GroupNormalization (CPU)#

  • Domain: ai.onnx

  • Device: CPU

  • Kernel: onnx_light_cpu::GroupNormalization

  • Supported types: FLOAT, DOUBLE, FLOAT16, BFLOAT16

  • Opset bounds: since opset 18

Operator support#

The standard ONNX schema and any standard shape inference come from onnx-light. No additional shape, peak-memory, fusion-pattern, or gradient implementation is registered by onnx-light-cpu for this operator.