Tuning in PyTorch-based Online Inference Scenarios

For the Atlas A2 training products/Atlas A2 inference products, tuning in online inference is not supported.

For the Atlas A3 training products/Atlas A3 inference products, tuning in online inference is not supported.

Operator tuning is supported in PyTorch-based online inference scenarios. The tuning process here is basically the same as that in PyTorch-based training scenarios. For details, see Offline Tuning in PyTorch-based Training Scenarios.

The difference between tuning in PyTorch-based online inference scenarios and tuning in PyTorch-based training scenarios is that the online inference script is executed for online inference, and the training script is executed for training.