Training After Migration
To run a migrated training script on an NPU, perform the following steps:
- Choose on the main menu.
- Go to the run configuration page. Figure 1 shows the parameter configuration. Click OK.
- Select the migrated training script.
- Specify ResNet-50 as the training network.
- Click
to start training.
Figure 2 shows the normal training process. Check that the training process is normal and the loss is converged.
If an error related to Protobuf is reported during training, install Protobuf of the corresponding version as prompted.
Parent topic: Migration Tool (PyTorch GPU2Ascend)

