SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1904.02882 · 2019

LibriTTS: A Corpus Derived from LibriSpeech for Text-to-Speech

arXiv · PDF · Open in the Atlas

Code that ran

14 of 17 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 13 have no contract verdict; 1 failure of our instrument; 3 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/1904.02882.json · badge: /api/ran/1904.02882.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
zooeygirl/waveletProsody community (classifier label)
classifier basis not recorded on the link
5 ran · 0 no recorded run
skakouros/bert-prosody community (classifier label)
classifier basis not recorded on the link
4 ran · 1 no recorded run
mamachengcheng/prosody community (classifier label)
classifier basis not recorded on the link
3 ran · 1 no recorded run
Helsinki-NLP/prosody community (classifier label)
classifier basis not recorded on the link
2 ran · 1 no recorded run
FunctionStatusWhere it lives
chunk Ran
ran on a synthesized input
skakouros/bert-prosody/convert_dataset.py
code served (permissive licence) · get_code("f50aeb5281088ff3")
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 torch
import numpy as np
from typing import List, Iterable

torch.manual_seed(0)

# The function takes an iterable and chunks it into tuples of size 'size'
# It returns an iterable (list of tuples)

# Create a simple iterable - a list of integers
test_iterable = [1, 2, 3, 4, 5, 6, 7, 8, 9]

# Define chunk size
chunk_size = 3

ARGS = {
    "it": test_iterable,
    "size": chunk_size
}
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, 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
cosine_sim Ran
ran; our contract draft was wrong, not the code
skakouros/bert-prosody/losses.py
code served (permissive licence) · get_code("55b7eacdfa96a1c5")
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 (2, 3) 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 torch

torch.manual_seed(0)

# Create two small tensors for cosine similarity computation
# x1: shape (2, 4) - 2 samples, 4 dimensions
# x2: shape (3, 4) - 3 samples, 4 dimensions
x1 = torch.randn(2, 4)
x2 = torch.randn(3, 4)

ARGS = {
    "x1": x1,
    "x2": x2,
    "dim": 1,
    "eps": 1e-08
}
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 Tensor, ndim 2, dtype family float, shape (2, 3) — shape (2, 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
generate_square_subsequent_mask Ran
ran on a synthesized input
zooeygirl/waveletProsody/transformer.py
code served (permissive licence) · get_code("934e64a68c92b447")
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 (4, 4) 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 torch
from torch import nn, Tensor

torch.manual_seed(0)

def generate_square_subsequent_mask(sz: int) -> Tensor:
    return torch.triu(torch.ones(sz, sz) * float('-inf'), diagonal=2)

ARGS = {
    "sz": 4
}
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 Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 4) 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
load_dataset Ran
ran on a synthesized input
Helsinki-NLP/prosody/prosody_dataset.py
code served (permissive licence) · get_code("69e2f9f2018bfea5")
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 random
import tempfile
import os

class Config:
    def __init__(self):
        self.train_set = 'train'
        self.datadir = None  # Will be set to temp directory
        self.fraction_of_train_data = 1.0
        self.nclasses = 2
        self.shuffle_sentences = False
        self.sorted_batches = False

# Create temporary directory with test data files
temp_dir = tempfile.mkdtemp()

# Create train.txt
train_content = """<file>\tNA\tNA\tNA\tNA
word1\t0\t0\t1.5\t0
word2\t1\t1\t3.2\t1
<file>\tNA\tNA\tNA\tNA
word3\t0\t0\t2.1\t0
"""

# Create dev.txt
dev_content = """<file>\tNA\tNA\tNA\tNA
word4\t1\t1\t4.5\t1
word5\t0\t0\t1.2\t0
"""

# Create test.txt
test_content = """<file>\tNA\tNA\tNA\tNA
word6\t0\t0\t2.8\t0
word7\t1\t1\t5.1\t1
"""

with open(os.path.join(temp_dir, 'train.txt'), 'w') as f:
    f.write(train_content)

with open(os.path.join(temp_dir, 'dev.txt'), 'w') as f:
    f.write(dev_content)

with open(os.path.join(temp_dir, 'test.txt'), 'w') as f:
    f.write(test_content)

config = Config()
config.datadir = temp_dir

ARGS = {
    'config': config
}
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
load_dataset Ran
ran on a synthesized input
skakouros/bert-prosody/prosody_dataset.py
code served (permissive licence) · get_code("509c06b2e68c301a")
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 random
from pathlib import Path
from unittest.mock import Mock, mock_open, patch
import tempfile
import os

# Create a minimal config class that the function expects
class Config:
    def __init__(self):
        self.datadir = tempfile.gettempdir()
        self.split_dir = 'test_split'
        self.train_set = 'train'
        self.fraction_of_train_data = 1.0
        self.use_pos = False
        self.nclasses = 2
        self.shuffle_sentences = False
        self.sorted_batches = False

# Create temporary directory structure and files
config = Config()
split_dir_path = Path(config.datadir) / config.split_dir
split_dir_path.mkdir(exist_ok=True)

# Create minimal data files for train, dev, test
sample_data = """<file>\tNA\tNA\tNA\tNA\tNA
word1\t0\t0\t0.5\t0.5\tADJ
word2\t1\t1\t3.0\t1.0\tNOUN
<file>\tNA\tNA\tNA\tNA\tNA
word3\t0\t0\t0.2\t0.5\tVERB
word4\t1\t1\t5.0\t1.0\tADJ
"""

# Write files
for split_name in ['train', 'dev', 'test']:
    filename = config.train_set if split_name == 'train' else split_name
    filepath = Path(config.datadir).joinpath(config.split_dir, filename).with_suffix('.txt')
    with open(filepath, 'w') as f:
        f.write(sample_data)

ARGS = {
    'config': config
}
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
load_dataset Ran
ran on a synthesized input
zooeygirl/waveletProsody/prosody_dataset.py
code served (permissive licence) · get_code("7cb334ebfb9ae647")
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 random
import tempfile
import os

class Config:
    def __init__(self):
        self.train_set = 'train'
        self.test_set = 'test'
        self.datadir = None
        self.fraction_of_train_data = 1.0
        self.nclasses = 2
        self.shuffle_sentences = True
        self.sorted_batches = False

# Create a temporary directory with test data files
temp_dir = tempfile.mkdtemp()

# Create train.txt
train_content = """<file>\tbook1/spk1/utt1
word1\t1\tB\t0.5\t0.1
word2\t1\tI\t0.3\t0.2
<file>\tbook1/spk1/utt2
word3\t2\tB\t1.5\t0.3
word4\t1\tI\t0.2\t0.1
"""

# Create dev.txt
dev_content = """<file>\tbook1/spk1/utt3
word5\t1\tB\t0.4\t0.1
word6\t1\tI\t0.6\t0.2
"""

# Create test.txt
test_content = """<file>\tbook1/spk1/utt4
word7\t2\tB\t1.2\t0.3
word8\t1\tI\t0.5\t0.1
"""

with open(os.path.join(temp_dir, 'train.txt'), 'w') as f:
    f.write(train_content)

with open(os.path.join(temp_dir, 'dev.txt'), 'w') as f:
    f.write(dev_content)

with open(os.path.join(temp_dir, 'test.txt'), 'w') as f:
    f.write(test_content)

config = Config()
config.datadir = temp_dir

ARGS = {
    'config': config
}
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
load_embeddings Ran
ran on a synthesized input
Helsinki-NLP/prosody/prosody_dataset.py
code served (permissive licence) · get_code("c3b2065399bb567e")
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
import tempfile
import os

# Create a temporary embedding file
temp_dir = tempfile.mkdtemp()
embedding_file = os.path.join(temp_dir, "embeddings.txt")

# Write a small embedding file with 2 words, each with 300 dimensions
with open(embedding_file, 'w', encoding='utf8') as f:
    # Write "hello" with 300 floats
    hello_vec = ' '.join([str(0.1 * i) for i in range(300)])
    f.write(f"hello {hello_vec}\n")
    # Write "world" with 300 floats
    world_vec = ' '.join([str(0.2 * i) for i in range(300)])
    f.write(f"world {world_vec}\n")

# Create a config object with embedding_file attribute
class Config:
    def __init__(self, embedding_file):
        self.embedding_file = embedding_file

config = Config(embedding_file)

# Create a vocab as a set of words
vocab = {'hello', 'world', 'test'}

ARGS = {
    'config': config,
    'vocab': vocab,
}
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
load_embeddings Ran
ran on a synthesized input
mamachengcheng/prosody/prosody_dataset.py
code served (permissive licence) · get_code("b6f866c5e91fb701")
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
import tempfile
import os

class Config:
    def __init__(self, embedding_file):
        self.embedding_file = embedding_file

class Vocab:
    def __init__(self, words):
        self.words = list(words)
    
    def __iter__(self):
        return iter(self.words)
    
    def __len__(self):
        return len(self.words)
    
    def add(self, word):
        if word not in self.words:
            self.words.append(word)

# Create a temporary embedding file
embedding_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt', encoding='utf8')
embedding_file.write("hello " + " ".join([str(x) for x in np.random.normal(0, 1, 300)]) + "\n")
embedding_file.write("world " + " ".join([str(x) for x in np.random.normal(0, 1, 300)]) + "\n")
embedding_file.close()

vocab = Vocab(['hello', 'world', 'test'])
config = Config(embedding_file.name)

ARGS = {
    'config': config,
    'vocab': vocab
}

# Cleanup will happen after function execution
import atexit
atexit.register(lambda: os.unlink(embedding_file.name))
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
pad Ran
ran on a synthesized input
mamachengcheng/prosody/prosody_dataset.py
code served (permissive licence) · get_code("2eb70a3bb02b8fcb")
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 torch
import numpy as np

torch.manual_seed(0)

# The function expects batch to be a list of samples
# Each sample is a tuple with at least 8 elements:
# [0] words
# [1] word_ids (will be padded with 0)
# [2] is_main_piece
# [3] tags
# [4] labels (will be padded with 0)
# [5] seqlens
# [6] values (will be padded with invalid_set_to)
# [7] invalid_set_to value

sample1 = (
    ["hello", "world"],                    # words (index 0)
    [1, 2],                                # word_ids (index 1)
    [True, True],                          # is_main_piece (index 2)
    ["NOUN", "VERB"],                      # tags (index 3)
    [0, 1],                                # labels (index 4)
    2,                                     # seqlen (index 5)
    [0.5, 0.7],                            # values (index 6)
    -1                                     # invalid_set_to (index 7)
)

sample2 = (
    ["foo"],                               # words (index 0)
    [3],                                   # word_ids (index 1)
    [True],                                # is_main_piece (index 2)
    ["ADJ"],                               # tags (index 3)
    [1],                                   # labels (index 4)
    1,                                     # seqlen (index 5)
    [0.9],                                 # values (index 6)
    -1                                     # invalid_set_to (index 7)
)

batch = [sample1, sample2]

ARGS = {
    "batch": batch
}
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 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
pad Ran
ran on a synthesized input
mamachengcheng/prosody/process.py
code served (permissive licence) · get_code("82f5c2bcab653812")
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 torch
import numpy as np

torch.manual_seed(0)

# The function expects a batch, which is a list of samples
# Each sample has structure: (sent, tag, x, y, seqlen)
# where x and y are lists that will be padded

batch = [
    (
        "sent1",           # sample[0]: sents
        "tag1",            # sample[1]: tags
        [1, 2],            # sample[2]: x (list to be padded)
        [3, 4],            # sample[3]: y (list to be padded)
        2                  # sample[4]: seqlen
    ),
    (
        "sent2",           # sample[0]: sents
        "tag2",            # sample[1]: tags
        [5, 6, 7],         # sample[2]: x (list to be padded)
        [8, 9, 10],        # sample[3]: y (list to be padded)
        3                  # sample[4]: seqlen
    )
]

ARGS = {
    "batch": batch
}
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 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
pad Ran
ran on a synthesized input
zooeygirl/waveletProsody/prosody_dataset.py
code served (permissive licence) · get_code("3bc0d284301c0a82")
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 torch
import numpy as np

torch.manual_seed(0)

# The function expects a batch where each sample is a tuple/list with at least 15 elements (indices 0-14)
# Let's construct samples based on what the function accesses:
# 0: words (list)
# 1: features (list, gets padded)
# 2: is_main_piece (list)
# 3: tags (list)
# 4: labels (list, gets padded)
# 5: seqlens (int)
# 6: values (list, gets padded with invalid_set_to)
# 7: invalid_set_to marker (used as padding value)
# 8: prevSeq (list)
# 9: prvSeqLen (int)
# 10: pTags (list)
# 13: prevWords (list)
# 14: file_id (int or identifier)

sample1 = [
    ["word1", "word2"],           # 0: words
    [1, 2, 3],                     # 1: features (shorter)
    [1, 1],                        # 2: is_main_piece
    ["tag1", "tag2"],              # 3: tags
    [0, 1],                        # 4: labels (shorter)
    2,                             # 5: seqlens
    [0.5, 0.6, 0.7],              # 6: values (shorter)
    -100,                          # 7: invalid_set_to
    ["prev1"],                     # 8: prevSeq
    1,                             # 9: prvSeqLen
    ["ptag1", "ptag2"],            # 10: pTags
    None,                          # 11: unused
    None,                          # 12: unused
    ["pword1"],                    # 13: prevWords
    0,                             # 14: file_id
]

sample2 = [
    ["word3", "word4", "word5"],   # 0: words
    [4, 5],                        # 1: features (shorter)
    [1, 0, 1],                     # 2: is_main_piece
    ["tag3", "tag4", "tag5"],      # 3: tags
    [1, 0, 1],                     # 4: labels
    3,                             # 5: seqlens
    [0.1, 0.2],                    # 6: values (shorter)
    -100,                          # 7: invalid_set_to
    ["prev2", "prev3"],            # 8: prevSeq
    2,                             # 9: prvSeqLen
    ["ptag3", "ptag4", "ptag5"],   # 10: pTags
    None,                          # 11: unused
    None,                          # 12: unused
    ["pword2", "pword3"],          # 13: prevWords
    1,                             # 14: file_id
]

batch = [sample1, sample2]

ARGS = {
    "batch": batch
}
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 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
rediscretize_tag Ran
ran on a synthesized input
zooeygirl/waveletProsody/prosody_dataset_no_config.py
code served (permissive licence) · get_code("cfd81aa27089ed35")
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 torch
import numpy as np

torch.manual_seed(0)

ARGS = {
    'value_prominance': 3.5,
    'nclasses': 10
}
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 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
split_punctuation Ran
ran on a synthesized input
skakouros/bert-prosody/convert_dataset.py
code served (permissive licence) · get_code("b1a0158abfab316c")
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 2, 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 numpy as np
from typing import List

# Construct valid arguments for split_punctuation
utterance = ["hello,", "world", "how", "are."]
prominence = ["1", "2", "1", "0"]
duration = ["0.5", "0.3", "0.4", "0.2"]
add_eos_mark = False

ARGS = {
    "utterance": utterance,
    "prominence": prominence,
    "duration": duration,
    "add_eos_mark": add_eos_mark,
}
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 tuple, ndim 2, 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
weighted_mse_loss Ran
ran on a synthesized input
zooeygirl/waveletProsody/main2.py
code served (permissive licence) · get_code("3ef7a35a2729d0a8")
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 0, dtype family float, shape () 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

torch.manual_seed(0)

# Create small input and target tensors
# The function expects tensors that support .device, arithmetic operations, and .cuda()
input_tensor = torch.randn(2, 4)
target_tensor = torch.randn(2, 4)

ARGS = {
    "input": input_tensor,
    "target": target_tensor,
}
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 Tensor, ndim 0, dtype family float, shape () — shape () 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
load_dataset Not yet run
unverified (harvested, no recorded run)
mamachengcheng/prosody/prosody_dataset.py
code served (permissive licence) · get_code("aab5d0a9d341d750")
How it ran

no invocation recorded.

pad Not yet run
unverified (harvested, no recorded run)
Helsinki-NLP/prosody/prosody_dataset.py
code served (permissive licence) · get_code("6ba17c97a01e6a49")
How it ran

no invocation recorded.

pad Not yet run
unverified (harvested, no recorded run)
skakouros/bert-prosody/prosody_dataset.py
code served (permissive licence) · get_code("ef8ae41c68caaf41")
How it ran

no invocation recorded.

Also present in other papers' code

Of 17 samples lifted from this paper's linked repositories, 17 carry a function identity and 0 do not. Of the 15 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 14 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
cosine_sim 2 other papers
2 later
aaai_26699 (2023 (year only))
MixFairFace: Towards Ultimate Fairness via MixFair Adapter in Face Recognition
2404.13788 (2024-04)
AnyPattern: Towards In-context Image Copy Detection

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

This paper introduces a new speech corpus called "LibriTTS" designed for text-to-speech use. It is derived from the original audio and text materials of the LibriSpeech corpus, which has been used for training and evaluating automatic speech recognition systems. The new corpus inherits desired properties of the LibriSpeech corpus while addressing a number of issues which make LibriSpeech less than ideal for text-to-speech work. The released corpus consists of 585 hours of speech data at 24kHz sampling rate from 2,456 speakers and the corresponding texts. Experimental results show that neural end-to-end TTS models trained from the LibriTTS corpus achieved above 4.0 in mean opinion scores in naturalness in five out of six evaluation speakers. The corpus is freely available for download from http://www.openslr.org/60/.

For agents

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

get_harvested_code_for_paper("1904.02882")
get_code_for_paper("1904.02882")
have("1904.02882")

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

curl https://syntology.ai/api/ran/1904.02882.json

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

[![Syntology run record](https://syntology.ai/api/ran/1904.02882.svg)](https://syntology.ai/paper/1904.02882)

Connect an agent — have() is free.