标准化归一化 batch norm, layer norm, group norm, instance norm
LayerNormalization-EXPLAINED(inTransformerNeuralNetworks)LayerNormalization-EXPLAINED(inTransformerNeuralNetworks)0~4min:什么是multi-headattention5~7min:layernorm图示7~9min:公式举例layernorm9:54-end:layernorm的