Figure 3. Using gradual pruning and dynamic quantization to control the accuracy-efficiency trade-off (IMAGE)

Tokyo Institute of Technology

Figure 3. Using gradual pruning and dynamic quantization to control the accuracy-efficiency trade-off

Caption

The trained model was pruned by removing the lowest weight in each channel. Only one element remains after 8 rounds of pruning (pruned to 1/9). Each of the pruned models is then subjected to dynamic quantization.

Credit

Hot Chips

Usage Restrictions

Nothing

License

Original content

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.