Merge pull request #2832 from activeloopai/docs_fix #1550
22436 tests run, 11800 passed, 10632 skipped, 4 failed.
Annotations
Check failure on line 656 in deeplake/core/vectorstore/test_deeplake_vectorstore.py
github-actions / JUnit Test Report
test_deeplake_vectorstore.test_search_quantitative[COS]
assert False
+ where False = all([True, True, True, False])
Raw output
distance_metric = 'COS'
hub_cloud_dev_token = 'eyJhbGciOiJub25lIiwidHlwIjoiSldUIn0.eyJpZCI6InRlc3RpbmdhY2MyIiwiYXBpX2tleSI6IjU4Y0tLb1p6UE1BbThPU2RpbTRiZ2tBekhWekt1VUE3MFJpNTNyZUpKRTJuaiJ9.'
@pytest.mark.slow
@requires_libdeeplake
@pytest.mark.parametrize("distance_metric", ["L1", "L2", "COS", "MAX"])
def test_search_quantitative(distance_metric, hub_cloud_dev_token):
"""Test whether TQL and Python return the same results"""
# initialize vector store object:
vector_store = DeepLakeVectorStore(
path="hub://testingacc2/vectorstore_test",
read_only=True,
token=hub_cloud_dev_token,
)
# use python implementation to search the data
data_p = vector_store.search(
embedding=query_embedding, exec_option="python", distance_metric=distance_metric
)
# use indra implementation to search the data
data_ce = vector_store.search(
embedding=query_embedding,
exec_option="compute_engine",
distance_metric=distance_metric,
)
assert len(data_p["score"]) == len(data_ce["score"])
> assert all(
[
isclose(
data_p["score"][i],
data_ce["score"][i],
abs_tol=0.00001
* (abs(data_p["score"][i]) + abs(data_ce["score"][i]))
/ 2,
)
for i in range(len(data_p["score"]))
]
)
E assert False
E + where False = all([True, True, True, False])
deeplake/core/vectorstore/test_deeplake_vectorstore.py:656: AssertionError
Check failure on line 726 in deeplake/core/vectorstore/test_deeplake_vectorstore.py
github-actions / JUnit Test Report
test_deeplake_vectorstore.test_search_managed
assert False
+ where False = all([True, True, True, False])
Raw output
hub_cloud_dev_token = 'eyJhbGciOiJub25lIiwidHlwIjoiSldUIn0.eyJpZCI6InRlc3RpbmdhY2MyIiwiYXBpX2tleSI6IjU4Y0tLb1p6UE1BbThPU2RpbTRiZ2tBekhWekt1VUE3MFJpNTNyZUpKRTJuaiJ9.'
@requires_libdeeplake
@pytest.mark.slow
def test_search_managed(hub_cloud_dev_token):
"""Test whether managed TQL and client-side TQL return the same results"""
# initialize vector store object:
vector_store = DeepLakeVectorStore(
path="hub://testingacc2/vectorstore_test_managed",
read_only=True,
token=hub_cloud_dev_token,
)
# use indra implementation to search the data
data_ce = vector_store.search(
embedding=query_embedding,
exec_option="compute_engine",
)
data_db = vector_store.search(
embedding=query_embedding,
exec_option="tensor_db",
)
assert "vectordb/" in vector_store.dataset_handler.dataset.base_storage.path
assert len(data_ce["score"]) == len(data_db["score"])
> assert all(
[
isclose(
data_ce["score"][i],
data_db["score"][i],
abs_tol=0.00001
* (abs(data_ce["score"][i]) + abs(data_db["score"][i]))
/ 2,
)
for i in range(len(data_ce["score"]))
]
)
E assert False
E + where False = all([True, True, True, False])
deeplake/core/vectorstore/test_deeplake_vectorstore.py:726: AssertionError
Check failure on line 222 in deeplake/api/tests/test_json.py
github-actions / JUnit Test Report
test_json.test_json_transform[None-gdrive_ds]
deeplake.util.exceptions.TransformError: Transform failed. See traceback for more details.
Raw output
self = <deeplake.core.transform.transform.Pipeline object at 0x0000022ECBAC2700>
data_in = [{'x': [1, 2, 3], 'y': [4, [5, 6]]}, {'x': [1, 2, 3], 'y': [4, {'z': [0.1, 0.2, []]}]}, ['a', ['b', 'c'], {'d': 1.0}], [1.0, 2.0, 3.0, 4.0], ['a', 'b', 'c', 'd'], 1, ...]
ds_out = Dataset(path='gdrive://hubtest/tmp63f6', tensors=['json'])
num_workers = 2, scheduler = 'threaded', progressbar = True, skip_ok = False
check_lengths = True, pad_data_in = False, read_only_ok = False, cache_size = 16
checkpoint_interval = 0, ignore_errors = False, verbose = True, kwargs = {}
overwrite = False
original_data_in = [{'x': [1, 2, 3], 'y': [4, [5, 6]]}, {'x': [1, 2, 3], 'y': [4, {'z': [0.1, 0.2, []]}]}, ['a', ['b', 'c'], {'d': 1.0}], [1.0, 2.0, 3.0, 4.0], ['a', 'b', 'c', 'd'], 1, ...]
initial_padding_state = None
target_ds = Dataset(path='gdrive://hubtest/tmp63f6', tensors=['json'])
compute_provider = <deeplake.core.compute.thread.ThreadProvider object at 0x0000022ECBAC2880>
compute_id = '3f8d9560b8444a7092d16ab2027d555b', initial_autoflush = True
def eval(
self,
data_in,
ds_out: Optional[deeplake.Dataset] = None,
num_workers: int = 0,
scheduler: str = "threaded",
progressbar: bool = True,
skip_ok: bool = False,
check_lengths: bool = True,
pad_data_in: bool = False,
read_only_ok: bool = False,
cache_size: int = DEFAULT_TRANSFORM_SAMPLE_CACHE_SIZE,
checkpoint_interval: int = 0,
ignore_errors: bool = False,
verbose: bool = True,
**kwargs,
):
"""Evaluates the pipeline on ``data_in`` to produce an output dataset ``ds_out``.
Args:
data_in: Input passed to the transform to generate output dataset. Should support \__getitem__ and \__len__. Can be a Deep Lake dataset.
ds_out (Dataset, optional): - The dataset object to which the transform will get written. If this is not provided, ``data_in`` will be overwritten if it is a Deep Lake dataset, otherwise error will be raised.
- It should have all keys being generated in output already present as tensors. It's initial state should be either:
- **Empty**, i.e., all tensors have no samples. In this case all samples are added to the dataset.
- **All tensors are populated and have same length.** In this case new samples are appended to the dataset.
num_workers (int): The number of workers to use for performing the transform. Defaults to 0. When set to 0, it will always use serial processing, irrespective of the scheduler.
scheduler (str): The scheduler to be used to compute the transformation. Supported values include: 'serial', 'threaded', 'processed' and 'ray'.
Defaults to 'threaded'.
progressbar (bool): Displays a progress bar if ``True`` (default).
skip_ok (bool): If ``True``, skips the check for output tensors generated. This allows the user to skip certain tensors in the function definition.
This is especially useful for inplace transformations in which certain tensors are not modified. Defaults to ``False``.
check_lengths (bool): If ``True``, checks whether ``ds_out`` has tensors of same lengths initially.
pad_data_in (bool): If ``True``, pads tensors of ``data_in`` to match the length of the largest tensor in ``data_in``.
Defaults to ``False``.
read_only_ok (bool): If ``True`` and output dataset is same as input dataset, the read-only check is skipped.
Defaults to False.
cache_size (int): Cache size to be used by transform per worker.
checkpoint_interval (int): If > 0, the transform will be checkpointed with a commit every ``checkpoint_interval`` input samples to avoid restarting full transform due to intermitten failures. If the transform is interrupted, the intermediate data is deleted and the dataset is reset to the last commit.
If <= 0, no checkpointing is done. Checkpoint interval should be a multiple of num_workers if num_workers > 0. Defaults to 0.
ignore_errors (bool): If ``True``, input samples that causes transform to fail will be skipped and the errors will be ignored **if possible**.
verbose (bool): If ``True``, prints additional information about the transform.
**kwargs: Additional arguments.
Raises:
InvalidInputDataError: If ``data_in`` passed to transform is invalid. It should support \__getitem__ and \__len__ operations. Using scheduler other than "threaded" with deeplake dataset having base storage as memory as ``data_in`` will also raise this.
InvalidOutputDatasetError: If all the tensors of ``ds_out`` passed to transform don't have the same length. Using scheduler other than "threaded" with deeplake dataset having base storage as memory as ``ds_out`` will also raise this.
TensorMismatchError: If one or more of the outputs generated during transform contain different tensors than the ones present in 'ds_out' provided to transform.
UnsupportedSchedulerError: If the scheduler passed is not recognized. Supported values include: 'serial', 'threaded', 'processed' and 'ray'.
TransformError: All other exceptions raised if there are problems while running the pipeline.
ValueError: If ``num_workers`` > 0 and ``checkpoint_interval`` is not a multiple of ``num_workers`` or if ``checkpoint_interval`` > 0 and ds_out is None.
# noqa: DAR401
Example::
@deeplake.compute
def my_fn(sample_in: Any, samples_out, my_arg0, my_arg1=0):
samples_out.my_tensor.append(my_arg0 * my_arg1)
# This transform can be used using the eval method in one of these 2 ways:-
# Directly evaluating the method
# here arg0 and arg1 correspond to the 3rd and 4th argument in my_fn
my_fn(arg0, arg1).eval(data_in, ds_out, scheduler="threaded", num_workers=5)
# As a part of a Transform pipeline containing other functions
pipeline = deeplake.compose([my_fn(a, b), another_function(x=2)])
pipeline.eval(data_in, ds_out, scheduler="processed", num_workers=2)
Note:
``pad_data_in`` is only applicable if ``data_in`` is a Deep Lake dataset.
"""
num_workers, scheduler = sanitize_workers_scheduler(num_workers, scheduler)
overwrite = ds_out is None
deeplake_reporter.feature_report(
feature_name="eval",
parameters={"Num_Workers": str(num_workers), "Scheduler": scheduler},
)
check_transform_data_in(data_in, scheduler)
data_in, original_data_in, initial_padding_state = prepare_data_in(
data_in, pad_data_in, overwrite
)
target_ds = data_in if overwrite else ds_out
check_transform_ds_out(
target_ds, scheduler, check_lengths, read_only_ok and overwrite
)
# if overwrite then we've already flushed and autocheckecked out data_in which is target_ds now
if not overwrite:
target_ds.flush()
auto_checkout(target_ds)
compute_provider = get_compute_provider(scheduler, num_workers)
compute_id = str(uuid4().hex)
target_ds._send_compute_progress(compute_id=compute_id, start=True, progress=0)
initial_autoflush = target_ds.storage.autoflush
target_ds.storage.autoflush = False
if not check_lengths or read_only_ok:
skip_ok = True
checkpointing_enabled = checkpoint_interval > 0
total_samples = len_data_in(data_in)
if checkpointing_enabled:
check_checkpoint_interval(
data_in,
checkpoint_interval,
num_workers,
overwrite,
verbose,
)
datas_in = [
data_in[i : i + checkpoint_interval]
for i in range(0, len_data_in(data_in), checkpoint_interval)
]
else:
datas_in = [data_in]
samples_processed = 0
desc = get_pbar_description(self.functions)
if progressbar:
pbar = get_progress_bar(len_data_in(data_in), desc)
pqueue = compute_provider.create_queue()
else:
pbar, pqueue = None, None
try:
desc = desc.split()[1]
completed = False
progress = 0.0
for data_in in datas_in:
if checkpointing_enabled:
target_ds._commit(
f"Auto-commit during deeplake.compute of {desc} after {progress}% progress",
None,
False,
is_checkpoint=True,
total_samples_processed=samples_processed,
)
progress = round(
(samples_processed + len_data_in(data_in)) / total_samples * 100, 2
)
end = progress == 100
progress_args = {
"compute_id": compute_id,
"progress": progress,
"end": end,
}
try:
> self.run(
data_in,
target_ds,
compute_provider,
num_workers,
scheduler,
progressbar,
overwrite,
skip_ok,
read_only_ok and overwrite,
cache_size,
pbar,
pqueue,
ignore_errors,
**kwargs,
)
deeplake\core\transform\transform.py:288:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
deeplake\core\transform\transform.py:430: in run
result = compute.map_with_progress_bar(
deeplake\core\compute\provider.py:62: in map_with_progress_bar
result = self.map(sub_func, iterable)
deeplake\core\compute\thread.py:13: in map
return self.pool.map(func, iterable)
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\pathos\threading.py:144: in map
return _pool.map(star(f), zip(*args), **kwds)
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\multiprocess\pool.py:364: in map
return self._map_async(func, iterable, mapstar, chunksize).get()
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\multiprocess\pool.py:771: in get
raise self._value
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\multiprocess\pool.py:125: in worker
result = (True, func(*args, **kwds))
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\multiprocess\pool.py:48: in mapstar
return list(map(*args))
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\pathos\helpers\mp_helper.py:15: in <lambda>
func = lambda args: f(*args)
deeplake\core\compute\provider.py:57: in sub_func
return func(pg_callback, *args, **kwargs)
deeplake\util\transform.py:345: in store_data_slice_with_pbar
all_chunk_engines = create_worker_chunk_engines(
deeplake\util\transform.py:439: in create_worker_chunk_engines
storage_chunk_engine = ChunkEngine(tensor, storage_cache, version_state)
deeplake\core\chunk_engine.py:230: in __init__
tensor_meta = self.tensor_meta
deeplake\core\chunk_engine.py:321: in tensor_meta
self._tensor_meta = self.meta_cache.get_deeplake_object(key, TensorMeta)
deeplake\core\storage\lru_cache.py:166: in get_deeplake_object
item = self[path]
deeplake\core\storage\lru_cache.py:217: in __getitem__
result = self.next_storage[path]
deeplake\core\storage\google_drive.py:287: in __getitem__
return self.get_object_by_id(id)
deeplake\core\storage\google_drive.py:277: in get_object_by_id
status, done = downloader.next_chunk()
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\googleapiclient\_helpers.py:131: in positional_wrapper
return wrapped(*args, **kwargs)
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\googleapiclient\http.py:740: in next_chunk
resp, content = _retry_request(
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\googleapiclient\http.py:221: in _retry_request
raise exception
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\googleapiclient\http.py:190: in _retry_request
resp, content = http.request(uri, method, *args, **kwargs)
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\google_auth_httplib2.py:218: in request
response, content = self.http.request(
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\httplib2\__init__.py:1724: in request
(response, content) = self._request(
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\httplib2\__init__.py:1444: in _request
(response, content) = self._conn_request(conn, request_uri, method, body, headers)
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\site-packages\httplib2\__init__.py:1396: in _conn_request
response = conn.getresponse()
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\http\client.py:1344: in getresponse
response.begin()
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\http\client.py:307: in begin
version, status, reason = self._read_status()
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\http\client.py:268: in _read_status
line = str(self.fp.readline(_MAXLINE + 1), "iso-8859-1")
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\socket.py:669: in readinto
return self._sock.recv_into(b)
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\ssl.py:1241: in recv_into
return self.read(nbytes, buffer)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = <ssl.SSLSocket [closed] fd=-1, family=AddressFamily.AF_INET, type=SocketKind.SOCK_STREAM, proto=0>
len = 8192, buffer = <memory at 0x0000022EC95E8280>
def read(self, len=1024, buffer=None):
"""Read up to LEN bytes and return them.
Return zero-length string on EOF."""
self._checkClosed()
if self._sslobj is None:
raise ValueError("Read on closed or unwrapped SSL socket.")
try:
if buffer is not None:
> return self._sslobj.read(len, buffer)
E socket.timeout: The read operation timed out
c:\hostedtoolcache\windows\python\3.8.10\x64\lib\ssl.py:1099: timeout
The above exception was the direct cause of the following exception:
ds = Dataset(path='gdrive://hubtest/tmp63f6', tensors=['json'])
compression = None, scheduler = 'threaded'
@enabled_non_gcs_datasets
@pytest.mark.parametrize("compression", ["lz4", None])
@pytest.mark.slow
def test_json_transform(ds, compression, scheduler="threaded"):
ds.create_tensor("json", htype="json", sample_compression=compression)
items = [
{"x": [1, 2, 3], "y": [4, [5, 6]]},
{"x": [1, 2, 3], "y": [4, {"z": [0.1, 0.2, []]}]},
["a", ["b", "c"], {"d": 1.0}],
[1.0, 2.0, 3.0, 4.0],
["a", "b", "c", "d"],
1,
5.0,
True,
False,
None,
] * 5
expected = [*items[:9], {}] * 5
@deeplake.compute
def upload(stuff, ds):
ds.json.append(stuff)
return ds
> upload().eval(items, ds, num_workers=2, scheduler=scheduler)
deeplake\api\tests\test_json.py:222:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
deeplake\core\transform\transform.py:105: in eval
pipeline.eval(
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = <deeplake.core.transform.transform.Pipeline object at 0x0000022ECBAC2700>
data_in = [{'x': [1, 2, 3], 'y': [4, [5, 6]]}, {'x': [1, 2, 3], 'y': [4, {'z': [0.1, 0.2, []]}]}, ['a', ['b', 'c'], {'d': 1.0}], [1.0, 2.0, 3.0, 4.0], ['a', 'b', 'c', 'd'], 1, ...]
ds_out = Dataset(path='gdrive://hubtest/tmp63f6', tensors=['json'])
num_workers = 2, scheduler = 'threaded', progressbar = True, skip_ok = False
check_lengths = True, pad_data_in = False, read_only_ok = False, cache_size = 16
checkpoint_interval = 0, ignore_errors = False, verbose = True, kwargs = {}
overwrite = False
original_data_in = [{'x': [1, 2, 3], 'y': [4, [5, 6]]}, {'x': [1, 2, 3], 'y': [4, {'z': [0.1, 0.2, []]}]}, ['a', ['b', 'c'], {'d': 1.0}], [1.0, 2.0, 3.0, 4.0], ['a', 'b', 'c', 'd'], 1, ...]
initial_padding_state = None
target_ds = Dataset(path='gdrive://hubtest/tmp63f6', tensors=['json'])
compute_provider = <deeplake.core.compute.thread.ThreadProvider object at 0x0000022ECBAC2880>
compute_id = '3f8d9560b8444a7092d16ab2027d555b', initial_autoflush = True
def eval(
self,
data_in,
ds_out: Optional[deeplake.Dataset] = None,
num_workers: int = 0,
scheduler: str = "threaded",
progressbar: bool = True,
skip_ok: bool = False,
check_lengths: bool = True,
pad_data_in: bool = False,
read_only_ok: bool = False,
cache_size: int = DEFAULT_TRANSFORM_SAMPLE_CACHE_SIZE,
checkpoint_interval: int = 0,
ignore_errors: bool = False,
verbose: bool = True,
**kwargs,
):
"""Evaluates the pipeline on ``data_in`` to produce an output dataset ``ds_out``.
Args:
data_in: Input passed to the transform to generate output dataset. Should support \__getitem__ and \__len__. Can be a Deep Lake dataset.
ds_out (Dataset, optional): - The dataset object to which the transform will get written. If this is not provided, ``data_in`` will be overwritten if it is a Deep Lake dataset, otherwise error will be raised.
- It should have all keys being generated in output already present as tensors. It's initial state should be either:
- **Empty**, i.e., all tensors have no samples. In this case all samples are added to the dataset.
- **All tensors are populated and have same length.** In this case new samples are appended to the dataset.
num_workers (int): The number of workers to use for performing the transform. Defaults to 0. When set to 0, it will always use serial processing, irrespective of the scheduler.
scheduler (str): The scheduler to be used to compute the transformation. Supported values include: 'serial', 'threaded', 'processed' and 'ray'.
Defaults to 'threaded'.
progressbar (bool): Displays a progress bar if ``True`` (default).
skip_ok (bool): If ``True``, skips the check for output tensors generated. This allows the user to skip certain tensors in the function definition.
This is especially useful for inplace transformations in which certain tensors are not modified. Defaults to ``False``.
check_lengths (bool): If ``True``, checks whether ``ds_out`` has tensors of same lengths initially.
pad_data_in (bool): If ``True``, pads tensors of ``data_in`` to match the length of the largest tensor in ``data_in``.
Defaults to ``False``.
read_only_ok (bool): If ``True`` and output dataset is same as input dataset, the read-only check is skipped.
Defaults to False.
cache_size (int): Cache size to be used by transform per worker.
checkpoint_interval (int): If > 0, the transform will be checkpointed with a commit every ``checkpoint_interval`` input samples to avoid restarting full transform due to intermitten failures. If the transform is interrupted, the intermediate data is deleted and the dataset is reset to the last commit.
If <= 0, no checkpointing is done. Checkpoint interval should be a multiple of num_workers if num_workers > 0. Defaults to 0.
ignore_errors (bool): If ``True``, input samples that causes transform to fail will be skipped and the errors will be ignored **if possible**.
verbose (bool): If ``True``, prints additional information about the transform.
**kwargs: Additional arguments.
Raises:
InvalidInputDataError: If ``data_in`` passed to transform is invalid. It should support \__getitem__ and \__len__ operations. Using scheduler other than "threaded" with deeplake dataset having base storage as memory as ``data_in`` will also raise this.
InvalidOutputDatasetError: If all the tensors of ``ds_out`` passed to transform don't have the same length. Using scheduler other than "threaded" with deeplake dataset having base storage as memory as ``ds_out`` will also raise this.
TensorMismatchError: If one or more of the outputs generated during transform contain different tensors than the ones present in 'ds_out' provided to transform.
UnsupportedSchedulerError: If the scheduler passed is not recognized. Supported values include: 'serial', 'threaded', 'processed' and 'ray'.
TransformError: All other exceptions raised if there are problems while running the pipeline.
ValueError: If ``num_workers`` > 0 and ``checkpoint_interval`` is not a multiple of ``num_workers`` or if ``checkpoint_interval`` > 0 and ds_out is None.
# noqa: DAR401
Example::
@deeplake.compute
def my_fn(sample_in: Any, samples_out, my_arg0, my_arg1=0):
samples_out.my_tensor.append(my_arg0 * my_arg1)
# This transform can be used using the eval method in one of these 2 ways:-
# Directly evaluating the method
# here arg0 and arg1 correspond to the 3rd and 4th argument in my_fn
my_fn(arg0, arg1).eval(data_in, ds_out, scheduler="threaded", num_workers=5)
# As a part of a Transform pipeline containing other functions
pipeline = deeplake.compose([my_fn(a, b), another_function(x=2)])
pipeline.eval(data_in, ds_out, scheduler="processed", num_workers=2)
Note:
``pad_data_in`` is only applicable if ``data_in`` is a Deep Lake dataset.
"""
num_workers, scheduler = sanitize_workers_scheduler(num_workers, scheduler)
overwrite = ds_out is None
deeplake_reporter.feature_report(
feature_name="eval",
parameters={"Num_Workers": str(num_workers), "Scheduler": scheduler},
)
check_transform_data_in(data_in, scheduler)
data_in, original_data_in, initial_padding_state = prepare_data_in(
data_in, pad_data_in, overwrite
)
target_ds = data_in if overwrite else ds_out
check_transform_ds_out(
target_ds, scheduler, check_lengths, read_only_ok and overwrite
)
# if overwrite then we've already flushed and autocheckecked out data_in which is target_ds now
if not overwrite:
target_ds.flush()
auto_checkout(target_ds)
compute_provider = get_compute_provider(scheduler, num_workers)
compute_id = str(uuid4().hex)
target_ds._send_compute_progress(compute_id=compute_id, start=True, progress=0)
initial_autoflush = target_ds.storage.autoflush
target_ds.storage.autoflush = False
if not check_lengths or read_only_ok:
skip_ok = True
checkpointing_enabled = checkpoint_interval > 0
total_samples = len_data_in(data_in)
if checkpointing_enabled:
check_checkpoint_interval(
data_in,
checkpoint_interval,
num_workers,
overwrite,
verbose,
)
datas_in = [
data_in[i : i + checkpoint_interval]
for i in range(0, len_data_in(data_in), checkpoint_interval)
]
else:
datas_in = [data_in]
samples_processed = 0
desc = get_pbar_description(self.functions)
if progressbar:
pbar = get_progress_bar(len_data_in(data_in), desc)
pqueue = compute_provider.create_queue()
else:
pbar, pqueue = None, None
try:
desc = desc.split()[1]
completed = False
progress = 0.0
for data_in in datas_in:
if checkpointing_enabled:
target_ds._commit(
f"Auto-commit during deeplake.compute of {desc} after {progress}% progress",
None,
False,
is_checkpoint=True,
total_samples_processed=samples_processed,
)
progress = round(
(samples_processed + len_data_in(data_in)) / total_samples * 100, 2
)
end = progress == 100
progress_args = {
"compute_id": compute_id,
"progress": progress,
"end": end,
}
try:
self.run(
data_in,
target_ds,
compute_provider,
num_workers,
scheduler,
progressbar,
overwrite,
skip_ok,
read_only_ok and overwrite,
cache_size,
pbar,
pqueue,
ignore_errors,
**kwargs,
)
target_ds._send_compute_progress(**progress_args, status="success")
samples_processed += len_data_in(data_in)
completed = end
except Exception as e:
if checkpointing_enabled:
print(
"Transform failed. Resetting back to last committed checkpoint."
)
target_ds.reset(force=True)
target_ds._send_compute_progress(**progress_args, status="failed")
index, sample, suggest = None, None, False
if isinstance(e, TransformError):
index, sample, suggest = e.index, e.sample, e.suggest
if checkpointing_enabled and isinstance(index, int):
index = samples_processed + index
e = e.__cause__ # type: ignore
if isinstance(e, AllSamplesSkippedError):
raise e
> raise TransformError(
index=index,
sample=sample,
samples_processed=samples_processed,
suggest=suggest,
) from e
E deeplake.util.exceptions.TransformError: Transform failed. See traceback for more details.
deeplake\core\transform\transform.py:322: TransformError
Check failure on line 726 in deeplake/core/vectorstore/test_deeplake_vectorstore.py
github-actions / JUnit Test Report
test_deeplake_vectorstore.test_search_managed
assert False
+ where False = all([True, True, True, False])
Raw output
hub_cloud_dev_token = 'eyJhbGciOiJub25lIiwidHlwIjoiSldUIn0.eyJpZCI6InRlc3RpbmdhY2MyIiwiYXBpX2tleSI6IjU4Y0tLb1p6UE1BbThPU2RpbTRiZ2tBekhWekt1VUE3MFJpNTNyZUpKRTJuaiJ9.'
@requires_libdeeplake
@pytest.mark.slow
def test_search_managed(hub_cloud_dev_token):
"""Test whether managed TQL and client-side TQL return the same results"""
# initialize vector store object:
vector_store = DeepLakeVectorStore(
path="hub://testingacc2/vectorstore_test_managed",
read_only=True,
token=hub_cloud_dev_token,
)
# use indra implementation to search the data
data_ce = vector_store.search(
embedding=query_embedding,
exec_option="compute_engine",
)
data_db = vector_store.search(
embedding=query_embedding,
exec_option="tensor_db",
)
assert "vectordb/" in vector_store.dataset_handler.dataset.base_storage.path
assert len(data_ce["score"]) == len(data_db["score"])
> assert all(
[
isclose(
data_ce["score"][i],
data_db["score"][i],
abs_tol=0.00001
* (abs(data_ce["score"][i]) + abs(data_db["score"][i]))
/ 2,
)
for i in range(len(data_ce["score"]))
]
)
E assert False
E + where False = all([True, True, True, False])
deeplake/core/vectorstore/test_deeplake_vectorstore.py:726: AssertionError