SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1612.01744 · 2016

Listen and Translate: A Proof of Concept for End-to-End Speech-to-Text Translation

arXiv · PDF · Open in the Atlas

Code that ran

10 of 15 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 9 have no contract verdict; 1 failure of our instrument; 5 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1612.01744.json · badge: /api/ran/1612.01744.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
eske/seq2seq canonical (classifier label)
classifier basis: census
10 ran · 5 no recorded run
FunctionStatusWhere it lives
auto_reuse Ran
ran on a synthesized input
eske/seq2seq/translate/models.py
code served (permissive licence) · get_code("68d6e78c95902674")
How it ran

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

import tensorflow as tf

def dummy_function_with_reuse():
    """A simple function that uses variable_scope to demonstrate reuse behavior"""
    with tf.variable_scope("test_scope"):
        var = tf.get_variable("test_var", shape=[2, 3])
    return var

def sample_fun(x):
    """A function that might raise ValueError with 'reuse' in the message"""
    with tf.variable_scope("sample_scope"):
        var = tf.get_variable("sample_var", shape=[2, 3])
    return var + x

def another_fun(y, z=None):
    """Another sample function to wrap"""
    if z is None:
        z = 0
    return y * 2 + z

ARGS = {
    "fun": sample_fun
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_shape Ran
ran; our contract draft was wrong, not the code
eske/seq2seq/translate/beam_search.py
code served (permissive licence) · get_code("350b8500eef3645a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family 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 tensorflow as tf

# Create a small tensor with known shape
# Using batch size 2 and feature dimension 4
tensor = tf.constant([[1.0, 2.0, 3.0, 4.0],
                      [5.0, 6.0, 7.0, 8.0]])

ARGS = {
    "tensor": tensor
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family 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
get_state_size Ran
ran on a synthesized input
eske/seq2seq/translate/rnn.py
code served (permissive licence) · get_code("06352e12d02c8738")
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 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

torch.manual_seed(0)

ARGS = {
    'cell_type': 'lstm',
    'cell_size': 64,
    'proj_size': 32,
    'layers': 2
}
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 int — 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
initialize_vocabulary Ran
ran on a synthesized input
eske/seq2seq/translate/utils.py
code served (permissive licence) · get_code("b64547c9e827d468")
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 vocab. 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
from collections import namedtuple

# Create a temporary vocabulary file
temp_dir = tempfile.mkdtemp()
vocab_file = os.path.join(temp_dir, "vocab.txt")

# Write vocabulary file with sample words
with open(vocab_file, 'w') as f:
    f.write("dog\n")
    f.write("cat\n")
    f.write("bird\n")
    f.write("fish\n")

ARGS = {
    "vocabulary_path": vocab_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 vocab — a vocab 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
levenshtein Ran
ran on a synthesized input
eske/seq2seq/translate/evaluation.py
code served (permissive licence) · get_code("d8042102ebe27b4c")
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 random

# Set seed for reproducibility
random.seed(42)
np.random.seed(42)

# Define the arguments
ARGS = {
    "src": "kitten",
    "trg": "sitting",
    "sub_cost": 1.0,
    "del_cost": 1.0,
    "ins_cost": 1.0,
    "randomize": False
}
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
resize_like Ran
ran on a synthesized input
eske/seq2seq/translate/beam_search.py
code served (permissive licence) · get_code("86988a69e7729805")
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 EagerTensor, ndim 2, dtype family float, shape (6, 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 tensorflow as tf
import numpy as np

# Set random seed for reproducibility
tf.random.set_seed(0)

# Create small test tensors
# src: batch_size=2, feature_dim=4
src = tf.constant(np.random.randn(2, 4).astype(np.float32))

# dst: batch_size*beam_size=6 (so batch_size=2, beam_size=3), feature_dim=5
# This should have shape [batch_size * beam_size, ...other dims...]
dst = tf.constant(np.random.randn(6, 5).astype(np.float32))

ARGS = {
    "src": src,
    "dst": dst,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 2, dtype family float, shape (6, 4) — shape (6, 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
score_function_decorator Ran
ran on a synthesized input
eske/seq2seq/translate/evaluation.py
code served (permissive licence) · get_code("c2cc86fca14350af")
How it ran

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

import torch
import torch.nn as nn

torch.manual_seed(0)

# The function score_function_decorator takes a single parameter 'reversed'
# which is a boolean flag. It returns a decorator that attaches the 'reversed'
# attribute to a function.

ARGS = {
    "reversed": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
sentence_bleu Ran
ran on a synthesized input
eske/seq2seq/translate/evaluation.py
code served (permissive licence) · get_code("fb51636d55302b3b")
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 float, 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 math
from collections import Counter, OrderedDict

# Define ARGS with valid arguments for sentence_bleu function
ARGS = {
    "hypothesis": ["the", "cat", "sat", "on", "the", "mat"],
    "reference": ["the", "cat", "is", "sitting", "on", "the", "mat"],
    "smoothing": True,
    "order": 4
}
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 float, 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
sentence_to_token_ids Ran
ran on a synthesized input
eske/seq2seq/translate/utils.py
code served (permissive licence) · get_code("4f84c74cf4c69c83")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family 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.

UNK_ID = 2

# Construct a simple vocabulary dictionary
vocabulary = {
    "I": 1,
    "have": 2,
    "a": 4,
    "dog": 7,
    "cat": 5,
    "the": 3,
}

# Test sentence - words not in vocabulary will map to UNK_ID
sentence = "I have a dog"

# Character-level test vocabulary (single characters)
char_vocabulary = {
    "I": 1,
    " ": 2,
    "h": 3,
    "a": 4,
    "v": 5,
    "e": 6,
    "d": 7,
    "o": 8,
    "g": 9,
}

ARGS = {
    "sentence": sentence,
    "vocabulary": vocabulary,
    "character_level": False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family 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
sort_key Ran
ran on a synthesized input
eske/seq2seq/config/AMU/avg.py
code served (permissive licence) · get_code("94dc1f1512737035")
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.

keys = ['ter', 'bleu', 'bleu1', 'wer']

def sort_key(item):
    key, _ = item
    if key in keys:
        return keys.index(key)
    else:
        return len(keys)

# The function takes a single parameter 'item' which should be a tuple/iterable
# with at least 2 elements: (key, value)
# It unpacks key and _ from item, then checks if key is in the keys list

ARGS = {
    'item': ('bleu', 42)
}
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 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
apply_time_pooling Not yet run
unverified (harvested, no recorded run)
eske/seq2seq/translate/rnn.py
code served (permissive licence) · get_code("1ff3291db854041a")
How it ran

no invocation recorded.

get_best_score Not yet run
unverified (harvested, no recorded run)
eske/seq2seq/run-tests.py
code served (permissive licence) · get_code("1141b89c91c7f61c")
How it ran

no invocation recorded.

get_weights Not yet run
unverified (harvested, no recorded run)
eske/seq2seq/translate/beam_search.py
code served (permissive licence) · get_code("2ac7677b1e6b3956")
How it ran

no invocation recorded.

reverse_edits Not yet run
unverified (harvested, no recorded run)
eske/seq2seq/translate/utils.py
code served (permissive licence) · get_code("57a6b91d7cad71f0")
How it ran

no invocation recorded.

stack_bidirectional_dynamic_rnn Not yet run
unverified (harvested, no recorded run)
eske/seq2seq/translate/rnn.py
code served (permissive licence) · get_code("92c33312592e9111")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 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
get_shape 1 other paper
1 later
1808.09560 (2018-08)
On Learning 3D Face Morphable Model from In-the-wild Images

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 proposes a first attempt to build an end-to-end speech-to-text translation system, which does not use source language transcription during learning or decoding. We propose a model for direct speech-to-text translation, which gives promising results on a small French-English synthetic corpus. Relaxing the need for source language transcription would drastically change the data collection methodology in speech translation, especially in under-resourced scenarios. For instance, in the former project DARPA TRANSTAC (speech translation from spoken Arabic dialects), a large effort was devoted to the collection of speech transcripts (and a prerequisite to obtain transcripts was often a detailed transcription guide for languages with little standardized spelling). Now, if end-to-end approaches for speech-to-text translation are successful, one might consider collecting data by asking bilingual speakers to directly utter speech in the source language from target language text utterances. Such an approach has the advantage to be applicable to any unwritten (source) language.

For agents

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

get_harvested_code_for_paper("1612.01744")
get_code_for_paper("1612.01744")
have("1612.01744")

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

curl https://syntology.ai/api/ran/1612.01744.json

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

[![Syntology run record](https://syntology.ai/api/ran/1612.01744.svg)](https://syntology.ai/paper/1612.01744)

Connect an agent — have() is free.