Skip to content

optimize clear_low_latency_buffer when mixed_infer #4170

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: feature/experimental_feature_20250908
Choose a base branch
from
Open

optimize clear_low_latency_buffer when mixed_infer #4170

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
09e4e91
optimize clear_low_latency_buffer when mixed_infer
carryyu Sep 17, 2025
1f2fefd
optimize clear_low_latency_buffer when mixed_infer
carryyu Sep 18, 2025
d0aebab
Merge branch 'feature/experimental_feature_20250908' into opt_clear_exp
carryyu Sep 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 5 additions & 3 deletions fastdeploy/model_executor/layers/moe/fused_moe_backend_base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -160,12 +160,14 @@ def apply(
"""
if layer.ep_size > 1:
if layer.fd_config.parallel_config.moe_phase.phase == "prefill":
if layer.fd_config.parallel_config.splitwise_role == "mixed":
self.ep_prefill_runner.clean_low_latency_buffer()
return self.apply_ep_prefill(layer, x, gate)
else:
if layer.fd_config.parallel_config.splitwise_role == "mixed":
self.ep_decoder_runner.clean_low_latency_buffer()
if layer.layer_idx == 0 and (
layer.fd_config.parallel_config.moe_phase.last_phase
!= layer.fd_config.parallel_config.moe_phase.phase
):
self.ep_decoder_runner.clean_low_latency_buffer()
return self.apply_ep_decode(layer, x, gate)
else:
return self.apply_tp(layer, x, gate)
Expand Down
1 change: 1 addition & 0 deletions fastdeploy/worker/dcu_model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,6 +68,7 @@ def initialize_forward_meta(self):
prefill_exists = self.exist_prefill()
paddle.distributed.all_gather_object(only_decode_batch_list, not prefill_exists)
only_decode_batch = all(only_decode_batch_list)
self.fd_config.parallel_config.moe_phase.last_phase = self.fd_config.parallel_config.moe_phase.phase
self.fd_config.parallel_config.moe_phase.phase = "decode" if only_decode_batch else "prefill"

self.forward_meta.step_use_cudagraph = (
Expand Down
1 change: 1 addition & 0 deletions fastdeploy/worker/gpu_model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -975,6 +975,7 @@ def initialize_forward_meta(self):
prefill_exists = self.exist_prefill()
paddle.distributed.all_gather_object(only_decode_batch_list, not prefill_exists)
only_decode_batch = all(only_decode_batch_list)
self.fd_config.parallel_config.moe_phase.last_phase = self.fd_config.parallel_config.moe_phase.phase
self.fd_config.parallel_config.moe_phase.phase = "decode" if only_decode_batch else "prefill"

self.forward_meta.step_use_cudagraph = (
Expand Down
1 change: 1 addition & 0 deletions fastdeploy/worker/metax_model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -910,6 +910,7 @@ def initialize_forward_meta(self):
prefill_exists = self.exist_prefill()
paddle.distributed.all_gather_object(only_decode_batch_list, not prefill_exists)
only_decode_batch = all(only_decode_batch_list)
self.fd_config.parallel_config.moe_phase.last_phase = self.fd_config.parallel_config.moe_phase.phase
self.fd_config.parallel_config.moe_phase.phase = "decode" if only_decode_batch else "prefill"

self.forward_meta.step_use_cudagraph = (
Expand Down
Loading