Concept十亿次浮点运算
GFLOPs
A common unit for measuring the computational cost of neural network inference. One FLOP is one floating-point addition or multiplication, and the total number of operations a model requires to process each image frame is its computational cost.
Timeline (last 90 days)
Oct 16
GFLOPs is a commonly used unit for measuring the computational cost of neural network inference, and 0.18 GFLOPs means a single frame requires only about 180 million floating-point operations.
Unverified50%