Lucene search
+L

99 matches found

NVD
NVD
added 2 days ago5 views

CVE-2026-58659

PyTorch Lightning through 2.6.5, fixed in commit d710d68, contains a remote code execution vulnerability in the loadstate function that imports and executes attacker-controlled module names from checkpoint instantiator hyperparameters. Attackers can craft malicious checkpoint files that bypass...

8.4CVSS0.00235EPSS
Exploits0References4
EUVD
EUVD
added 2 days ago6 views

EUVD-2026-44751

PyTorch Lightning through 2.6.5, fixed in commit d710d68, contains a remote code execution vulnerability in the loadstate function that imports and executes attacker-controlled module names from checkpoint instantiator hyperparameters. Attackers can craft malicious checkpoint files that bypass...

8.4CVSS6.6AI score0.00235EPSS
Exploits0References4
OSV
OSV
added 2026/07/07 4:3 p.m.6 views

PYSEC-2026-1856 PyTorch Vulnerable to Remote Code Execution via Untrusted Checkpoint Files

Summary A vulnerability in PyTorch's weightsonly unpickler allows an attacker to craft a malicious checkpoint file .pth that, when loaded with torch.load..., weightsonly=True, can corrupt memory and potentially lead to arbitrary code execution. Vulnerability Details The weightsonly=True unpickler...

8.8CVSS7.4AI score0.00695EPSS
Exploits1References8
OSV
OSV
added 2026/06/29 11:50 a.m.7 views

PYSEC-2026-405 Ludwig framework is vulnerable to insecure deserialization in its model serving component

The Ludwig framework thru 0.10.4 is vulnerable to insecure deserialization CWE-502 in its model serving component. When starting a model server with the ludwig serve command, the framework loads model weight files using torch.load without enabling the security-restrictive weightsonly=True...

9.8CVSS6.5AI score0.00497EPSS
Exploits0References5
RedhatCVE
RedhatCVE
added 2026/06/05 7:45 p.m.13 views

CVE-2026-31229

A flaw was found in the Adversarial Robustness Toolbox ART, specifically within its Kubeflow component. This vulnerability, categorized as insecure deserialization CWE-502, occurs when loading model weights without proper security restrictions. A remote attacker can exploit this by uploading a...

9.8CVSS6.6AI score0.006EPSS
Exploits0References5
RedhatCVE
RedhatCVE
added 2026/06/05 7:45 p.m.10 views

CVE-2026-31224

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability CWE-502 in the MultitaskClassifier.load method of the MultitaskClassifier class. The method loads model weight files using torch.load without enabling the security-restrictive weightsonly=True parameter. This...

8.8CVSS6.3AI score0.00392EPSS
Exploits0References1
RedhatCVE
RedhatCVE
added 2026/05/15 7:57 p.m.11 views

CVE-2026-31222

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability CWE-502 in the Trainer.load method of the Trainer class. The method loads model checkpoint files using torch.load without enabling the security-restrictive weightsonly=True parameter. This default behavior allows...

8.8CVSS6.3AI score0.00392EPSS
Exploits0References1
RedhatCVE
RedhatCVE
added 2026/05/15 7:57 p.m.12 views

CVE-2026-31219

The loadmodel function in the neuralmagictraining.py script of the optimate project in commit a6d302f912b481c94370811af6b11402f51d377f 2024-07-21 is vulnerable to insecure deserialization CWE-502. When a user provides a single model file path e.g., .pt or .pth via the --model command-line argumen...

8.8CVSS6.3AI score0.00559EPSS
Exploits0References1
RedhatCVE
RedhatCVE
added 2026/05/15 1:57 a.m.9 views

CVE-2026-31232

The CosyVoice project thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e 2025-30-21 contains an insecure deserialization vulnerability CWE-502 in its model loading process. When loading model files .pt from a user-specified directory via the --modeldir argument, the code uses torch.load without...

8.8CVSS6.5AI score0.00458EPSS
Exploits0References1
RedhatCVE
RedhatCVE
added 2026/05/12 8:21 p.m.14 views

CVE-2026-31249

CosyVoice thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e 2025-30-21 contains an insecure deserialization vulnerability CWE-502 in its makeparquetlist.py data processing tool. The script loads PyTorch .pt files utterance embeddings, speaker embeddings, speech tokens using torch.load without...

7.3CVSS6.1AI score0.0021EPSS
Exploits0References1
RedhatCVE
RedhatCVE
added 2026/05/12 8:21 p.m.11 views

CVE-2026-31250

CosyVoice thru commit 6e01309e01bc93bbeb83bdd996b1182a81aaf11e 2025-30-21 contains an insecure deserialization vulnerability CWE-502 in its averagemodel.py model averaging tool. The script loads PyTorch checkpoint files epoch.pt for model averaging using torch.load without enabling the...

7.3CVSS6.1AI score0.00222EPSS
Exploits0References1
OSV
OSV
added 2026/05/12 6:30 p.m.7 views

GHSA-XP5Q-5Q7G-Q26R Ludwig framework is vulnerable to insecure deserialization in its model serving component

The Ludwig framework thru 0.10.4 is vulnerable to insecure deserialization CWE-502 in its model serving component. When starting a model server with the ludwig serve command, the framework loads model weight files using torch.load without enabling the security-restrictive weightsonly=True...

9.8CVSS6.3AI score0.00497EPSS
Exploits0References3
EUVD
EUVD
added 2026/05/12 6:30 p.m.14 views

EUVD-2026-29503

The loadmodel function in the neuralmagictraining.py script of the optimate project in commit a6d302f912b481c94370811af6b11402f51d377f 2024-07-21 is vulnerable to insecure deserialization CWE-502. When a user provides a single model file path e.g., .pt or .pth via the --model command-line argumen...

6.3AI score0.00559EPSS
Exploits0References3
OSV
OSV
added 2026/05/12 6:30 p.m.10 views

GHSA-78CP-F66X-QMH5 Snorkel Trainer.load uses an unsafe torch.load

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability CWE-502 in the Trainer.load method of the Trainer class. The method loads model checkpoint files using torch.load without enabling the security-restrictive weightsonly=True parameter. This default behavior allows...

8.8CVSS6.3AI score0.00392EPSS
Exploits0References3
NVD
NVD
added 2026/05/12 6:16 p.m.10 views

CVE-2026-31238

The Ludwig framework thru 0.10.4 is vulnerable to insecure deserialization CWE-502 in its model serving component. When starting a model server with the ludwig serve command, the framework loads model weight files using torch.load without enabling the security-restrictive weightsonly=True...

9.8CVSS0.00497EPSS
Exploits0References2
NVD
NVD
added 2026/05/12 6:16 p.m.11 views

CVE-2026-31239

The mamba language model framework thru 2.2.6 is vulnerable to insecure deserialization CWE-502 when loading pre-trained models from HuggingFace Hub. The MambaLMHeadModel.frompretrained method uses torch.load to load the pytorchmodel.bin weight file without enabling the security-restrictive...

9.8CVSS0.00409EPSS
Exploits0References2
NVD
NVD
added 2026/05/12 6:16 p.m.11 views

CVE-2026-31229

The Adversarial Robustness Toolbox ART thru 1.20.1 contains an insecure deserialization vulnerability CWE-502 in its Kubeflow component's model loading functionality. When loading model weights from a file e.g., model.pt during robustness evaluation, the code uses torch.load without the...

9.8CVSS0.006EPSS
Exploits0References2
NVD
NVD
added 2026/05/12 4:16 p.m.14 views

CVE-2026-31222

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability CWE-502 in the Trainer.load method of the Trainer class. The method loads model checkpoint files using torch.load without enabling the security-restrictive weightsonly=True parameter. This default behavior allows...

8.8CVSS0.00392EPSS
Exploits0References2
NVD
NVD
added 2026/05/12 4:16 p.m.14 views

CVE-2026-31224

The snorkel library thru v0.10.0 contains an insecure deserialization vulnerability CWE-502 in the MultitaskClassifier.load method of the MultitaskClassifier class. The method loads model weight files using torch.load without enabling the security-restrictive weightsonly=True parameter. This...

8.8CVSS0.00392EPSS
Exploits0References2
NVD
NVD
added 2026/05/12 4:16 p.m.12 views

CVE-2026-31218

The loadmodel function in the neuralmagictraining.py script of the optimate project in commit a6d302f912b481c94370811af6b11402f51d377f 2024-07-21 is vulnerable to insecure deserialization CWE-502. When loading a model state dictionary from a statedict.pt file via torch.load, the function does not...

8.8CVSS0.00559EPSS
Exploits0References2
Rows per page
Query Builder