vLLM is an inference and serving engine for large language models (LLMs). From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing (to_mono), while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results in inconsistency between audio heard by humans (e.g., through headphones/regular speakers) and a…

CVSS: 5.9 · CWE: CWE-20

View on NVD

Agenta is an open-source LLMOps platform. In Agenta-API prior to version 0.48.1, a Python sandbox escape vulnerability existed in Agenta's custom code evaluator. Agenta used RestrictedPython as a sandboxing mechanism for user-supplied evaluator code, but incorrectly whitelisted the `numpy` package as safe within the sandbox. This allowed authenticated users to bypass the sandbox and achieve arbit…

CVSS: 8.8 · CWE: CWE-94

View on NVD

MLX is an array framework for machine learning on Apple silicon. Prior to version 0.29.4, there is a heap buffer overflow in mlx::core::load() when parsing malicious NumPy .npy files. Attacker-controlled file causes 13-byte out-of-bounds read, leading to crash or information disclosure. This issue has been patched in version 0.29.4.

CVSS: 9.1 · CWE: CWE-122

View on NVD

A vulnerability, which was classified as problematic, has been found in MindSpore 2.5.0. Affected by this issue is the function mindspore.numpy.fft.rfft2. The manipulation leads to memory corruption. The attack needs to be approached locally. The exploit has been disclosed to the public and may be used.

CVSS: 3.3 · CWE: CWE-119

View on NVD

A vulnerability classified as problematic was found in MindSpore 2.5.0. Affected by this vulnerability is the function mindspore.numpy.fft.hfftn. The manipulation leads to memory corruption. It is possible to launch the attack on the local host. The exploit has been disclosed to the public and may be used.

CVSS: 3.3 · CWE: CWE-119

View on NVD

A pickle deserialization vulnerability exists in the Latex English error correction plug-in function of binary-husky/gpt_academic versions up to and including 3.83. This vulnerability allows attackers to achieve remote command execution by deserializing untrusted data. The issue arises from the inclusion of numpy in the deserialization whitelist, which can be exploited by constructing a malicious…

CVSS: 8.8 · CWE: CWE-502

View on NVD

joblib v1.4.2 was discovered to contain a deserialization vulnerability via the component joblib.numpy_pickle::NumpyArrayWrapper().read_array(). NOTE: this is disputed by the supplier because NumpyArrayWrapper is only used during caching of trusted content.

CVSS: 7.5 · CWE: CWE-502

View on NVD

sagemaker-python-sdk is a library for training and deploying machine learning models on Amazon SageMaker. The sagemaker.base_deserializers.NumpyDeserializer module before v2.218.0 allows potentially unsafe deserialization when untrusted data is passed as pickled object arrays. This consequently may allow an unprivileged third party to cause remote code execution, denial of service, affecting both…

CVSS: 7.8 · CWE: CWE-502

View on NVD

TensorFlow is an open source platform for machine learning. If `BCast::ToShape` is given input larger than an `int32`, it will crash, despite being supposed to handle up to an `int64`. An example can be seen in `tf.experimental.numpy.outer` by passing in large input to the input `b`. We have patched the issue in GitHub commit 8310bf8dd188ff780e7fc53245058215a05bdbe5. The fix will be included in T…

CVSS: 4.8 · CWE: CWE-704

View on NVD

TensorFlow is an open source platform for machine learning. If a numpy array is created with a shape such that one element is zero and the others sum to a large number, an error will be raised. We have patched the issue in GitHub commit 2b56169c16e375c521a3bc8ea658811cc0793784. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and Tensor…

CVSS: 4.8 · CWE: CWE-670

View on NVD

TensorFlow is an open source platform for machine learning. Prior to versions 2.9.0, 2.8.1, 2.7.2, and 2.6.4, TensorFlow's `saved_model_cli` tool is vulnerable to a code injection. This can be used to open a reverse shell. This code path was maintained for compatibility reasons as the maintainers had several test cases where numpy expressions were used as arguments. However, given that the tool i…

CVSS: 7.8 · CWE: CWE-94

View on NVD

Buffer overflow in the array_from_pyobj function of fortranobject.c in NumPy < 1.19, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally)

CVSS: 5.5 · CWE: CWE-120

View on NVD

Null Pointer Dereference vulnerability exists in numpy.sort in NumPy < and 1.19 in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privile…

CVSS: 5.3 · CWE: CWE-476

View on NVD

An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. NOTE: the vendor states that this reported code behavior is "completely harmless."

CVSS: 5.3 · CWE: CWE-697

View on NVD

A Buffer Overflow vulnerability exists in NumPy 1.9.x in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. NOTE: The vendor does not agree this is a vulneraility; In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely al…

CVSS: 5.3 · CWE: CWE-120

View on NVD

An issue was discovered in NumPy before 1.16.3. It uses the pickle Python module unsafely, which allows remote attackers to execute arbitrary code via a crafted serialized object, as demonstrated by a numpy.load call. NOTE: third parties dispute this issue because it is a behavior that might have legitimate applications in (for example) loading serialized Python object arrays from trusted and aut…

CVSS: 9.8 · CWE: CWE-502

View on NVD

(1) core/tests/test_memmap.py, (2) core/tests/test_multiarray.py, (3) f2py/f2py2e.py, and (4) lib/tests/test_io.py in NumPy before 1.8.1 allow local users to write to arbitrary files via a symlink attack on a temporary file.

CVSS: 5.5 · CWE: CWE-59

View on NVD

__init__.py in f2py in NumPy before 1.8.1 allows local users to write to arbitrary files via a symlink attack on a temporary file.

CVSS: 5.5 · CWE: CWE-20

View on NVD

The numpy.pad function in Numpy 1.13.1 and older versions is missing input validation. An empty list or ndarray will stick into an infinite loop, which can allow attackers to cause a DoS attack.

CVSS: 7.5 · CWE: CWE-835

View on NVD