Gahow Wang
445e491123
third_party/vllm/ now tracked in git for direct patch management.
Based on vLLM v0.18.1 release with one patch applied:
vllm/v1/core/sched/scheduler.py:
Replace fatal assert with graceful skip when KV transfer callback
arrives for an already-aborted request during PD disaggregated serving.
Future vLLM modifications should be made directly in third_party/vllm/
and committed normally. The patches/ directory is kept as documentation
of what changed from upstream.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
43 lines
888 B
Python
43 lines
888 B
Python
# SPDX-License-Identifier: Apache-2.0
|
|
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
|
|
|
"""vLLM Attention Benchmarking Suite."""
|
|
|
|
from .batch_spec import (
|
|
BatchRequest,
|
|
format_batch_spec,
|
|
get_batch_stats,
|
|
parse_batch_spec,
|
|
reorder_for_flashinfer,
|
|
split_by_type,
|
|
)
|
|
from .common import (
|
|
BenchmarkConfig,
|
|
BenchmarkResult,
|
|
MockLayer,
|
|
ResultsFormatter,
|
|
get_attention_scale,
|
|
is_mla_backend,
|
|
setup_mla_dims,
|
|
)
|
|
|
|
__all__ = [
|
|
# Batch specification
|
|
"BatchRequest",
|
|
"parse_batch_spec",
|
|
"format_batch_spec",
|
|
"reorder_for_flashinfer",
|
|
"split_by_type",
|
|
"get_batch_stats",
|
|
# Benchmarking infrastructure
|
|
"BenchmarkConfig",
|
|
"BenchmarkResult",
|
|
"ResultsFormatter",
|
|
# Mock objects
|
|
"MockLayer",
|
|
# Utilities
|
|
"setup_mla_dims",
|
|
"get_attention_scale",
|
|
"is_mla_backend",
|
|
]
|