TensorPlay
latest (dev)
Copy
View Markdown

Latest development documentation · Updated 2026-10-08

Schedule1F1B

class tensorplay.distributed.pipelining.Schedule1F1B(stage: Any, n_microbatches: int, loss_fn: Any = None, args_chunk_spec: Any = None, kwargs_chunk_spec: Any = None, output_merge_spec: Any = None, scale_grads: bool = True)[source]

Overlap forward and backward microbatches after warmup.

On this page

Ask DeepWiki