[Bugfix] Allow to enable HSDP alone (#1567)

Signed-off-by: gcanlin <canlinguosdu@gmail.com>
This commit is contained in:
Canlin Guo authored and GitHub committed 2026-03-09 06:36:13 +00:00
1 parent d6871db6a3
commit 29357027b0
4 files changed
+182 -6

No files matched your search

+5 -4
View File
@@ -111,11 +111,12 @@ class DiffusionParallelConfig:
# 1. Standalone: when other parallelism is all 1, HSDP determines world_size
# 2. Combined: HSDP overlays on top of other parallelism
if self.use_hsdp:
if self.tensor_parallel_size > 1:
if self.tensor_parallel_size > 1 or self.data_parallel_size > 1:
raise ValueError(
"HSDP (use_hsdp=True) cannot be used with Tensor Parallelism "
f"(tensor_parallel_size={self.tensor_parallel_size}). "
"Set tensor_parallel_size=1 when using HSDP."
"HSDP (use_hsdp=True) cannot be used with TP or DP "
f"(tensor_parallel_size={self.tensor_parallel_size}, "
f"data_parallel_size={self.data_parallel_size}). "
"Set tensor_parallel_size=1 and data_parallel_size=1 when using HSDP."
)
if self.hsdp_shard_size == -1:
# Auto-calculate: use other_parallel_world_size as shard_size