Unverified Commit 5041df2d authored by Cheng Wan's avatar Cheng Wan Committed by GitHub
Browse files

Fix 7285 Merge Conflicts (#7403)

parent 256801e9
...@@ -1438,7 +1438,7 @@ class Scheduler( ...@@ -1438,7 +1438,7 @@ class Scheduler(
if need_dp_attn_preparation and not self.spec_algorithm.is_none(): if need_dp_attn_preparation and not self.spec_algorithm.is_none():
# In speculative decoding, prefill batches and decode batches cannot be processed in the same DP attention group. # In speculative decoding, prefill batches and decode batches cannot be processed in the same DP attention group.
# We prepare idle batches in advance to skip preparing decode batches when there are prefill batches in the group. # We prepare idle batches in advance to skip preparing decode batches when there are prefill batches in the group.
new_batch, _ = self.prepare_dp_attn_batch(new_batch) new_batch, _ = self.prepare_mlp_sync_batch(new_batch)
need_dp_attn_preparation = new_batch is None need_dp_attn_preparation = new_batch is None
if new_batch is not None: if new_batch is not None:
...@@ -1454,7 +1454,7 @@ class Scheduler( ...@@ -1454,7 +1454,7 @@ class Scheduler(
# Handle DP attention # Handle DP attention
if need_dp_attn_preparation: if need_dp_attn_preparation:
ret, _ = self.prepare_dp_attn_batch(ret) ret, _ = self.prepare_mlp_sync_batch(ret)
return ret return ret
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment