Skip to content

[BugFix] Fix chunked prefill bugs in engine v1 #844

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

Merged
merged 2 commits into from
May 22, 2025
Merged

[BugFix] Fix chunked prefill bugs in engine v1 #844

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
418ed3d
[BugFix] Fix chunked prefill bugs
rjg-lyh May 14, 2025
f0dac90
allow deepseek models to enable chunked prefill on NPUs
rjg-lyh May 14, 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
3 changes: 3 additions & 0 deletions vllm_ascend/platform.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -204,6 +204,9 @@ def check_and_update_config(cls, vllm_config: VllmConfig) -> None:
"ascend_scheduler_config", None) is not None:
additional_scheduler_config = additional_config.get(
"ascend_scheduler_config")
if vllm_config.scheduler_config.enable_chunked_prefill:
additional_scheduler_config[
"enable_chunked_prefill"] = True
from vllm_ascend.core.schedule_config import \
AscendSchedulerConfig
ascend_scheduler_config = AscendSchedulerConfig.initialize_from_config(
Expand Down
14 changes: 11 additions & 3 deletions vllm_ascend/worker/model_runner_v1.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -120,6 +120,13 @@ def __init__(self, vllm_config: VllmConfig, device: torch.device):
self.max_num_tokens = self.scheduler_config.max_num_batched_tokens
self.max_num_reqs = self.scheduler_config.max_num_seqs

additional_config = vllm_config.additional_config
if additional_config and additional_config.get(
"ascend_scheduler_config", None) is not None:
self.use_v0_scheduler = True
else:
self.use_v0_scheduler = False

self.graph_block_tables = np.zeros(
(self.vllm_config.scheduler_config.max_num_seqs,
(self.model_config.max_model_len + self.block_size - 1) //
Expand Down Expand Up @@ -545,13 +552,14 @@ def _process_reqs(
block_offsets,
out=self.slot_mapping_np[:total_num_scheduled_tokens])

if self.chunked_prefill_enabled:
attn_state = AscendAttentionState.ChunkedPrefill
elif np.array_equal(self.seq_lens_np[:num_reqs], num_scheduled_tokens):
if np.array_equal(self.seq_lens_np[:num_reqs], num_scheduled_tokens):
attn_state = AscendAttentionState.PrefillNoCache
# We assume it is the decode stage, where prefill occurs but only one token is not hit in cache.
elif np.all(num_scheduled_tokens == 1):
attn_state = AscendAttentionState.DecodeOnly
# splitfuse
elif not self.use_v0_scheduler or self.chunked_prefill_enabled:
attn_state = AscendAttentionState.ChunkedPrefill
else:
attn_state = AscendAttentionState.PrefillCacheHit

Expand Down