**TOSA** (Tensor Operator Set Architecture) is a set of tensor operators that solves issues with the proliferation of machine learning frameworks and operators that make common implementations difficult.
TOSA has the following goals:
• A minimal and stable set of tensor-level operators to which machine learning framework operators can be reduced.
• Full support for both quantized integer and floating-point content.
• Precise functional description of the behavior of every operator, including the treatment of their numerical behavior in the case of precision, saturation, scaling, etc. as required by quantized datatypes.
• Agnostic to any single high-level framework, compiler backend stack or particular target.
• The detailed functional and numerical description enables precise code construction for a diverse range of targets – SIMD CPUs, GPUs and custom hardware like NPUs/TPUs.
The TOSA specification is located at [[ https://git.mlplatform.org/tosa/specification.git/ | https://git.mlplatform.org/tosa/specification.git/ ]]