From 84be03385cbd04975ffaf069a40844c81bddc4a1 Mon Sep 17 00:00:00 2001 From: liyurui Date: Wed, 9 Oct 2024 19:42:35 +0800 Subject: [PATCH] support best unbalaced pp scheduler --- paddlenlp/trainer/training_args.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/paddlenlp/trainer/training_args.py b/paddlenlp/trainer/training_args.py index e1f453806ac8..596332fc567d 100644 --- a/paddlenlp/trainer/training_args.py +++ b/paddlenlp/trainer/training_args.py @@ -1074,6 +1074,7 @@ def __post_init__(self): "enable_clear_every_step_cache", "enable_overlap_p2p_comm", "disable_batch_p2p_comm", + "best_unbalanced_scheduler", ]: raise ValueError( f"Found unknown pipeline mode config {x}, accpet config is disable_p2p_cache_shape, disable_partial_send_recv." @@ -1112,6 +1113,7 @@ def __post_init__(self): "overlap_p2p_comm": "enable_overlap_p2p_comm" in pipeline_parallel_config, "clear_every_step_cache": "enable_clear_every_step_cache" in pipeline_parallel_config, "use_batch_p2p_comm": "disable_batch_p2p_comm" not in pipeline_parallel_config, + "best_unbalanced_scheduler": "best_unbalanced_scheduler" in pipeline_parallel_config, } if dygraph_pp_configs["dp_comm_overlap"]: raise ValueError("overlap has accuracy issue") # TODO: fix `overalap` + `delay_scale` issue