update

llm/run_pretrain.py

@@ -20,6 +20,7 @@
 from typing import List, Optional
 
 import paddle
+from paddle.io.reader import use_pinned_memory
 
 from paddlenlp.data.causal_dataset import (
     build_train_valid_test_datasets,
@@ -47,7 +48,6 @@
 from paddlenlp.utils.batch_sampler import DistributedBatchSampler
 from paddlenlp.utils.log import logger
 from paddlenlp.utils.tools import get_env_device
-from paddle.io.reader import use_pinned_memory
 
 # Pretaining Environment Variables to support sharding stage1 overlap optimization.
 os.environ["USE_CASUAL_MASK"] = "True"
@@ -403,6 +403,10 @@ def main():
     else:
         model_args, data_args, training_args = parser.parse_args_into_dataclasses()
 
+    print("--888--" * 100)
+    print("training_args:", training_args)
+    print("--888--" * 100)
+
     if training_args.enable_linear_fused_grad_add:
         from fused_layers import mock_layers
 
@@ -499,6 +503,14 @@ def main():
         config.seq_length % config.context_parallel_degree == 0
     ), f"seq_length:{config.seq_length} must be divisible by context_parallel_degree {config.context_parallel_degree}"
 
+    if training_args.sharding_parallel_config is not None:
+        # for stage1 overlap optimization
+        if (
+            "enable_stage1_allgather_overlap" in training_args.sharding_parallel_config
+            or "enable_stage1_broadcast_overlap" in training_args.sharding_parallel_config
+        ):
+            use_pinned_memory(False)
+
     if get_env_device() == "xpu" and training_args.gradient_accumulation_steps > 1:
         try:
             from paddle_xpu.layers.nn.linear import LinearConfig  # noqa: F401
@@ -635,5 +647,4 @@ def main():
 
 
 if __name__ == "__main__":
-    use_pinned_memory(False)
     main()