Add attention benchmarking tools (#26835)

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Claude <noreply@anthropic.com>

benchmarks/attention_benchmarks/__init__.py

@@ -0,0 +1,44 @@
+# 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,
+    MockModelConfig,
+    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",
+    "MockModelConfig",
+    # Utilities
+    "setup_mla_dims",
+    "get_attention_scale",
+    "is_mla_backend",
+]