ONNX Untrusted Model Repository Warnings Suppressed by silent=True in onnx.hub.load() — Silent Supply-Chain Attack

A vulnerability in the `download_model` function of the onnx/onnx framework, before and including version 1.16.1, allows for arbitrary file overwrite due to inadequate prevention…

ONNX: Malicious ONNX models can crash servers by exploiting unprotected object settings.

ONNX: Path Traversal via Symlink

ONNX symlink-following and path-traversal arbitrary file write

ONNX: Denial of service via unchecked expansion of packed sub-byte tensors in ONNX Python APIs

ONNX: Null Pointer Dereference in ScanInferenceFunction Due to Missing Required Attribute

ONNX: Out-of-bounds read in ConvTranspose shape inference via undersized weight rank

ONNX: Out-of-Bounds Read in RNN/GRU/LSTM Shape Inference (Old Opsets, Missing Rank Check)

ONNX: Null Pointer Dereference in Upsample Version Converter Adapter (Zero Inputs)

ONNX: External Data Symlink Traversal

ONNX: Arbitrary File Read via ExternalData Hardlink Bypass in ONNX load

onnx onnxruntime old.cc convPoolShapeInference_opset19 out-of-bounds

ONNX: Heap-Buffer-Overflow READ in Gemm Version Converter Adapter via Undersized Input Shape