CVE-2026-112304

docTR through 1.1.0 contains an unsafe deserialization vulnerability in loadpretrainedparams that loads checkpoints with torch.load without weightsonly=True. Attackers who control a local or URL-supplied checkpoint can embed a malicious pickle payload to execute Python code with the loading...

CVE-2026-112304

docTR through 1.1.0 contains an unsafe deserialization vulnerability in loadpretrainedparams that loads checkpoints with torch.load without weightsonly=True. Attackers who control a local or URL-supplied checkpoint can embed a malicious pickle payload to execute Python code with the loading...

EUVD-2026-104396

docTR through 1.1.0 contains an unsafe deserialization vulnerability in loadpretrainedparams that loads checkpoints with torch.load without weightsonly=True. Attackers who control a local or URL-supplied checkpoint can embed a malicious pickle payload to execute Python code with the loading...

CVE-2026-112304 docTR through 1.1.0 Unsafe Deserialization RCE via load_pretrained_params()

docTR through 1.1.0 contains an unsafe deserialization vulnerability in loadpretrainedparams that loads checkpoints with torch.load without weightsonly=True. Attackers who control a local or URL-supplied checkpoint can embed a malicious pickle payload to execute Python code with the loading...

CVE-2026-112304 docTR through 1.1.0 Unsafe Deserialization RCE via load_pretrained_params()

docTR through 1.1.0 contains an unsafe deserialization vulnerability in loadpretrainedparams that loads checkpoints with torch.load without weightsonly=True. Attackers who control a local or URL-supplied checkpoint can embed a malicious pickle payload to execute Python code with the loading...

CVE-2026-112304

The docTR library (versions through 1.1.0 ) is vulnerable to Remote Code Execution (RCE) via unsafe deserialization in the load_pretrained_params() function. The root cause is the use of torch.load without setting weights_only=True, which allows attackers to execute arbitrary Python code by provi...

CVE-2026-112304: Deserialization of Untrusted Data

docTR through 1.1.0 contains an unsafe deserialization vulnerability in loadpretrainedparams that loads checkpoints with torch.load without weightsonly=True. Attackers who control a local or URL-supplied checkpoint can embed a malicious pickle payload to execute Python code with the loading...

Mindee docTR deserialization injection vulnerability

Mindee docTR 1.1.0 and earlier versions have a deserialization injection vulnerability. This vulnerability stems from the use of torch.load to load checkpoints in loadpretrainedparams, where weightsonly=True is not set. This allows attackers to inject malicious pickle payloads by controlling loca...