CVE-2025-23265

NVIDIA Megatron-LM vulnerable in a Python component to Arbitrary Code Injection via a malicious file, potentially enabling Code Execution, Privilege Escalation, Information Disclosure and Data Tampering. Remediation per NVIDIA advisory: upgrade to Megatron-LM 0.12.1 or later (from NVIDIA’s bullet...

CVE-2025-23306

CVE-2025-23306 affects NVIDIA Megatron-LM, specifically the megatron/training/arguments.py component. The vulnerability permits code injection through malicious input, with potential outcomes including code execution, privilege escalation, information disclosure, and data tampering. Public refere...

CVE-2025-23305

CVE-2025-23305 affects NVIDIA Megatron-LM across all platforms, describing a vulnerability in the tools component that enables code injection. A successful exploit could lead to code execution, privilege escalation, information disclosure, and data tampering. The Red Hat entry and NVIDIA security...

CVE-2025-23264

CVE-2025-23264 affects NVIDIA Megatron-LM across platforms due to a vulnerability in a Python component that allows code injection via a malicious file. Exploitation may lead to Code Execution, Privilege Escalation, Information Disclosure, and Data Tampering. Public sources (NVD/Red Hat) describe...

CVE-2025-23353

CVE-2025-23353 affects NVIDIA Megatron-LM across platforms due to a vulnerability in the msdp preprocessing script. The issue enables code injection that can lead to code execution, privilege escalation, information disclosure, and data tampering. Public details identify the affected component as...

CVE-2025-23354

The CVE-2025-23354 issue affects NVIDIA Megatron-LM, specifically the ensemble_classifer script, with a code injection vulnerability that attacker-supplied data can trigger. The vulnerability may enable code execution, privilege escalation, information disclosure, and data tampering. Affected com...

CVE-2026-24152

NVIDIA Megatron-LM is affected by CVE-2026-24152 through a vulnerability in checkpoint loading that could allow an attacker to cause remote code execution by convincing a user to load a maliciously crafted file. The NVIDIA security bulletin states this vulnerability could lead to code execution, ...

CVE-2025-23349

CVE-2025-23349 affects NVIDIA Megatron-LM. The vulnerability exists in the tasks/orqa/unsupervised/nq.py component and can enable code injection, potentially leading to code execution, privilege escalation, information disclosure, and data tampering. Public documents consistently cite Megatron-LM...

CVE-2025-33247

NVIDIA Megatron LM is affected by CVE-2025-33247 due to a vulnerability in quantization configuration loading that could allow remote code execution. The security bulletin states this could lead to code execution, elevation of privileges, information disclosure, and data tampering. Affected produ...

CVE-2025-33248

NVIDIA Megatron-LM contains a vulnerability in the hybrid conversion script where an attacker may cause remote code execution (RCE) by convincing a user to load a maliciously crafted file (CVE-2025-33248). Impact includes code execution, privilege escalation, information disclosure, and data tamp...

CVE-2026-24151

NVIDIA Megatron-LM contains a vulnerability in inferencing: an attacker may cause remote code execution by convincing a user to load a malicious input. This CVE-2026-24151 is addressed in the NVIDIA security bulletin, which recommends updating to Megatron-LM 0.15.3 or later. The bulletin lists CV...

CVE-2026-24150

NVIDIA Megatron-LM is affected by CVE-2026-24150 through a vulnerability in checkpoint loading that could allow an attacker to induce remote code execution by convincing a user to load a maliciously crafted file. The advisory indicates the vulnerability is addressed by updating Megatron LM to ver...

CVE-2025-23348

CVE-2025-23348 affects NVIDIA Megatron-LM across platforms, with a vulnerability in the pretrain_gpt script where malicious input data may trigger code injection, potentially enabling code execution, privilege escalation, information disclosure, and data tampering. Documents consistently identify...