| batch_slice |
Ran ran on a synthesized input |
jzbjyb/rri_match/rri.py code served (permissive licence) · get_code("3f6f8ca566974e84") How it ranRan 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 3, dtype family float, shape (2, 3, 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 tensorflow as tf
import numpy as np
# Create a small batch tensor with shape [batch_size, sequence_length, feature_dim]
batch = tf.constant(np.random.randn(2, 4, 3).astype(np.float32))
# Create start indices for each batch element
start = tf.constant([0, 1], dtype=tf.int32)
# Create offset values (how many elements to slice)
offset = tf.constant([2, 3], dtype=tf.int32)
# Pad value for constant padding
pad_values = 0.0
ARGS = {
"batch": batch,
"start": start,
"offset": offset,
"pad_values": pad_values,
}
- 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 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 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
|
| batch_slice_slow |
Ran ran on a synthesized input |
jzbjyb/rri_match/rri.py code served (permissive licence) · get_code("33b35e906cc63cae") How it ranRan 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 3, dtype family float, shape (2, 3, 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 tensorflow as tf
import numpy as np
# Set seed for reproducibility
tf.random.set_seed(0)
np.random.seed(0)
# Create a small batch tensor with shape (batch_size=2, sequence_length=4, feature_dim=3)
batch = tf.constant([
[[1.0, 2.0, 3.0], [4.0, 5.0, 6.0], [7.0, 8.0, 9.0], [10.0, 11.0, 12.0]],
[[13.0, 14.0, 15.0], [16.0, 17.0, 18.0], [19.0, 20.0, 21.0], [22.0, 23.0, 24.0]]
], dtype=tf.float32)
# Create start indices for each batch element
start = tf.constant([0, 1], dtype=tf.int32)
# Create offset values for each batch element
offset = tf.constant([2, 3], dtype=tf.int32)
# Create pad values (scalar)
pad_values = 0.0
ARGS = {
"batch": batch,
"start": start,
"offset": offset,
"pad_values": pad_values
}
- 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 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 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
|
| batch_where |
Ran ran on a synthesized input |
jzbjyb/rri_match/rri.py code served (permissive licence) · get_code("e93059a43aa4dbfa") How it ranRan 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 3, 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 tensorflow as tf
# Create a condition tensor (batch_size=2, dim=4)
cond = tf.constant([[True, False, True, False],
[False, True, False, True]])
# Create xs as a list of tensors (when not equal to 1)
xs = [
tf.constant([[1.0, 2.0, 3.0, 4.0],
[5.0, 6.0, 7.0, 8.0]], dtype=tf.float32),
tf.constant([[9.0, 10.0, 11.0, 12.0],
[13.0, 14.0, 15.0, 16.0]], dtype=tf.float32)
]
# Create ys as a list of tensors with matching shape
ys = [
tf.constant([[17.0, 18.0, 19.0, 20.0],
[21.0, 22.0, 23.0, 24.0]], dtype=tf.float32),
tf.constant([[25.0, 26.0, 27.0, 28.0],
[29.0, 30.0, 31.0, 32.0]], dtype=tf.float32)
]
ARGS = {
"cond": cond,
"xs": xs,
"ys": ys
}
- 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 3, dtype family float — a list output has no array shape of its own; none is stated
- dtype family
- int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
- Environment
- image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
|
| dcg |
Ran ran on a synthesized input |
jzbjyb/rri_match/metric.py code served (permissive licence) · get_code("34a2a38a33c779ed") How it ranRan 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 float64, ndim 0, 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 numpy as np
# Construct arguments for dcg function
query = "query_1"
doc_list = ["doc_0", "doc_1", "doc_2", "doc_3", "doc_4"]
# qd_judge is a nested dict mapping query -> doc -> relevance score
# The function accesses qd_judge[query][doc_list[i]] for each i in range(real_top_k)
qd_judge = {
"query_1": {
"doc_0": 3,
"doc_1": 2,
"doc_2": 1,
"doc_3": 0,
"doc_4": 2,
}
}
top_k = 5
ARGS = {
"query": query,
"doc_list": doc_list,
"qd_judge": qd_judge,
"top_k": top_k,
}
- 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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
- 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
|
| dynamic_split_and_pad_map_fn |
Ran ran on a synthesized input |
jzbjyb/rri_match/represent.py code served (permissive licence) · get_code("39edbcf27e85fbea") How it ranRan 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 3, dtype family float, shape (3, 3, 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
# Create small tensors for testing
x = tf.constant([[1.0, 2.0, 3.0, 4.0],
[5.0, 6.0, 7.0, 8.0],
[9.0, 10.0, 11.0, 12.0],
[13.0, 14.0, 15.0, 16.0],
[17.0, 18.0, 19.0, 20.0],
[21.0, 22.0, 23.0, 24.0]])
# sp represents split offsets - how many elements to take for each piece
sp = tf.constant([2, 3, 1], dtype=tf.int32)
ARGS = {
"x": x,
"sp": sp
}
- 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 EagerTensor, ndim 3, dtype family float, shape (3, 3, 4) — shape (3, 3, 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
|
| dynamic_split_and_pad_map_fn_ |
Ran ran on a synthesized input |
jzbjyb/rri_match/represent.py code served (permissive licence) · get_code("4cfeb52c620297d9") How it ranRan 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 3, dtype family float, shape (2, 3, 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 seed for reproducibility
tf.random.set_seed(0)
# Create a small tensor to split and pad
# Shape: (10, 4) - 10 elements along axis 0, 4 features
x = tf.constant(np.arange(40, dtype=np.float32).reshape(10, 4))
# Create start indices - where each piece begins
# Using 2 pieces for this example
start = tf.constant([0, 5], dtype=tf.int32)
# Create offset lengths - how long each piece is
# First piece: 3 elements, Second piece: 2 elements
offset = tf.constant([3, 2], dtype=tf.int32)
ARGS = {
'x': x,
'start': start,
'offset': offset
}
- 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 3, dtype family float, shape (2, 3, 4) — shape (2, 3, 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
|
| eval_mm |
Ran ran on a synthesized input |
jzbjyb/rri_match/ir_match_matrix.py code served (permissive licence) · get_code("96826e3e74c40415") How it ranRan 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 numpy as np
# Construct a small match matrix (bs=2, h=3, w=4)
# Values represent match scores between 0 and 1
mms = np.array([
[[0.5, 0.2, 0.8, 0.1],
[0.3, 0.9, 0.0, 0.4],
[0.7, 0.1, 0.6, 0.2]],
[[0.0, 0.0, 0.0, 0.0],
[1.0, 0.0, 1.0, 0.0],
[0.0, 0.5, 0.0, 0.5]]
], dtype=np.float32)
features = 'tf_proximity'
ARGS = {
'mms': mms,
'features': features
}
- 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
|
| fft_ind_gen |
Ran ran on a synthesized input |
jzbjyb/rri_match/data_gen.py code served (permissive licence) · get_code("5c573a033c9bdc92") How it ranRan 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 torch
ARGS = {
"n": 8
}
- 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 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
|
| gaussian_random_field |
Ran ran on a synthesized input |
jzbjyb/rri_match/data_gen.py code served (permissive licence) · get_code("939949c527bd2577") How it ranRan 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 (64, 48) 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
def custom_pk(k):
"""Custom power spectrum function"""
return k ** (-2.5)
ARGS = {
"pk": custom_pk,
"size1": 64,
"size2": 48,
"anisotropy": True
}
- 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 (64, 48) — shape (64, 48) 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_arguments |
Ran ran; our contract draft was wrong, not the code |
philipperemy/tensorflow-multi-dimensional-lstm/trainer.py code served (permissive licence) · get_code("bf26c8d187534a2d") How it ranRan 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 Namespace. 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 argparse
# Create a minimal ArgumentParser with some basic arguments
parser = argparse.ArgumentParser(
description="Test parser for get_arguments function"
)
parser.add_argument('--name', type=str, default='test', help='A test argument')
parser.add_argument('--value', type=int, default=42, help='A numeric argument')
parser.add_argument('--flag', action='store_true', help='A boolean flag')
# Mock sys.argv to avoid parsing actual command line arguments
import sys
original_argv = sys.argv
sys.argv = ['test_script.py']
ARGS = {
'parser': parser
}
# Restore original argv
sys.argv = original_argv
- 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 Namespace — a Namespace output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Namespace 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_padding |
Ran ran on a synthesized input |
jzbjyb/rri_match/represent.py code served (permissive licence) · get_code("3e2732158df7801d") How it ranRan 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 tensorflow as tf
torch.manual_seed(0)
# Create a TensorFlow tensor (the function calls .get_shape() which is TF-style)
x = tf.constant([[1.0, 2.0, 3.0, 4.0],
[5.0, 6.0, 7.0, 8.0]])
axis = 1
pad = [1, 2]
ARGS = {
"x": x,
"axis": axis,
"pad": pad
}
- Setup block drafted by
- a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
- Imports
- tensorflow, 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
|
| ndcg |
Ran ran on a synthesized input |
jzbjyb/rri_match/metric.py code served (permissive licence) · get_code("e41abfe79a1abfca") How it ranRan 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 float64, ndim 0, 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 numpy as np
# Build qd_judge: a dict mapping queries to dicts of {doc: relevance_score}
# The function accesses qd_judge[query] and qd_judge[query][doc]
qd_judge = {
"query1": {
"doc1": 3,
"doc2": 2,
"doc3": 1,
"doc4": 0,
},
"query2": {
"doc5": 2,
"doc6": 1,
"doc7": 0,
}
}
# Build doc_list: a list of document IDs
doc_list = ["doc1", "doc2", "doc3", "doc4"]
# query: a string key that exists in qd_judge
query = "query1"
# top_k: integer parameter
top_k = 5
ARGS = {
"query": query,
"doc_list": doc_list,
"qd_judge": qd_judge,
"top_k": top_k,
}
- 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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
- 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
|
| next_batch_gaussian |
Ran ran on a synthesized input |
jzbjyb/rri_match/data_gen.py code served (permissive licence) · get_code("0eb4737d6f16a402") How it ranRan 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 4, 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 numpy as np
ARGS = {
"bs": 2,
"h": 4,
"w": 3,
"anisotropy": True
}
- 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, ndim 4, 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
|
| parse_sogou_qcl_one_qd |
Ran ran on a synthesized input |
jzbjyb/rri_match/utils.py code served (permissive licence) · get_code("3be55b1f27889067") How it ranRan in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.
import torch
import torch.nn as nn
import numpy as np
# The function parse_sogou_qcl_one_qd expects a list of lines (strings)
# that represent XML-like structured data for search queries and documents.
# Let's construct a realistic example with the nested structure:
# <q>
# <query>...</query>
# <query_frequency>...</query_frequency>
# <query_id>...</query_id>
# <doc>
# <url>...</url>
# <doc_id>...</doc_id>
# <title>...</title>
# <content>...</content>
# <html>...</html>
# <doc_frequency>...</doc_frequency>
# <relevance>
# <TCM>...</TCM>
# <DBN>...</DBN>
# <PSCM>...</PSCM>
# <TACM>...</TACM>
# <UBM>...</UBM>
# </relevance>
# </doc>
# </q>
lines = [
'<q>',
'<query>python programming</query>',
'<query_frequency>100</query_frequency>',
'<query_id>q001</query_id>',
'<doc>',
'<url>https://example.com/page1</url>',
'<doc_id>d001</doc_id>',
'<title>Python Guide</title>',
'<content>Learn Python programming basics</content>',
'<html><html>Python content</html></html>',
'<doc_frequency>50</doc_frequency>',
'<relevance>',
'<TCM>0.95</TCM>',
'<DBN>0.87</DBN>',
'<PSCM>0.92</PSCM>',
'<TACM>0.89</TACM>',
'<UBM>0.91</UBM>',
'</relevance>',
'</doc>',
'<doc>',
'<url>https://example.com/page2</url>',
'<doc_id>d002</doc_id>',
'<title>Advanced Python</title>',
'<content>Advanced Python concepts</content>',
'<html><html>Advanced Python</html></html>',
'<doc_frequency>45</doc_frequency>',
'<relevance>',
'<TCM>0.88</TCM>',
'<DBN>0.80</DBN>',
'<PSCM>0.85</PSCM>',
'<TACM>0.82</TACM>',
'<UBM>0.84</UBM>',
'</relevance>',
'</doc>',
'</q>',
]
ARGS = {
'lines': lines
}
- 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 1, dtype family float — a list output has no array shape of its own; none is stated
- dtype family
- int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
- Environment
- image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
|
| precision |
Ran ran on a synthesized input |
jzbjyb/rri_match/metric.py code served (permissive licence) · get_code("2007c4e9c81aaf2c") How it ranRan 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 torch
torch.manual_seed(0)
# Construct arguments for the precision function
# The function expects:
# - query: a key to index into qd_judge dict
# - doc_list: a list of document identifiers
# - qd_judge: a nested dict mapping query -> doc -> relevance score
# - top_k: integer parameter (has default value of 5)
query = "q1"
doc_list = ["doc1", "doc2", "doc3", "doc4", "doc5", "doc6"]
# qd_judge is a nested dict: qd_judge[query][doc] should return a relevance score
# Positive values indicate relevant documents
qd_judge = {
"q1": {
"doc1": 1, # relevant
"doc2": 0, # not relevant
"doc3": 2, # relevant
"doc4": 0, # not relevant
"doc5": 1, # relevant
"doc6": 0, # not relevant
}
}
top_k = 5
ARGS = {
"query": query,
"doc_list": doc_list,
"qd_judge": qd_judge,
"top_k": top_k
}
- 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 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
|
| word_segment |
Ran ran on a synthesized input |
jzbjyb/rri_match/utils.py code served (permissive licence) · get_code("f64b418ef9cd57ca") How it ranRan in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.
import torch
import numpy as np
import jieba
# The function word_segment takes a single parameter: text
# It uses jieba.cut() to segment the text, which expects a string
ARGS = {
"text": "今天天气很好,我想去公园散步。这是一个测试句子。"
}
- Setup block drafted by
- a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
- Imports
- jieba, 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 1, dtype family float — a list output has no array shape of its own; none is stated
- dtype family
- int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
- Environment
- image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
|
| conv_knrm |
Not yet run unverified (harvested, no recorded run) |
jzbjyb/rri_match/cnn.py code served (permissive licence) · get_code("b6525b4986ae0270") How it ran |
| knrm |
Not yet run unverified (harvested, no recorded run) |
jzbjyb/rri_match/cnn.py code served (permissive licence) · get_code("c7d7c11bd2988e25") How it ran |
| ln |
Not yet run unverified (harvested, no recorded run) |
jzbjyb/rri_match/md_lstm.py code served (permissive licence) · get_code("03595c4790906612") How it ran |
| map_fn_concat |
Not yet run unverified (harvested, no recorded run) |
jzbjyb/rri_match/basic_op.py code served (permissive licence) · get_code("3f90df6c4fe61b20") How it ran |
| mlp |
Not yet run unverified (harvested, no recorded run) |
jzbjyb/rri_match/cnn.py code served (permissive licence) · get_code("6fb3d9e6ce59a987") How it ran |
| save_graph |
Not yet run unverified (harvested, no recorded run) |
T-Almeida/tensorflow-keras-multidimensional-rnn/utils.py code served (permissive licence) · get_code("9d88c7a8d285c466") How it ran |
| tf_jacobian |
Not yet run unverified (harvested, no recorded run) |
jzbjyb/rri_match/utils.py code served (permissive licence) · get_code("f4f484080e6ec5d4") How it ran |