Prediction

Model

ttsds/fishspeech_1_2_sft:34a7e498

50t54px1jxrme0cmnrxaz05sp4

Status

Succeeded

Source

Web

Hardware

L40S

Total duration

78.3s

Created

11 months ago

Webhook

–

Input

text: With tenure, Suzie'd have all the more leisure for yachting, but her publications are no good.
speaker_reference: https://replicate.delivery/pbxt/MNFXdPaUPOwYCZjZM4azsymbzE2TCV2WJXfGpeV2DrFWaSq8/example_en.wav
text_reference: and keeping eternity before the eyes, though much

{
  "speaker_reference": "https://replicate.delivery/pbxt/MNFXdPaUPOwYCZjZM4azsymbzE2TCV2WJXfGpeV2DrFWaSq8/example_en.wav",
  "text": "With tenure, Suzie'd have all the more leisure for yachting, but her publications are no good.",
  "text_reference": "and keeping eternity before the eyes, though much"
}

Install Replicate’s Node.js client library:

npm install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=r8_ZF7**********************************

This is your API token. Keep it to yourself.

Import and set up the client:

import Replicate from "replicate";
import fs from "node:fs";

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

Run ttsds/fishspeech_1_2_sft using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "ttsds/fishspeech_1_2_sft:34a7e498d81e49e7200ee9aaa52f18b2529f30162ba6afd163b43037aa5a5d20",
  {
    input: {
      speaker_reference: "https://replicate.delivery/pbxt/MNFXdPaUPOwYCZjZM4azsymbzE2TCV2WJXfGpeV2DrFWaSq8/example_en.wav",
      text: "With tenure, Suzie'd have all the more leisure for yachting, but her publications are no good.",
      text_reference: "and keeping eternity before the eyes, though much"
    }
  }
);

// To access the file URL:
console.log(output.url()); //=> "http://example.com"

// To write the file to disk:
fs.writeFile("my-image.png", 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=r8_ZF7**********************************

This is your API token. Keep it to yourself.

Import the client:

import replicate

Run ttsds/fishspeech_1_2_sft using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "ttsds/fishspeech_1_2_sft:34a7e498d81e49e7200ee9aaa52f18b2529f30162ba6afd163b43037aa5a5d20",
    input={
        "speaker_reference": "https://replicate.delivery/pbxt/MNFXdPaUPOwYCZjZM4azsymbzE2TCV2WJXfGpeV2DrFWaSq8/example_en.wav",
        "text": "With tenure, Suzie'd have all the more leisure for yachting, but her publications are no good.",
        "text_reference": "and keeping eternity before the eyes, though much"
    }
)

# To access the file URL:
print(output.url())
#=> "http://example.com"

# To write the file to disk:
with open("my-image.png", "wb") as file:
    file.write(output.read())

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=r8_ZF7**********************************

This is your API token. Keep it to yourself.

Run ttsds/fishspeech_1_2_sft 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": "ttsds/fishspeech_1_2_sft:34a7e498d81e49e7200ee9aaa52f18b2529f30162ba6afd163b43037aa5a5d20",
    "input": {
      "speaker_reference": "https://replicate.delivery/pbxt/MNFXdPaUPOwYCZjZM4azsymbzE2TCV2WJXfGpeV2DrFWaSq8/example_en.wav",
      "text": "With tenure, Suzie\'d have all the more leisure for yachting, but her publications are no good.",
      "text_reference": "and keeping eternity before the eyes, though much"
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "id": "50t54px1jxrme0cmnrxaz05sp4",
  "model": "ttsds/fishspeech_1_2_sft",
  "version": "34a7e498d81e49e7200ee9aaa52f18b2529f30162ba6afd163b43037aa5a5d20",
  "input": {
    "speaker_reference": "https://replicate.delivery/pbxt/MNFXdPaUPOwYCZjZM4azsymbzE2TCV2WJXfGpeV2DrFWaSq8/example_en.wav",
    "text": "With tenure, Suzie'd have all the more leisure for yachting, but her publications are no good.",
    "text_reference": "and keeping eternity before the eyes, though much"
  },
  "logs": "2025-01-28 18:45:39.050 | INFO     | tools.llama.generate:generate_long:432 - Encoded text: With tenure, Suzie'd have all the more leisure for yachting, but her publications are no good.\n2025-01-28 18:45:39.050 | INFO     | tools.llama.generate:generate_long:450 - Generating sentence 1/1 of sample 1/1\n  0%|          | 0/3892 [00:00<?, ?it/s]/root/.pyenv/versions/3.11.10/lib/python3.11/site-packages/torch/backends/cuda/__init__.py:342: FutureWarning: torch.backends.cuda.sdp_kernel() is deprecated. In the future, this context manager will be removed. Please see, torch.nn.attention.sdpa_kernel() for the new context manager, with updated signature.\nwarnings.warn(\n  0%|          | 6/3892 [00:00<01:10, 54.92it/s]\n  0%|          | 12/3892 [00:00<01:08, 56.30it/s]\n  0%|          | 18/3892 [00:00<01:08, 56.74it/s]\n  1%|          | 24/3892 [00:00<01:07, 56.99it/s]\n  1%|          | 30/3892 [00:00<01:07, 57.13it/s]\n  1%|          | 36/3892 [00:00<01:07, 57.23it/s]\n  1%|          | 42/3892 [00:00<01:07, 57.29it/s]\n  1%|          | 48/3892 [00:00<01:07, 57.26it/s]\n  1%|▏         | 54/3892 [00:00<01:07, 57.27it/s]\n  2%|▏         | 60/3892 [00:01<01:06, 57.21it/s]\n  2%|▏         | 66/3892 [00:01<01:06, 57.23it/s]\n  2%|▏         | 72/3892 [00:01<01:06, 57.22it/s]\n  2%|▏         | 78/3892 [00:01<01:06, 57.22it/s]\n  2%|▏         | 84/3892 [00:01<01:07, 56.64it/s]\n  2%|▏         | 90/3892 [00:01<01:07, 56.65it/s]\n  2%|▏         | 96/3892 [00:01<01:06, 56.84it/s]\n  3%|▎         | 102/3892 [00:01<01:07, 56.24it/s]\n  3%|▎         | 108/3892 [00:01<01:06, 56.51it/s]\n  3%|▎         | 114/3892 [00:02<01:06, 56.73it/s]\n  3%|▎         | 120/3892 [00:02<01:06, 56.94it/s]\n  3%|▎         | 126/3892 [00:02<01:06, 57.04it/s]\n  3%|▎         | 132/3892 [00:02<01:05, 57.14it/s]\n  4%|▎         | 138/3892 [00:02<01:06, 56.88it/s]\n  4%|▎         | 144/3892 [00:02<01:05, 56.98it/s]\n  4%|▍         | 150/3892 [00:02<01:05, 57.13it/s]\n  4%|▍         | 156/3892 [00:02<01:05, 57.09it/s]\n  4%|▍         | 162/3892 [00:02<01:05, 57.00it/s]\n  4%|▍         | 168/3892 [00:02<01:05, 57.04it/s]\n  4%|▍         | 174/3892 [00:03<01:05, 57.15it/s]\n  5%|▍         | 180/3892 [00:03<01:04, 57.20it/s]\n  5%|▍         | 186/3892 [00:03<01:04, 57.18it/s]\n  5%|▍         | 192/3892 [00:03<01:04, 57.21it/s]\n  5%|▌         | 198/3892 [00:03<01:04, 57.24it/s]\n  5%|▌         | 204/3892 [00:03<01:04, 57.32it/s]\n  5%|▌         | 210/3892 [00:03<01:04, 57.38it/s]\n  6%|▌         | 216/3892 [00:03<01:04, 57.34it/s]\n  6%|▌         | 222/3892 [00:03<01:04, 57.27it/s]\n  6%|▌         | 228/3892 [00:03<01:03, 57.36it/s]\n  6%|▌         | 234/3892 [00:04<01:04, 56.65it/s]\n  6%|▌         | 240/3892 [00:04<01:04, 56.89it/s]\n6%|▌         | 243/3892 [00:04<01:04, 56.78it/s]\n2025-01-28 18:45:43.499 | INFO     | tools.llama.generate:generate_long:505 - Generated 245 tokens in 4.45 seconds, 55.07 tokens/sec\n2025-01-28 18:45:43.499 | INFO     | tools.llama.generate:generate_long:508 - Bandwidth achieved: 27.00 GB/s\n2025-01-28 18:45:43.499 | INFO     | tools.llama.generate:generate_long:513 - GPU Memory used: 1.56 GB\n/root/.pyenv/versions/3.11.10/lib/python3.11/site-packages/torch/nn/modules/conv.py:306: UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR: cudnnFinalize Descriptor Failed cudnn_status: CUDNN_STATUS_NOT_SUPPORTED (Triggered internally at ../aten/src/ATen/native/cudnn/Conv_v8.cpp:919.)\nreturn F.conv1d(input, weight, bias, self.stride,\nNext sample",
  "output": "https://replicate.delivery/xezq/rXAanJ47XKbWJ9gTbnNPibxbNBX3CmzjCKhUb1wyrL81XbCF/generated.wav",
  "data_removed": false,
  "error": null,
  "source": "web",
  "status": "succeeded",
  "created_at": "2025-01-28T18:44:25.623Z",
  "started_at": "2025-01-28T18:45:38.775187Z",
  "completed_at": "2025-01-28T18:45:43.867597Z",
  "urls": {
    "cancel": "https://api.replicate.com/v1/predictions/50t54px1jxrme0cmnrxaz05sp4/cancel",
    "get": "https://api.replicate.com/v1/predictions/50t54px1jxrme0cmnrxaz05sp4",
    "stream": "https://stream.replicate.com/v1/files/bsvm-peyax7uzinhbwwofrvbyjmtxaocejnhcttq6jp3mrgy5jvwh5xbq",
    "web": "https://replicate.com/p/50t54px1jxrme0cmnrxaz05sp4"
  },
  "metrics": {
    "predict_time": 5.09241007,
    "total_time": 78.244597
  }
}

Generated in

5.1 seconds

Tweak it Report

2025-01-28 18:45:39.050 | INFO | tools.llama.generate:generate_long:432 - Encoded text: With tenure, Suzie'd have all the more leisure for yachting, but her publications are no good. 2025-01-28 18:45:39.050 | INFO | tools.llama.generate:generate_long:450 - Generating sentence 1/1 of sample 1/1 0%| | 0/3892 [00:00<?, ?it/s]/root/.pyenv/versions/3.11.10/lib/python3.11/site-packages/torch/backends/cuda/__init__.py:342: FutureWarning: torch.backends.cuda.sdp_kernel() is deprecated. In the future, this context manager will be removed. Please see, torch.nn.attention.sdpa_kernel() for the new context manager, with updated signature. warnings.warn( 0%| | 6/3892 [00:00<01:10, 54.92it/s] 0%| | 12/3892 [00:00<01:08, 56.30it/s] 0%| | 18/3892 [00:00<01:08, 56.74it/s] 1%| | 24/3892 [00:00<01:07, 56.99it/s] 1%| | 30/3892 [00:00<01:07, 57.13it/s] 1%| | 36/3892 [00:00<01:07, 57.23it/s] 1%| | 42/3892 [00:00<01:07, 57.29it/s] 1%| | 48/3892 [00:00<01:07, 57.26it/s] 1%|▏ | 54/3892 [00:00<01:07, 57.27it/s] 2%|▏ | 60/3892 [00:01<01:06, 57.21it/s] 2%|▏ | 66/3892 [00:01<01:06, 57.23it/s] 2%|▏ | 72/3892 [00:01<01:06, 57.22it/s] 2%|▏ | 78/3892 [00:01<01:06, 57.22it/s] 2%|▏ | 84/3892 [00:01<01:07, 56.64it/s] 2%|▏ | 90/3892 [00:01<01:07, 56.65it/s] 2%|▏ | 96/3892 [00:01<01:06, 56.84it/s] 3%|▎ | 102/3892 [00:01<01:07, 56.24it/s] 3%|▎ | 108/3892 [00:01<01:06, 56.51it/s] 3%|▎ | 114/3892 [00:02<01:06, 56.73it/s] 3%|▎ | 120/3892 [00:02<01:06, 56.94it/s] 3%|▎ | 126/3892 [00:02<01:06, 57.04it/s] 3%|▎ | 132/3892 [00:02<01:05, 57.14it/s] 4%|▎ | 138/3892 [00:02<01:06, 56.88it/s] 4%|▎ | 144/3892 [00:02<01:05, 56.98it/s] 4%|▍ | 150/3892 [00:02<01:05, 57.13it/s] 4%|▍ | 156/3892 [00:02<01:05, 57.09it/s] 4%|▍ | 162/3892 [00:02<01:05, 57.00it/s] 4%|▍ | 168/3892 [00:02<01:05, 57.04it/s] 4%|▍ | 174/3892 [00:03<01:05, 57.15it/s] 5%|▍ | 180/3892 [00:03<01:04, 57.20it/s] 5%|▍ | 186/3892 [00:03<01:04, 57.18it/s] 5%|▍ | 192/3892 [00:03<01:04, 57.21it/s] 5%|▌ | 198/3892 [00:03<01:04, 57.24it/s] 5%|▌ | 204/3892 [00:03<01:04, 57.32it/s] 5%|▌ | 210/3892 [00:03<01:04, 57.38it/s] 6%|▌ | 216/3892 [00:03<01:04, 57.34it/s] 6%|▌ | 222/3892 [00:03<01:04, 57.27it/s] 6%|▌ | 228/3892 [00:03<01:03, 57.36it/s] 6%|▌ | 234/3892 [00:04<01:04, 56.65it/s] 6%|▌ | 240/3892 [00:04<01:04, 56.89it/s] 6%|▌ | 243/3892 [00:04<01:04, 56.78it/s] 2025-01-28 18:45:43.499 | INFO | tools.llama.generate:generate_long:505 - Generated 245 tokens in 4.45 seconds, 55.07 tokens/sec 2025-01-28 18:45:43.499 | INFO | tools.llama.generate:generate_long:508 - Bandwidth achieved: 27.00 GB/s 2025-01-28 18:45:43.499 | INFO | tools.llama.generate:generate_long:513 - GPU Memory used: 1.56 GB /root/.pyenv/versions/3.11.10/lib/python3.11/site-packages/torch/nn/modules/conv.py:306: UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR: cudnnFinalize Descriptor Failed cudnn_status: CUDNN_STATUS_NOT_SUPPORTED (Triggered internally at ../aten/src/ATen/native/cudnn/Conv_v8.cpp:919.) return F.conv1d(input, weight, bias, self.stride, Next sample