Remove compatibility restriction between Pipeline Parallelism and Mixed Chunked Prefill (#22920)
Co-authored-by: cyy <cy02433585@alibaba-inc.com>
This commit is contained in:
@@ -6465,10 +6465,8 @@ class ServerArgs:
|
||||
|
||||
if self.pp_size > 1:
|
||||
assert (
|
||||
self.disable_overlap_schedule
|
||||
and self.speculative_algorithm is None
|
||||
and not self.enable_mixed_chunk
|
||||
), "Pipeline parallelism is not compatible with overlap schedule, speculative decoding, mixed chunked prefill."
|
||||
self.disable_overlap_schedule and self.speculative_algorithm is None
|
||||
), "Pipeline parallelism is not compatible with overlap schedule, speculative decoding"
|
||||
|
||||
assert not (
|
||||
self.dp_size > 1 and self.nnodes != 1 and not self.enable_dp_attention
|
||||
|
||||
Reference in New Issue
Block a user