progress
Function
Performs pipeline training.
Prototype
1 | def progress(dataloader_iter: Iterator[In]) -> Out: |
Parameters
Parameter |
Type |
Mandatory/Optional |
Description |
|---|---|---|---|
dataloader_iter |
Iterator[In] |
Mandatory |
Dataset iterator. The iterator returns the Batch class used for table lookup and training. For details, see 1. |
Return Value
- Success: The model output is returned.
- Failure: An exception is thrown.
Example
1 | pipeline.progress(dataloader_iter) |
Parent topic: HybridTrainPipelineSparseDist