SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2005.09921 · 2020

End-to-End Speaker Diarization for an Unknown Number of Speakers with Encoder-Decoder Based Attractors

arXiv · PDF · Open in the Atlas

Code that ran

19 of 21 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 19 have no contract verdict; 0 failures of our instrument; 2 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/2005.09921.json · badge: /api/ran/2005.09921.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
butspeechfit/diacorrect community (classifier label)
classifier basis not recorded on the link
10 ran · 2 no recorded run
Yoshani/pytorch-sa-eend community (classifier label)
classifier basis not recorded on the link
9 ran · 0 no recorded run
FunctionStatusWhere it lives
batch_pit_loss Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/pytorch_backend/loss.py
code served (permissive licence) · get_code("89d4f325ff9c071b")
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 torch.nn.functional as F
import numpy as np
from itertools import permutations

torch.manual_seed(0)

# Create small sample tensors for batch_pit_loss
# ys: B-length list of predictions, each (T, C) shaped pre-activation values
# ts: B-length list of labels, each (T, C) shaped labels in {0, 1}

# Batch size = 2
# Time dimension T = 5 (distinct from C)
# Channel dimension C = 3 (distinct from T)

T = 5
C = 3
B = 2

# Create predictions (pre-activation logits)
ys = [
    torch.randn(T, C),  # First sample: (5, 3)
    torch.randn(T, C),  # Second sample: (5, 3)
]

# Create labels (binary values in {0, 1})
ts = [
    torch.randint(0, 2, (T, C)).float(),  # First sample: (5, 3)
    torch.randint(0, 2, (T, C)).float(),  # Second sample: (5, 3)
]

label_delay = 0

ARGS = {
    "ys": ys,
    "ts": ts,
    "label_delay": label_delay,
}
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
calc_diarization_error Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/pytorch_backend/loss.py
code served (permissive licence) · get_code("977833339e01af8c")
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 dict. 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

torch.manual_seed(0)

# Create small tensors with distinct dimensions
# T=5 time steps, C=3 channels
T = 5
C = 3

# pred: (T, C) pre-activation values
pred = torch.randn(T, C)

# label: (T, C) binary labels in {0, 1}
label = torch.randint(0, 2, (T, C)).float()

# label_delay: offset parameter
label_delay = 1

ARGS = {
    'pred': pred,
    'label': label,
    'label_delay': label_delay,
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
get_input_dim Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/feature.py
code served (permissive licence) · get_code("2f299371003065aa")
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 int, ndim 0, dtype family int, shape () 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
import torch.nn as nn
import numpy as np

torch.manual_seed(0)

# Analyze the function:
# - frame_size: int, used in bit_length() and arithmetic
# - context_size: int, used in arithmetic
# - transform_type: str, used with startswith()

# Construct valid arguments
frame_size = 8
context_size = 5
transform_type = 'logmel23'

ARGS = {
    'frame_size': frame_size,
    'context_size': context_size,
    'transform_type': transform_type
}
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 int, ndim 0, dtype family int, 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_sap Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/common_utils/dataset_correct.py
code served (permissive licence) · get_code("13bd5877d13d5e3e")
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 2, dtype family float, shape (2, 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 h5py
import tempfile
import os
import numpy as np

# Create a temporary HDF5 file with the required 'T_hat' dataset
temp_dir = tempfile.mkdtemp()
temp_filepath = os.path.join(temp_dir, "test_data.h5")

# Create a sample HDF5 file
with h5py.File(temp_filepath, 'w') as f:
    # Create a small sample tensor data
    sample_data = np.random.rand(2, 4).astype(np.float32)
    f.create_dataset('T_hat', data=sample_data)

ARGS = {
    'filepath': temp_filepath
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
h5py, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (2, 4) — shape (2, 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_segments Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/kaldi_data.py
code served (permissive licence) · get_code("216dff0f296d027f")
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 1, dtype family float, shape (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 os
import numpy as np
import tempfile

# Create a temporary segments file with valid data
temp_dir = tempfile.mkdtemp()
segments_file = os.path.join(temp_dir, "segments.txt")

# Write sample segment data in the format expected by loadtxt
with open(segments_file, 'w') as f:
    f.write("utt_001 rec_001 0.5 1.5\n")
    f.write("utt_002 rec_002 2.0 3.5\n")

ARGS = {
    'segments_file': segments_file
}
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 1, dtype family float, shape (2,) — shape (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
load_segments_hash Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/kaldi_data.py
code served (permissive licence) · get_code("805c93ef8140e802")
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 dict. 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 os
import tempfile

# Create a temporary segments file
temp_dir = tempfile.mkdtemp()
segments_file = os.path.join(temp_dir, "segments.txt")

# Write sample segment data
with open(segments_file, 'w') as f:
    f.write("utt1 rec1 0.0 1.5\n")
    f.write("utt2 rec1 1.5 3.2\n")
    f.write("utt3 rec2 0.5 2.1\n")

ARGS = {
    "segments_file": segments_file
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
load_segments_rechash Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/kaldi_data.py
code served (permissive licence) · get_code("11ff0f39a6fdcd1e")
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 dict. 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 os
import tempfile

# Create a temporary segments file with test data
with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt') as f:
    f.write("utt1 rec1 0.0 1.5\n")
    f.write("utt2 rec1 1.5 3.0\n")
    f.write("utt3 rec2 0.5 2.0\n")
    segments_file_path = f.name

ARGS = {
    "segments_file": segments_file_path
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
load_segments_rechash Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/common_utils/kaldi_data.py
code served (permissive licence) · get_code("8972f961b497d9da")
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 dict. 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 os
import tempfile

# Create a temporary segments file for testing
temp_dir = tempfile.mkdtemp()
segments_file = os.path.join(temp_dir, "segments.txt")

# Write sample segment data to the file
with open(segments_file, 'w') as f:
    f.write("utt1 rec1 0.0 1.5\n")
    f.write("utt2 rec1 1.5 3.0\n")
    f.write("utt3 rec2 0.5 2.0\n")

ARGS = {
    "segments_file": segments_file
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
load_wav_scp Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/common_utils/dataset_correct.py
code served (permissive licence) · get_code("fcdd1d00334209b8")
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 dict. 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 wav.scp file with valid content
temp_dir = tempfile.mkdtemp()
wav_scp_path = os.path.join(temp_dir, "wav.scp")

# Write sample wav.scp content
# Format: recording_id /path/to/wav/file
with open(wav_scp_path, 'w') as f:
    f.write("rec_001 /path/to/audio/file1.wav\n")
    f.write("rec_002 /path/to/audio/file2.wav\n")
    f.write("rec_003 /path/to/audio/file3.wav\n")

ARGS = {
    "wav_scp_file": wav_scp_path
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
load_wav_scp Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/common_utils/kaldi_data.py
code served (permissive licence) · get_code("4e95cba0484da8ff")
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 dict. 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 wav.scp file for testing
temp_dir = tempfile.mkdtemp()
wav_scp_path = os.path.join(temp_dir, "wav.scp")

# Write sample content to the wav.scp file
# Format: rec_id wav_rxfilename
with open(wav_scp_path, 'w') as f:
    f.write("rec001 /path/to/audio/file001.wav\n")
    f.write("rec002 /path/to/audio/file002.wav\n")
    f.write("rec003 /path/to/audio/file003.wav\n")

ARGS = {
    "wav_scp_file": wav_scp_path
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
my_collate Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/pytorch_backend/diarization_dataset.py
code served (permissive licence) · get_code("36210d944a56c3e1")
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

torch.manual_seed(0)

# The function my_collate expects a batch parameter
# which is an iterable of tuples, where each tuple contains (data, target)
# The function unpacks these using zip(*batch) and returns [data, target]

# Create a small batch with 2 samples
# Each sample is a tuple of (data, target)
batch = [
    (torch.randn(4), torch.tensor(0)),
    (torch.randn(4), torch.tensor(1))
]

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 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
osd_loss Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/backend/losses.py
code served (permissive licence) · get_code("c8758ad59f4753a7")
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. 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.functional as F

torch.manual_seed(0)

# ys: logits tensor, shape (batch, sequence, num_classes)
# ts: targets tensor, shape (batch, sequence, num_classes)
# The function uses these to compute OSD loss

batch_size = 2
sequence_length = 3
num_classes = 4

# ys should be logits (unbounded), used with sigmoid
ys = torch.randn(batch_size, sequence_length, num_classes)

# ts should be target labels, can contain -1 (ignored), 0, 1, etc.
# The function checks for values < 0 and uses them as ignore masks
ts = torch.tensor([
    [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0]],
    [[0, 0, 0, 1], [-1, -1, -1, -1], [1, 1, 0, 0]]
], dtype=torch.float32)

ARGS = {
    "ys": ys,
    "ts": ts,
}
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
pit_loss Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/pytorch_backend/loss.py
code served (permissive licence) · get_code("6c112a64bf8d5421")
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 torch.nn.functional as F

torch.manual_seed(0)

# pred: (T, C)-shaped pre-activation values
# T=10 (time steps), C=4 (number of channels/sources)
pred = torch.randn(10, 4)

# label: (T, C)-shaped labels in {0, 1}
# Same shape as pred
label = torch.randint(0, 2, (10, 4)).float()

# label_delay: integer delay parameter
label_delay = 0

ARGS = {
    "pred": pred,
    "label": label,
    "label_delay": label_delay,
}
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 — 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
postprocess_output Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/infer.py
code served (permissive licence) · get_code("380ee6404a7f71ea")
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 2, dtype family float, shape (20, 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 numpy as np
from scipy.signal import medfilt

# Create sample probability data
# Shape: (time_steps, num_speakers)
# Using 10 time steps and 3 speakers
probabilities = np.random.rand(10, 3)

# Subsampling factor
subsampling = 2

# Threshold for binarization
threshold = 0.5

# Median filter window length (must be odd)
median_window_length = 3

ARGS = {
    "probabilities": probabilities,
    "subsampling": subsampling,
    "threshold": threshold,
    "median_window_length": median_window_length,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (20, 3) — shape (20, 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
setup_optimizer Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/backend/updater.py
code served (permissive licence) · get_code("75294dcc89fb7683")
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 NoamOpt. 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
import torch.optim as optim
from types import SimpleNamespace

torch.manual_seed(0)

# Create a simple model with parameters
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

model = SimpleModel()

# Create args with different optimizer choices
args_adam = SimpleNamespace(
    optimizer='adam',
    lr=0.001,
    hidden_size=512,
    noam_warmup_steps=4000
)

args_sgd = SimpleNamespace(
    optimizer='sgd',
    lr=0.01,
    hidden_size=512,
    noam_warmup_steps=4000
)

args_noam = SimpleNamespace(
    optimizer='noam',
    lr=0.001,
    hidden_size=512,
    noam_warmup_steps=4000
)

ARGS = {
    'args': args_noam,
    'model': model
}
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 NoamOpt — a NoamOpt output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a NoamOpt 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
sigmoid_inverse Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/infer_sap.py
code served (permissive licence) · get_code("d9d88d2aeb7d7679")
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 1, dtype family float, shape (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

torch.manual_seed(0)

# sigmoid_inverse takes a single parameter 'x' which should be a tensor
# The function applies log((x + EPS) / (1 - x + EPS))
# This expects x to be in a valid range for the sigmoid inverse (roughly 0 to 1)
# Create a small tensor with values in the valid range (0, 1)

x = torch.tensor([0.2, 0.5, 0.7, 0.9], dtype=torch.float32)

ARGS = {
    "x": x
}
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 1, dtype family float, shape (4,) — shape (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
splice Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/common_utils/features.py
code served (permissive licence) · get_code("bdec7a99af6edff5")
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 2, dtype family float, shape (5, 15) 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

# Create a small feature array: (n_frames=5, n_featdim=3)
Y = np.random.randn(5, 3).astype(np.float32)

# Set context_size to a small value
context_size = 2

ARGS = {
    "Y": Y,
    "context_size": context_size,
}
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 2, dtype family float, shape (5, 15) — shape (5, 15) 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
subsample Ran
ran on a synthesized input
Yoshani/pytorch-sa-eend/eend/feature.py
code served (permissive licence) · get_code("08f6950bcdfafc26")
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)

# Y should be a tensor or array-like that supports slicing with [::]
# T should be a tensor or array-like that supports slicing with [::]
# subsampling is an integer that controls the step size

Y = torch.randn(8, 4)  # 8 frames, 4 features
T = torch.randn(8, 3)  # 8 frames, 3 different dimension

ARGS = {
    "Y": Y,
    "T": T,
    "subsampling": 2
}
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
vad_loss Ran
ran on a synthesized input
butspeechfit/diacorrect/diacorrect/backend/losses.py
code served (permissive licence) · get_code("c54d0aa33d0eb502")
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. 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)

# ys: logits tensor, shape (batch, time, features)
# ts: target tensor, shape (batch, time, features)
# The function performs operations like:
# - torch.sum on dimension 2
# - torch.sigmoid on ys
# - torch.prod on dimension 2
# - F.binary_cross_entropy
# - torch.max on dimension 2
# - torch.mean on dimension 2

batch_size = 2
time_steps = 4
features = 3

# Create ys: logits (can be any real values)
ys = torch.randn(batch_size, time_steps, features)

# Create ts: target tensor with values that are either -1 (padding) or non-negative
# Mix of -1 (padding) and valid values (0 or 1)
ts = torch.tensor([
    [[1, 0, 1], [1, 1, 0], [-1, -1, -1], [0, 1, 1]],
    [[1, 1, 1], [-1, 0, -1], [0, 0, 1], [1, 1, 0]]
], dtype=torch.float32)

ARGS = {
    "ys": ys,
    "ts": ts,
}
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
get_rate Not yet run
unverified (harvested, no recorded run)
butspeechfit/diacorrect/diacorrect/backend/updater.py
code served (permissive licence) · get_code("acb2c2991cd415dd")
How it ran

no invocation recorded.

pit_loss_multispk Not yet run
unverified (harvested, no recorded run)
butspeechfit/diacorrect/diacorrect/backend/losses.py
code served (permissive licence) · get_code("61a0fbfaadc3465b")
How it ran

no invocation recorded.

Also present in other papers' code

Of 21 samples lifted from this paper's linked repositories, 21 carry a function identity and 0 do not. Of the 21 distinct functions with an identity, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 4 later) and 16 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
load_segments 2 other papers
1 dated earlier, 1 later
1909.06247 (2019-09)
End-to-End Neural Speaker Diarization with Self-attention
2006.01796 (2020-06)
Neural Speaker Diarization with Speaker-Wise Chain Rule
load_segments_hash 2 other papers
1 dated earlier, 1 later
1909.06247 (2019-09)
End-to-End Neural Speaker Diarization with Self-attention
2006.01796 (2020-06)
Neural Speaker Diarization with Speaker-Wise Chain Rule
load_segments_rechash 2 other papers
1 dated earlier, 1 later
1909.06247 (2019-09)
End-to-End Neural Speaker Diarization with Self-attention
2006.01796 (2020-06)
Neural Speaker Diarization with Speaker-Wise Chain Rule
subsample 2 other papers
1 dated earlier, 1 later
1909.06247 (2019-09)
End-to-End Neural Speaker Diarization with Self-attention
2006.01796 (2020-06)
Neural Speaker Diarization with Speaker-Wise Chain Rule
get_input_dim 1 other paper
1 dated earlier
1909.06247 (2019-09)
End-to-End Neural Speaker Diarization with Self-attention

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

End-to-end speaker diarization for an unknown number of speakers is addressed in this paper. Recently proposed end-to-end speaker diarization outperformed conventional clustering-based speaker diarization, but it has one drawback: it is less flexible in terms of the number of speakers. This paper proposes a method for encoder-decoder based attractor calculation (EDA), which first generates a flexible number of attractors from a speech embedding sequence. Then, the generated multiple attractors are multiplied by the speech embedding sequence to produce the same number of speaker activities. The speech embedding sequence is extracted using the conventional self-attentive end-to-end neural speaker diarization (SA-EEND) network. In a two-speaker condition, our method achieved a 2.69 % diarization error rate (DER) on simulated mixtures and a 8.07 % DER on the two-speaker subset of CALLHOME, while vanilla SA-EEND attained 4.56 % and 9.54 %, respectively. In unknown numbers of speakers conditions, our method attained a 15.29 % DER on CALLHOME, while the x-vector-based clustering method achieved a 19.43 % DER.

For agents

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

get_harvested_code_for_paper("2005.09921")
get_code_for_paper("2005.09921")
have("2005.09921")

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

curl https://syntology.ai/api/ran/2005.09921.json

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

[![Syntology run record](https://syntology.ai/api/ran/2005.09921.svg)](https://syntology.ai/paper/2005.09921)

Connect an agent — have() is free.