More data in benchmarking (#41848)

remi-or · web-flow · commit ff0f7d649858 · 2025-11-03T18:05:26.000+01:00
* Reduce scope of cross-generate

* Rm generate_sall configs

* Workflow benchmarks more

* Prevent crash when FA is not installed
diff --git a/.github/workflows/benchmark.yml b/.github/workflows/benchmark.yml
@@ -52,7 +52,7 @@ jobs:
             commit_id=$GITHUB_SHA
           fi
           commit_msg=$(git show -s --format=%s | cut -c1-70)
-          python3 benchmark_v2/run_benchmarks.py -b 32 -s 128 -n 256 --branch-name "$BRANCH_NAME" --commit-id "$commit_id" --commit-message "$commit_msg" --model-id "$MODEL_ID" --log-level INFO --push-result-to-dataset "$DATASET_ID"
+          python3 benchmark_v2/run_benchmarks.py -b 32 -s 128 -n 256 --cross-generate --branch-name "$BRANCH_NAME" --commit-id "$commit_id" --commit-message "$commit_msg" --model-id "$MODEL_ID" --log-level INFO --push-result-to-dataset "$DATASET_ID"
         env:
           HF_TOKEN: ${{ secrets.HF_HUB_READ_TOKEN }}
           PUSH_TO_HUB_TOKEN: ${{ secrets.PUSH_TO_HUB_TOKEN }}
diff --git a/benchmark_v2/framework/benchmark_config.py b/benchmark_v2/framework/benchmark_config.py
@@ -3,6 +3,8 @@
 import logging
 from typing import Any
 
+from transformers.utils.import_utils import is_flash_attn_2_available
+
 
 KERNELIZATION_AVAILABLE = False
 try:
@@ -18,6 +20,16 @@
 class BenchmarkConfig:
     """Configuration for a single benchmark scenario."""
 
+    all_attn_implementations = [
+        ("flash_attention_2", None),
+        ("eager", None),
+        ("sdpa", "math"),
+        ("sdpa", "flash_attention"),
+        ("flex_attention", None),
+    ]
+
+    all_compiled_modes = [None, "default", "reduce-overhead", "max-autotune", "max-autotune-no-cudagraphs"]
+
     def __init__(
         self,
         warmup_iterations: int = 5,
@@ -59,6 +71,13 @@ def __init__(
     def check_validity(self, skip_validity_check: bool = False) -> None:
         if skip_validity_check:
             return
+        # Check FA is installed
+        if self.attn_implementation == "flash_attention_2" and not is_flash_attn_2_available():
+            logger.warning(
+                "Flash attention does not support compile mode. Defaulting to SDPA w/ flash attention backend."
+            )
+            self.attn_implementation = "sdpa"
+            self.sdpa_backend = "flash_attention"
         # Flash attention does not support compile mode, so we turn it off # FIXME: it would be better to support it
         is_fa = self.attn_implementation == "flash_attention_2"
         is_fa |= self.attn_implementation == "sdpa" and self.sdpa_backend == "flash_attention"
@@ -163,34 +182,6 @@ def cross_generate_configs(
     return configs
 
 
-def generate_all_configs(
-    warmup_iterations: int = 5,
-    measurement_iterations: int = 20,
-    batch_size: int = 1,
-    sequence_length: int = 128,
-    num_tokens_to_generate: int = 128,
-    gpu_monitoring: bool = True,
-) -> list[BenchmarkConfig]:
-    all_attn_implementations = [
-        ("flash_attention_2", None),
-        ("eager", None),
-        ("sdpa", "math"),
-        ("sdpa", "flash_attention"),
-        ("flex_attention", None),
-    ]
-    return cross_generate_configs(
-        attn_impl_and_sdpa_backend=all_attn_implementations,
-        compiled_mode=[None, "default", "reduce-overhead", "max-autotune", "max-autotune-no-cudagraphs"],
-        kernelized=[False, KERNELIZATION_AVAILABLE],
-        warmup_iterations=warmup_iterations,
-        measurement_iterations=measurement_iterations,
-        batch_size=batch_size,
-        sequence_length=sequence_length,
-        num_tokens_to_generate=num_tokens_to_generate,
-        gpu_monitoring=gpu_monitoring,
-    )
-
-
 def generate_main_configs(
     warmup_iterations: int = 5,
     measurement_iterations: int = 20,
diff --git a/benchmark_v2/run_benchmarks.py b/benchmark_v2/run_benchmarks.py
@@ -23,7 +23,12 @@
 import sys
 import uuid
 
-from framework.benchmark_config import BenchmarkConfig, generate_all_configs, generate_main_configs
+from framework.benchmark_config import (
+    KERNELIZATION_AVAILABLE,
+    BenchmarkConfig,
+    cross_generate_configs,
+    generate_main_configs,
+)
 from framework.benchmark_runner import BenchmarkRunner
 
 
@@ -82,7 +87,10 @@
     # If there is only one (batch_size, sequence_length, num_tokens_to_generate), we benchmark across configs
     elif len(args.batch_size) * len(args.sequence_length) * len(args.num_tokens_to_generate) == 1:
         if args.cross_generate:
-            benchmark_configs = generate_all_configs(
+            benchmark_configs = cross_generate_configs(
+                attn_impl_and_sdpa_backend=BenchmarkConfig.all_attn_implementations,
+                compiled_mode=[None, "default"],  # usually there is not much to gain by compiling with other modes
+                kernelized=[False, KERNELIZATION_AVAILABLE],
                 warmup_iterations=args.warmup,
                 measurement_iterations=args.iterations,
                 batch_size=args.batch_size[0],
