AIMET is Qualcomm's toolkit for quantizing and compressing trained neural networks. It provides post-training quantization, quantization-aware training and pruning APIs for PyTorch and ONNX models.
2748quantization,deep-learning,compressionAIMET is Qualcomm's toolkit for quantizing and compressing trained neural networks. It provides post-training quantization, quantization-aware training and pruning APIs for PyTorch and ONNX models.