SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1612.01840 · 2016

FMA: A Dataset For Music Analysis

arXiv · PDF · Open in the Atlas

Code that ran

9 of 14 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 9 have no contract verdict; 0 failures of our instrument; 5 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1612.01840.json · badge: /api/ran/1612.01840.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
StatisticDean/Crystal_Clear community (classifier label)
classifier basis not recorded on the link
5 ran · 2 no recorded run
weepingwillowben/sound-eval community (classifier label)
classifier basis not recorded on the link
4 ran · 3 no recorded run
FunctionStatusWhere it lives
batch_filenames Ran
ran on a synthesized input
weepingwillowben/sound-eval/spectrify_all.py
code served (permissive licence) · get_code("4bb277ee9b64c89a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

# Construct valid arguments for batch_filenames
# The function takes a list of filenames and batches them into chunks of size 100

# Create a list of absolute filenames
abs_filenames = [
    f"/path/to/file_{i:04d}.mp3" for i in range(250)
]

ARGS = {
    "abs_filenames": abs_filenames
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
flatten_audios Ran
ran on a synthesized input
weepingwillowben/sound-eval/spectrogram_doc2vec.py
code served (permissive licence) · get_code("667e6a01204b5c1e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create spec_list - a list of numpy arrays with different shapes
# Each spec represents an audio spectrogram with different time lengths
spec1 = np.random.randn(5, 128)  # 5 time steps, 128 frequency bins
spec2 = np.random.randn(3, 128)  # 3 time steps, 128 frequency bins
spec3 = np.random.randn(7, 128)  # 7 time steps, 128 frequency bins

spec_list = [spec1, spec2, spec3]

ARGS = {
    "spec_list": spec_list
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_all_paths Ran
ran on a synthesized input
weepingwillowben/sound-eval/process_many_files.py
code served (permissive licence) · get_code("aee92765b773b981")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import subprocess
import tempfile
import os

# Create a temporary directory with some test files
temp_dir = tempfile.mkdtemp()
os.makedirs(os.path.join(temp_dir, 'subdir'), exist_ok=True)

# Create some test files with different extensions
open(os.path.join(temp_dir, 'file1.txt'), 'w').close()
open(os.path.join(temp_dir, 'file2.txt'), 'w').close()
open(os.path.join(temp_dir, 'subdir', 'file3.txt'), 'w').close()
open(os.path.join(temp_dir, 'file1.py'), 'w').close()

ARGS = {
    'base_folder': temp_dir,
    'extension': 'txt'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
make_into_3D Ran
ran on a synthesized input
StatisticDean/Crystal_Clear/crystal_clear/prepare.py
code served (permissive licence) · get_code("341a34687ad2fba2")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (3, 4, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Test case 1: 2D array input
data_2d = np.random.randn(4, 5)

# Test case 2: 3D array with 1 channel
data_1ch = np.random.randn(1, 4, 5)

# Test case 3: 3D array with 2 channels
data_2ch = np.random.randn(2, 4, 5)

# We'll use the 2D case as the primary test
ARGS = {
    "data": data_2d
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (3, 4, 5) — shape (3, 4, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
make_weighted_average_on_window Ran
ran on a synthesized input
StatisticDean/Crystal_Clear/crystal_clear/utils.py
code served (permissive licence) · get_code("43f535babf9558b5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (3, 4, 2) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Analyze the function:
# - Takes 'sliding' parameter which is a numpy array
# - Expects shape (n, nc, sz, _) where n is batch, nc is channels, sz is size
# - Uses stride tricks and array operations
# - Returns a weighted average computed over a sliding window

# Create a small test array with distinct dimensions
# Shape: (n=5, nc=3, sz=4, last_dim=2)
n = 5
nc = 3
sz = 4
last_dim = 2

sliding = np.random.randn(n, nc, sz, last_dim).astype(np.float32)

ARGS = {
    "sliding": sliding
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (3, 4, 2) — shape (3, 4, 2) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
path_mp3 Ran
ran on a synthesized input
StatisticDean/Crystal_Clear/crystal_clear/prepare.py
code served (permissive licence) · get_code("735c3f6db89d9f8f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned PosixPath. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from pathlib import Path

# Analyze the function:
# - track_id: can be int or str, gets converted to str and zero-padded
# - folder: string parameter with default 'fma_small'
# The function constructs and returns a Path object

ARGS = {
    'track_id': 42,
    'folder': 'fma_small'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
predict_method Ran
ran on a synthesized input
StatisticDean/Crystal_Clear/crystal_clear/upscale.py
code served (permissive licence) · get_code("9cca6de4e1c97a8d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (8, 3) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

class SimpleProcessor:
    def f_in(self, data):
        # Returns processed data and metadata
        return data, {"meta": "input_metadata"}
    
    def f_batch_in(self, batch):
        # Returns tensor and batch metadata
        return batch, {"batch_meta": "batch_metadata"}
    
    def f_batch_out(self, output, batch_in):
        # Returns processed output
        return output
    
    def f_out(self, data, d_in):
        # Returns final output
        return data

model = SimpleModel()
proc = SimpleProcessor()
data = torch.randn(8, 4)
bs = 2
use_gpu = False

ARGS = {
    "data": data,
    "model": model,
    "proc": proc,
    "bs": bs,
    "use_gpu": use_gpu,
    "show_progress": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (8, 3) — shape (8, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
read_file Ran
ran on a synthesized input
weepingwillowben/sound-eval/gather_computed.py
code served (permissive licence) · get_code("c708ef13e6436433")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import tempfile
import os

# Create a temporary file with some content
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("Hello, this is test content for the read_file function.")
temp_file.close()
temp_filename = temp_file.name

try:
    ARGS = {
        'filename': temp_filename
    }
finally:
    # Clean up will happen after ARGS is defined
    pass
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
wrap_for_processor Ran
ran on a synthesized input
StatisticDean/Crystal_Clear/crystal_clear/utils.py
code served (permissive licence) · get_code("4791d3c12dee84d7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned function. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Define a simple function that takes input x and returns a transformed output
class SimpleProcessor(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create a simple function to wrap
def simple_function(x):
    processor = SimpleProcessor()
    return processor(x)

# Create test data
test_input = torch.randn(2, 4)
test_output = torch.randn(2, 3)

ARGS = {
    "f": simple_function,
    "out": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
fourrier_from_data Not yet run
unverified (harvested, no recorded run)
StatisticDean/Crystal_Clear/crystal_clear/tensor_pipeline.py
code served (permissive licence) · get_code("ea510ef359b43c33")
How it ran

no invocation recorded.

get_batch_from_var Not yet run
unverified (harvested, no recorded run)
weepingwillowben/sound-eval/spectrogram_doc2vec.py
code served (permissive licence) · get_code("a3b1c9b3c1692bbf")
How it ran

no invocation recorded.

reconstruct Not yet run
unverified (harvested, no recorded run)
StatisticDean/Crystal_Clear/crystal_clear/misc.py
code served (permissive licence) · get_code("a33053bfb337d4fd")
How it ran

no invocation recorded.

spectrify_audios Not yet run
unverified (harvested, no recorded run)
weepingwillowben/sound-eval/spectrify.py
code served (permissive licence) · get_code("ea75c7dde8960478")
How it ran

no invocation recorded.

tf_spectrify Not yet run
unverified (harvested, no recorded run)
weepingwillowben/sound-eval/spectrify.py
code served (permissive licence) · get_code("a3ce6f94d6f60428")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 7 are also present in other papers whose code we hold (7 with at least one paper dated earlier, 0 the same month, 0 later) and 7 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
batch_filenames 1 other paper
1 dated earlier
1511.06811 (2015-11)
Learning visual groups from co-occurrences in space and time
flatten_audios 1 other paper
1 dated earlier
1511.06811 (2015-11)
Learning visual groups from co-occurrences in space and time
get_all_paths 1 other paper
1 dated earlier
1511.06811 (2015-11)
Learning visual groups from co-occurrences in space and time
get_batch_from_var 1 other paper
1 dated earlier
1511.06811 (2015-11)
Learning visual groups from co-occurrences in space and time
read_file 1 other paper
1 dated earlier
1511.06811 (2015-11)
Learning visual groups from co-occurrences in space and time
spectrify_audios 1 other paper
1 dated earlier
1511.06811 (2015-11)
Learning visual groups from co-occurrences in space and time
tf_spectrify 1 other paper
1 dated earlier
1511.06811 (2015-11)
Learning visual groups from co-occurrences in space and time

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We introduce the Free Music Archive (FMA), an open and easily accessible dataset suitable for evaluating several tasks in MIR, a field concerned with browsing, searching, and organizing large music collections. The community's growing interest in feature and end-to-end learning is however restrained by the limited availability of large audio datasets. The FMA aims to overcome this hurdle by providing 917 GiB and 343 days of Creative Commons-licensed audio from 106,574 tracks from 16,341 artists and 14,854 albums, arranged in a hierarchical taxonomy of 161 genres. It provides full-length and high-quality audio, pre-computed features, together with track- and user-level metadata, tags, and free-form text such as biographies. We here describe the dataset and how it was created, propose a train/validation/test split and three subsets, discuss some suitable MIR tasks, and evaluate some baselines for genre recognition. Code, data, and usage examples are available at https://github.com/mdeff/fma

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("1612.01840")
get_code_for_paper("1612.01840")
have("1612.01840")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/1612.01840.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/1612.01840.svg)](https://syntology.ai/paper/1612.01840)

Connect an agent — have() is free.