SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.09871 · IJCAI · 2024

MuChin: A Chinese Colloquial Description Benchmark for Evaluating Language Models in the Field of Music

Yan Liu, Qi Wang, Zihao Wang, Tao Zhang, Pengfei Yu, Kejun Zhang, Shuyu Li, Jinyang Luo, Ming Xi

arXiv · PDF · Open in the Atlas

Code that ran

6 of 7 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; 2 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 4 failures of our instrument; 1 no recorded run), measured 2026-11-22; 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/2402.09871.json · badge: /api/ran/2402.09871.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
CarlWangChina/MuChin canonical (classifier label)
classifier basis: census
6 ran · 1 no recorded run
FunctionStatusWhere it lives
F1_score Ran
ran; our contract draft was wrong, not the code
CarlWangChina/MuChin/und_eval/models/mlp.py
code served (permissive licence) · get_code("db3261a7e3b7c9c3")
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, 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 torch

torch.manual_seed(0)

# Create small integer tensors for predictions and ground truth
# Using shape (2, 4) for clarity - batch of 2, 4 elements each
data1 = torch.tensor([[1, 0, 1, 0], [1, 1, 0, 1]], dtype=torch.int32)
data2 = torch.tensor([[1, 0, 1, 0], [1, 1, 0, 0]], dtype=torch.int32)

ARGS = {
    "data1": data1,
    "data2": data2,
}
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 tuple, ndim 1, dtype family float — a tuple 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
calculate_accuracy_and_recall Ran
ran, honoured the contract we drafted
CarlWangChina/MuChin/und_eval/models/modeltest.py
code served (permissive licence) · get_code("9036a848759857a8")
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, 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 torch
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.fc = nn.Linear(4, 2)
    
    def forward(self, x):
        return self.fc(x)

model = SimpleModel()

class SimpleDataset:
    def __init__(self):
        self.data = [
            (torch.randn(4), 0),
            (torch.randn(4), 1),
            (torch.randn(4), 0),
            (torch.randn(4), 1),
        ]
    
    def __len__(self):
        return len(self.data)
    
    def __getitem__(self, idx):
        return self.data[idx]

dataset = SimpleDataset()

ARGS = {
    "model": model,
    "dataset": dataset,
}
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 tuple, ndim 1, dtype family float — a tuple 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
find_indices Ran
ran, honoured the contract we drafted
CarlWangChina/MuChin/gen_eval/glrc_obj_eval.py
code served (permissive licence) · get_code("e9dc9d5aa138ee8c")
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 2, dtype family int; 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 re

def find_indices(pattern: str, text: str):
    """Return list of tuples, with matched start and end index for each."""
    matches = re.finditer(pattern, text)
    return [(m.start(), m.end()) for m in matches]


ARGS = {
    "pattern": r"\d+",
    "text": "The numbers are 42 and 123 and 7 here."
}
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 2, dtype family int — 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
find_last_zh_word Ran
ran; our contract draft was wrong, not the code
CarlWangChina/MuChin/gen_eval/glrc_obj_eval.py
code served (permissive licence) · get_code("5ca8664cbef23bbc")
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, 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.

ARGS = {
    "line": "hello世界你好鿿test一二三"
}
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 tuple, ndim 1, dtype family float — a tuple 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
mapping_rhythm_name Ran
ran; our contract draft was wrong, not the code
CarlWangChina/MuChin/gen_eval/glrc_obj_eval.py
code served (permissive licence) · get_code("3a94e5e9e04c5a5e")
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 re

RHYTHM_TABLE = {'ㄚ': '01M', 'ㄛ': '02B', 'ㄜ': '03G', 'ㄝ': '04J', 'ㄓ': '05Z', 'ㄔ': '05Z', 'ㄕ': '05Z', 'ㄖ': '05Z', 'ㄗ': '05Z', 'ㄘ': '05Z', 'ㄙ': '05Z', 'ㄦ': '06E', 'ㄧ': '07Q', 'ㄟ': '08W', 'ㄞ': '09K', 'ㄨ': '10G', 'ㄩ': '11Y', 'ㄡ': '12H', 'ㄠ': '13H', 'ㄢ': '14H', 'ㄣ': '15H', 'ㄤ': '16T', 'ㄥ': '17G', 'ㄧㄥ': '17G', 'ㄨㄥ': '17G', '+ㄨㄥ': '18D', 'ㄩㄥ': '18D'}

def mapping_rhythm_name(w_zhuyin: str):
    """Return value mapped from `RHYTHM_TABLE`, or word zhuyin without tone
    from the passed-in `w_zhuyin` if no any value can be mapped, e.g., empty
    `w_zhuyin` will get empty."""
    zy_notone = re.sub('[˙ˊˇˋ]', '', w_zhuyin)
    rhythm = zy_notone[-1:]
    rtname = RHYTHM_TABLE.get(rhythm)
    if not rtname:
        return rhythm
    if rhythm != 'ㄥ':
        return rtname
    rhythm = zy_notone[-2:]
    if 'ㄩㄥ' == rhythm:
        return RHYTHM_TABLE.get(rhythm, '18D')
    elif 'ㄧㄥ' == rhythm:
        return RHYTHM_TABLE.get(rhythm, '17G')
    elif 'ㄨㄥ' == rhythm:
        return RHYTHM_TABLE.get(rhythm, '17G') if rhythm == zy_notone else RHYTHM_TABLE.get('+' + rhythm, '18D')
    else:
        return RHYTHM_TABLE.get(rhythm[-1:], '17G')

ARGS = {
    'w_zhuyin': 'ㄓㄨˋ'
}
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
run_cmd_sync Ran
ran; our contract draft was wrong, not the code
CarlWangChina/MuChin/und_eval/dataset.py
code served (permissive licence) · get_code("ea4c62034920a7b7")
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, 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 pathlib
import tempfile
import os

# Define ARGS dict with valid arguments for run_cmd_sync
ARGS = {
    "cmd": "echo 'Hello World'",
    "cwd": None,
    "interactive": False,
    "timeout": None,
}
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 tuple, ndim 1, dtype family float — a tuple 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
get_approximate_audio_length Not yet run
unverified (harvested, no recorded run)
CarlWangChina/MuChin/und_eval/dataset.py
code served (permissive licence) · get_code("2f347415ba83a44a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 6 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
get_approximate_audio_length 1 other paper
1 later
2025.findings-acl.360 (2025 (year only))
Generative Music Models' Alignment with Professional and Amateur Users' Expectations

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

The rapidly evolving multimodal Large Language Models (LLMs) urgently require new benchmarks to uniformly evaluate their performance on understanding and textually describing music. However, due to semantic gaps between Music Information Retrieval (MIR) algorithms and human understanding, discrepancies between professionals and the public, and low precision of annotations, existing music description datasets cannot serve as benchmarks. To this end, we present MuChin, the first open-source music description benchmark in Chinese colloquial language, designed to evaluate the performance of multimodal LLMs in understanding and describing music. We established the Caichong Music Annotation Platform (CaiMAP) that employs an innovative multi-person, multi-stage assurance method, and recruited both amateurs and professionals to ensure the precision of annotations and alignment with popular semantics. Utilizing this method, we built a dataset with multidimensional, high-precision music annotations, the Caichong Music Dataset (CaiMD), and carefully selected 1,000 high-quality entries to serve as the test set for MuChin. Based on MuChin, we analyzed the discrepancies between professionals and amateurs in terms of music description, and empirically demonstrated the effectiveness of annotated data for fine-tuning LLMs. Ultimately, we employed MuChin to evaluate existing music understanding models on their ability to provide colloquial descriptions of music.

For agents

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

get_harvested_code_for_paper("2402.09871")
get_code_for_paper("2402.09871")
have("2402.09871")

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

curl https://syntology.ai/api/ran/2402.09871.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.09871.svg)](https://syntology.ai/paper/2402.09871)

Connect an agent — have() is free.