/s]
 97%|█████████▋| 113/116 [00:08<00:00, 14.87it/s]
 99%|█████████▉| 115/116 [00:08<00:00, 14.91it/s]
100%|██████████| 116/116 [00:08<00:00, 13.33it/s]
2024-02-25 12:49:19 | INFO | fairseq.tasks.hubert_pretraining | current directory is /src
2024-02-25 12:49:19 | 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}
2024-02-25 12:49:19 | 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}
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/nn/utils/weight_norm.py:28: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/nn/utils/weight_norm.py:28: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.
Note: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)
return _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.
Note: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)
return _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.
Note: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)
return _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.
Note: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)
return _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/functional.py:660: UserWarning: stft with return_complex=False is deprecated. In a future pytorch release, stft will return complex tensors for all inputs, and return_complex=False will raise an error.
Note: you can still call torch.view_as_real on the complex output to recover the old return format. (Triggered internally at ../aten/src/ATen/native/SpectralOps.cpp:874.)
return _VF.stft(input, n_fft, hop_length, win_length, window,  # type: ignore[attr-defined]
/root/.pyenv/versions/3.8.10/lib/python3.8/site-packages/torch/autograd/__init__.py:266: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed.  This is not an error, but may impair performance.
grad.sizes() = [64, 1, 4], strides() = [4, 1, 1]
bucket_view.sizes() = [64, 1, 4], strides() = [4, 4, 1] (Triggered internally at ../torch/csrc/distributed/c10d/reducer.cpp:322.)
Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/root/.pyenv/versions/3.8.10/lib/python3.8/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 20 leaked semaphore objects to clean up at shutdown
warnings.warn('resource_tracker: There appear to be %d '
  0%|          | 0/65 [00:00<?, ?it/s]
  2%|▏         | 1/65 [00:00<00:06,  9.82it/s]
  5%|▍         | 3/65 [00:00<00:05, 11.97it/s]
  8%|▊         | 5/65 [00:00<00:04, 13.09it/s]
 11%|█         | 7/65 [00:00<00:04, 13.78it/s]
 14%|█▍        | 9/65 [00:00<00:03, 14.08it/s]
 17%|█▋        | 11/65 [00:00<00:03, 14.28it/s]
 20%|██        | 13/65 [00:00<00:03, 14.47it/s]
 23%|██▎       | 15/65 [00:01<00:03, 14.56it/s]
 26%|██▌       | 17/65 [00:01<00:03, 14.66it/s]
 29%|██▉       | 19/65 [00:01<00:03, 14.64it/s]
 32%|███▏      | 21/65 [00:01<00:02, 14.76it/s]
 35%|███▌      | 23/65 [00:01<00:02, 14.84it/s]
 38%|███▊      | 25/65 [00:01<00:02, 14.95it/s]
 42%|████▏     | 27/65 [00:01<00:02, 15.00it/s]
 45%|████▍     | 29/65 [00:02<00:02, 15.02it/s]
 48%|████▊     | 31/65 [00:02<00:02, 14.98it/s]
 51%|█████     | 33/65 [00:02<00:02, 14.79it/s]
 54%|█████▍    | 35/65 [00:02<00:02, 14.89it/s]
 57%|█████▋    | 37/65 [00:02<00:01, 14.84it/s]
 60%|██████    | 39/65 [00:02<00:01, 14.79it/s]
 63%|██████▎   | 41/65 [00:02<00:01, 14.79it/s]
 66%|██████▌   | 43/65 [00:02<00:01, 14.84it/s]
 69%|██████▉   | 45/65 [00:03<00:01, 14.84it/s]
 72%|███████▏  | 47/65 [00:03<00:01, 14.78it/s]
 75%|███████▌  | 49/65 [00:03<00:01, 14.78it/s]
 78%|███████▊  | 51/65 [00:03<00:00, 14.82it/s]
 82%|████████▏ | 53/65 [00:03<00:00, 14.93it/s]
 85%|████████▍ | 55/65 [00:03<00:00, 14.89it/s]
 88%|████████▊ | 57/65 [00:03<00:00, 14.88it/s]
 91%|█████████ | 59/65 [00:04<00:00, 14.90it/s]
 94%|█████████▍| 61/65 [00:04<00:00, 14.90it/s]
 97%|█████████▋| 63/65 [00:04<00:00, 14.98it/s]
100%|██████████| 65/65 [00:04<00:00, 14.97it/s]
100%|██████████| 65/65 [00:04<00:00, 14.67it/s]

This output was created using a different version of the model, datong-new/rvc:07402961.

Run time and cost

This model runs on Nvidia T4 GPU hardware. We don't yet have enough runs of this model to provide performance information.

Readme

This model doesn't have a readme.