mtg / effnet-discogs

An EfficientNet for music style classification by 400 styles from the Discogs taxonomy

Warm

Public
153.9K runs
CPU
GitHub
License

Run with an API

Playground API Examples README Versions

Input

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

npx create-replicate --model=mtg/effnet-discogs

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 mtg/effnet-discogs using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "mtg/effnet-discogs:1532dd069fb4f0e27c6833e28815f6b8c194dfec76fd9cd73460540fd720ffe1",
  {
    input: {
      url: "https://www.youtube.com/watch?v=CHekNnySAfM",
      top_n: 10,
      output_format: "Visualization"
    }
  }
);
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 mtg/effnet-discogs using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "mtg/effnet-discogs:1532dd069fb4f0e27c6833e28815f6b8c194dfec76fd9cd73460540fd720ffe1",
    input={
        "url": "https://www.youtube.com/watch?v=CHekNnySAfM",
        "top_n": 10,
        "output_format": "Visualization"
    }
)
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 mtg/effnet-discogs 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": "1532dd069fb4f0e27c6833e28815f6b8c194dfec76fd9cd73460540fd720ffe1",
    "input": {
      "url": "https://www.youtube.com/watch?v=CHekNnySAfM",
      "top_n": 10,
      "output_format": "Visualization"
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2022-03-16T13:56:03.874572Z",
  "created_at": "2022-03-16T13:54:24.327306Z",
  "data_removed": false,
  "error": null,
  "id": "mnjyyydfrjdwjp7jtsfj2gnysm",
  "input": {
    "url": "https://www.youtube.com/watch?v=CHekNnySAfM",
    "top_n": "10",
    "output_format": "Visualization"
  },
  "logs": "[youtube] CHekNnySAfM: Downloading webpage\n[youtube] CHekNnySAfM: Downloading player bd67d609\n[download] Destination: /tmp/tmp5hp8fkyh/audio.webm\n\n[download]   0.0% of 3.82MiB at  7.33KiB/s ETA 08:53\n[download]   0.1% of 3.82MiB at 21.97KiB/s ETA 02:57\n[download]   0.2% of 3.82MiB at 51.24KiB/s ETA 01:16\n[download]   0.4% of 3.82MiB at 109.68KiB/s ETA 00:35\n[download]   0.8% of 3.82MiB at 88.27KiB/s ETA 00:43\n[download]   1.6% of 3.82MiB at 76.81KiB/s ETA 00:50\n[download]   3.2% of 3.82MiB at 73.03KiB/s ETA 00:51\n[download]   5.0% of 3.82MiB at 68.01KiB/s ETA 00:54\n[download]   6.6% of 3.82MiB at 67.48KiB/s ETA 00:54\n[download]   8.3% of 3.82MiB at 68.29KiB/s ETA 00:52\n[download]  10.1% of 3.82MiB at 69.88KiB/s ETA 00:50\n[download]  12.1% of 3.82MiB at 69.50KiB/s ETA 00:49\n[download]  13.8% of 3.82MiB at 70.00KiB/s ETA 00:48\n[download]  15.7% of 3.82MiB at 69.33KiB/s ETA 00:47\n[download]  17.3% of 3.82MiB at 69.45KiB/s ETA 00:46\n[download]  19.2% of 3.82MiB at 70.08KiB/s ETA 00:45\n[download]  21.1% of 3.82MiB at 69.75KiB/s ETA 00:44\n[download]  22.8% of 3.82MiB at 69.95KiB/s ETA 00:43\n[download]  24.7% of 3.82MiB at 69.37KiB/s ETA 00:42\n[download]  26.3% of 3.82MiB at 69.24KiB/s ETA 00:41\n[download]  28.0% of 3.82MiB at 69.47KiB/s ETA 00:40\n[download]  29.9% of 3.82MiB at 69.07KiB/s ETA 00:39\n[download]  31.5% of 3.82MiB at 69.07KiB/s ETA 00:38\n[download]  33.3% of 3.82MiB at 69.38KiB/s ETA 00:37\n[download]  35.2% of 3.82MiB at 69.21KiB/s ETA 00:36\n[download]  36.9% of 3.82MiB at 69.35KiB/s ETA 00:35\n[download]  38.7% of 3.82MiB at 69.73KiB/s ETA 00:34\n[download]  40.7% of 3.82MiB at 69.65KiB/s ETA 00:33\n[download]  42.5% of 3.82MiB at 69.82KiB/s ETA 00:32\n[download]  44.4% of 3.82MiB at 69.56KiB/s ETA 00:31\n[download]  46.0% of 3.82MiB at 69.57KiB/s ETA 00:30\n[download]  47.8% of 3.82MiB at 69.80KiB/s ETA 00:29\n[download]  49.8% of 3.82MiB at 69.65KiB/s ETA 00:28\n[download]  51.4% of 3.82MiB at 69.72KiB/s ETA 00:27\n[download]  53.3% of 3.82MiB at 69.48KiB/s ETA 00:26\n[download]  54.9% of 3.82MiB at 69.43KiB/s ETA 00:25\n[download]  56.6% of 3.82MiB at 69.56KiB/s ETA 00:24\n[download]  58.5% of 3.82MiB at 69.36KiB/s ETA 00:23\n[download]  60.2% of 3.82MiB at 69.37KiB/s ETA 00:22\n[download]  61.9% of 3.82MiB at 69.54KiB/s ETA 00:21\n[download]  63.9% of 3.82MiB at 69.42KiB/s ETA 00:20\n[download]  65.6% of 3.82MiB at 69.48KiB/s ETA 00:19\n[download]  67.4% of 3.82MiB at 69.68KiB/s ETA 00:18\n[download]  69.4% of 3.82MiB at 69.62KiB/s ETA 00:17\n[download]  71.1% of 3.82MiB at 69.71KiB/s ETA 00:16\n[download]  73.0% of 3.82MiB at 69.54KiB/s ETA 00:15\n[download]  74.6% of 3.82MiB at 69.53KiB/s ETA 00:14\n[download]  76.4% of 3.82MiB at 69.66KiB/s ETA 00:13\n[download]  78.3% of 3.82MiB at 69.58KiB/s ETA 00:12\n[download]  80.0% of 3.82MiB at 69.64KiB/s ETA 00:11\n[download]  81.9% of 3.82MiB at 69.44KiB/s ETA 00:10\n[download]  83.4% of 3.82MiB at 69.40KiB/s ETA 00:09\n[download]  85.2% of 3.82MiB at 69.48KiB/s ETA 00:08\n[download]  87.0% of 3.82MiB at 69.67KiB/s ETA 00:07\n[download]  89.1% of 3.82MiB at 69.66KiB/s ETA 00:06\n[download]  90.9% of 3.82MiB at 69.45KiB/s ETA 00:05\n[download]  92.4% of 3.82MiB at 69.69KiB/s ETA 00:04\n[download]  94.6% of 3.82MiB at 69.55KiB/s ETA 00:03\n[download]  96.3% of 3.82MiB at 69.56KiB/s ETA 00:02\n[download]  98.1% of 3.82MiB at 69.67KiB/s ETA 00:01\n[download] 100.0% of 3.82MiB at 69.70KiB/s ETA 00:00\n[download] 100% of 3.82MiB in 00:56\n[ffmpeg] Destination: /tmp/tmp5hp8fkyh/audio.wav\nDeleting original file /tmp/tmp5hp8fkyh/audio.webm (pass -k to keep)\nrunning the inference network...\n2022-03-16 13:55:58.101663: I tensorflow/core/platform/profile_utils/cpu_utils.cc:114] CPU Frequency: 2299995000 Hz\nplotting...\ndone!",
  "metrics": {
    "predict_time": 65.628347,
    "total_time": 99.547266
  },
  "output": "https://replicate.delivery/mgxm/ed180118-13c0-410b-bec1-478391f5bd67/out.png",
  "started_at": "2022-03-16T13:54:58.246225Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/mnjyyydfrjdwjp7jtsfj2gnysm",
    "cancel": "https://api.replicate.com/v1/predictions/mnjyyydfrjdwjp7jtsfj2gnysm/cancel"
  },
  "version": "3b1d08bbeea1030fa31a71eef8e9440857cd3b3c780fd928b4d3068366d3cabb"
}

Generated in

65.6 seconds

Tweak itReport

[youtube] CHekNnySAfM: Downloading webpage
[youtube] CHekNnySAfM: Downloading player bd67d609
[download] Destination: /tmp/tmp5hp8fkyh/audio.webm

[download]   0.0% of 3.82MiB at  7.33KiB/s ETA 08:53
[download]   0.1% of 3.82MiB at 21.97KiB/s ETA 02:57
[download]   0.2% of 3.82MiB at 51.24KiB/s ETA 01:16
[download]   0.4% of 3.82MiB at 109.68KiB/s ETA 00:35
[download]   0.8% of 3.82MiB at 88.27KiB/s ETA 00:43
[download]   1.6% of 3.82MiB at 76.81KiB/s ETA 00:50
[download]   3.2% of 3.82MiB at 73.03KiB/s ETA 00:51
[download]   5.0% of 3.82MiB at 68.01KiB/s ETA 00:54
[download]   6.6% of 3.82MiB at 67.48KiB/s ETA 00:54
[download]   8.3% of 3.82MiB at 68.29KiB/s ETA 00:52
[download]  10.1% of 3.82MiB at 69.88KiB/s ETA 00:50
[download]  12.1% of 3.82MiB at 69.50KiB/s ETA 00:49
[download]  13.8% of 3.82MiB at 70.00KiB/s ETA 00:48
[download]  15.7% of 3.82MiB at 69.33KiB/s ETA 00:47
[download]  17.3% of 3.82MiB at 69.45KiB/s ETA 00:46
[download]  19.2% of 3.82MiB at 70.08KiB/s ETA 00:45
[download]  21.1% of 3.82MiB at 69.75KiB/s ETA 00:44
[download]  22.8% of 3.82MiB at 69.95KiB/s ETA 00:43
[download]  24.7% of 3.82MiB at 69.37KiB/s ETA 00:42
[download]  26.3% of 3.82MiB at 69.24KiB/s ETA 00:41
[download]  28.0% of 3.82MiB at 69.47KiB/s ETA 00:40
[download]  29.9% of 3.82MiB at 69.07KiB/s ETA 00:39
[download]  31.5% of 3.82MiB at 69.07KiB/s ETA 00:38
[download]  33.3% of 3.82MiB at 69.38KiB/s ETA 00:37
[download]  35.2% of 3.82MiB at 69.21KiB/s ETA 00:36
[download]  36.9% of 3.82MiB at 69.35KiB/s ETA 00:35
[download]  38.7% of 3.82MiB at 69.73KiB/s ETA 00:34
[download]  40.7% of 3.82MiB at 69.65KiB/s ETA 00:33
[download]  42.5% of 3.82MiB at 69.82KiB/s ETA 00:32
[download]  44.4% of 3.82MiB at 69.56KiB/s ETA 00:31
[download]  46.0% of 3.82MiB at 69.57KiB/s ETA 00:30
[download]  47.8% of 3.82MiB at 69.80KiB/s ETA 00:29
[download]  49.8% of 3.82MiB at 69.65KiB/s ETA 00:28
[download]  51.4% of 3.82MiB at 69.72KiB/s ETA 00:27
[download]  53.3% of 3.82MiB at 69.48KiB/s ETA 00:26
[download]  54.9% of 3.82MiB at 69.43KiB/s ETA 00:25
[download]  56.6% of 3.82MiB at 69.56KiB/s ETA 00:24
[download]  58.5% of 3.82MiB at 69.36KiB/s ETA 00:23
[download]  60.2% of 3.82MiB at 69.37KiB/s ETA 00:22
[download]  61.9% of 3.82MiB at 69.54KiB/s ETA 00:21
[download]  63.9% of 3.82MiB at 69.42KiB/s ETA 00:20
[download]  65.6% of 3.82MiB at 69.48KiB/s ETA 00:19
[download]  67.4% of 3.82MiB at 69.68KiB/s ETA 00:18
[download]  69.4% of 3.82MiB at 69.62KiB/s ETA 00:17
[download]  71.1% of 3.82MiB at 69.71KiB/s ETA 00:16
[download]  73.0% of 3.82MiB at 69.54KiB/s ETA 00:15
[download]  74.6% of 3.82MiB at 69.53KiB/s ETA 00:14
[download]  76.4% of 3.82MiB at 69.66KiB/s ETA 00:13
[download]  78.3% of 3.82MiB at 69.58KiB/s ETA 00:12
[download]  80.0% of 3.82MiB at 69.64KiB/s ETA 00:11
[download]  81.9% of 3.82MiB at 69.44KiB/s ETA 00:10
[download]  83.4% of 3.82MiB at 69.40KiB/s ETA 00:09
[download]  85.2% of 3.82MiB at 69.48KiB/s ETA 00:08
[download]  87.0% of 3.82MiB at 69.67KiB/s ETA 00:07
[download]  89.1% of 3.82MiB at 69.66KiB/s ETA 00:06
[download]  90.9% of 3.82MiB at 69.45KiB/s ETA 00:05
[download]  92.4% of 3.82MiB at 69.69KiB/s ETA 00:04
[download]  94.6% of 3.82MiB at 69.55KiB/s ETA 00:03
[download]  96.3% of 3.82MiB at 69.56KiB/s ETA 00:02
[download]  98.1% of 3.82MiB at 69.67KiB/s ETA 00:01
[download] 100.0% of 3.82MiB at 69.70KiB/s ETA 00:00
[download] 100% of 3.82MiB in 00:56
[ffmpeg] Destination: /tmp/tmp5hp8fkyh/audio.wav
Deleting original file /tmp/tmp5hp8fkyh/audio.webm (pass -k to keep)
running the inference network...
2022-03-16 13:55:58.101663: I tensorflow/core/platform/profile_utils/cpu_utils.cc:114] CPU Frequency: 2299995000 Hz
plotting...
done!

This example was created by a different version, mtg/effnet-discogs:3b1d08bb.

Examples

View more examples

Run time and cost

This model costs approximately $0.00025 to run on Replicate, or 4000 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker.

This model runs on CPU hardware. Predictions typically complete within 3 seconds.

Readme

effnet-discogs

effnet-discogs is an EfficientNet architecture trained to predict music styles for 400 of the most popular Discogs music styles. The output plot also shows the Discogs genre the predicted style belongs to.

This model was trained in more than two million music recordings from an in-house dataset annotated by Discogs metadata and is part of an ongoing research.

The architecture consists of an EfficientNet on its B0 configuration with an additional penultimate dense layer plus batch normalization to facilitate using the model as an embedding extractor.

This demo outputs the top_n music style activations, summarized as their mean and standard deviation through time.

License

This model is part of Essentia Models made by MTG-UPF.