mindspore.nn.AdaSumByGradWrapCell

class mindspore.nn.AdaSumByGradWrapCell(optimizer)[source]

Enable the adasum in “auto_parallel/semi_auto_parallel” mode. The implementation of the Adaptive Summation (AdaSum) algorithm is calculated by gradients. See the paper AdaSum: Scaling Distributed Training with Adaptive Summation.

\[\begin{split}\begin{array}{ll} w_{t+1}=w_{t} - \alpha \cdot Adasum(g_{1}, g_{2}) \\ w_{t+1}=w_{t} - \alpha \cdot [(1 - \frac{g_2^{T}\cdot g_1}{2\cdot \left \| g_1 \right \|^2 })\cdot g_1 + (1 - \frac{g_1^{T}\cdot g_2}{2\cdot \left \| g_2 \right \|^2 })\cdot g_2] \\ \end{array}\end{split}\]

In this implementation, \(g\) represents the gradient of the weights, and the subscripts represent different devices in the data-parallel dimension.

Note

When using AdaSum, the number of traning cards needs to be a power of 2 and at least 16 cards are required. Currently, the optimizer sharding and pipeline parallel is not supported when using AdaSum. It is recommended to using AdaSumByGradWrapCell in semi auto parallel/auto parallel mode, and in data parallel mode, we recommend to using mindspore.boost to applying AdaSum.

Parameters

optimizer (Union[Cell]) – Optimizer for updating the weights. The construct function of the optimizer requires only one input.

Inputs:

  • grads (Tuple(Tensor)) - Tuple of gradients, same with the input of passed optimizer.

Raises

  • RuntimeError – If parallel_mode uses stand_alone mode, AdaSum only supports use in distributed scenarios.

  • RuntimeError – If the optimizer parallel is used when using AdaSum.

  • RuntimeError – If the pipeline parallel is used when using AdaSum.

  • RuntimeError – If device_num is not a power of 2, or less than 16.

Supported Platforms:

Ascend GPU

Examples

>>> from mindspore import nn
>>> from mindspore.nn import AdaSumByGradWrapCell
>>> net = Net()
>>> optim = AdaSumByGradWrapCell(nn.Momentum(params=net.trainable_params(), learning_rate=0.1, momentum=0.9))
>>> loss = nn.SoftmaxCrossEntropyWithLogits()
>>> model = Model(net, loss_fn=loss, optimizer=optim, metrics=None)