set_device_sat_mode

Applicability

Description

Sets the process-level overflow/underflow mode for floating-point computation.

• Saturation mode: When overflow/underflow occurs during computing, the compute result is saturated as the floating-point extremum (±MAX).
• Inf/NaN mode: Complies with IEEE 754 and outputs the Inf/NaN compute result based on the definition.

For the Atlas A3 training products/Atlas A3 inference products, the overflow/underflow mode can be saturation or Inf/NaN. Retain the default Inf/NaN mode. The saturation mode is used only for compatibility with earlier versions and will not evolve in the future. In addition, the computing accuracy in this mode may be unreliable.

For the Atlas A2 training products/Atlas A2 inference products, the overflow/underflow mode can be saturation or Inf/NaN. Retain the default Inf/NaN mode. The saturation mode is used only for compatibility with earlier versions and will not evolve in the future. In addition, the computing accuracy in this mode may be unreliable.

For the Atlas training products, the default (and the only supported) mode is saturation mode.

Prototype

1

def set_device_sat_mode(mode)

Parameters

Returns

None

Restrictions

This API needs to be configured during running and called before the network script is executed.

Example

The following example applies only to the Atlas A3 training products/Atlas A3 inference productsAtlas A2 training products/Atlas A2 inference products. For other processors, you do not need to explicitly call this API.

1
2
3
4
5
6
7

import tensorflow as tf
from npu_bridge.npu_init import *

......
# The following API is called to set the overflow/underflow mode during network execution:
npu_plugin.set_device_sat_mode(1)
sess.run(xxx)