fix: auto backend in profiler

components/src/dynamo/profiler/profile_sla.py

@@ -33,7 +33,10 @@
 from dynamo.profiler.utils.config_modifiers.protocol import apply_dgd_overrides
 from dynamo.profiler.utils.defaults import SearchStrategy
 from dynamo.profiler.utils.dgd_generation import generate_dgd_config_with_planner
-from dynamo.profiler.utils.dgdr_v1beta1_types import DynamoGraphDeploymentRequestSpec
+from dynamo.profiler.utils.dgdr_v1beta1_types import (
+    BackendType,
+    DynamoGraphDeploymentRequestSpec,
+)
 from dynamo.profiler.utils.dgdr_validate import (
     run_gate_checks,
     validate_dgdr_for_profiler,
@@ -51,11 +54,23 @@
 
 logger = logging.getLogger(__name__)
 
+_CONCRETE_BACKENDS = ["trtllm", "sglang", "vllm"]
+
+
+def _check_auto_backend_support(model: str, system: str) -> bool:
+    """
+    Return True if *any* concrete backend is AIC-supported for this model/system.
+    TODO: move this function to AIC and handle partially supported model x backend x hardware
+    """
+    return any(
+        check_model_hardware_support(model, system, b) for b in _CONCRETE_BACKENDS
+    )
+
 
 def _extract_profiler_params(dgdr: DynamoGraphDeploymentRequestSpec) -> tuple:
     """Pull all profiler parameters from dgdr and log them."""
     model = dgdr.model
-    backend = dgdr.backend.value.lower()
+    backend = BackendType(dgdr.backend).value.lower()
     system = dgdr.hardware.gpuSku.lower()
     total_gpus = dgdr.hardware.totalGpus
     isl = dgdr.workload.isl
@@ -313,7 +328,10 @@ async def run_profile(
             picking_mode,
         ) = _extract_profiler_params(dgdr)
 
-        aic_supported = check_model_hardware_support(model, system, backend)
+        if backend == "auto":
+            aic_supported = _check_auto_backend_support(model, system)
+        else:
+            aic_supported = check_model_hardware_support(model, system, backend)
         run_gate_checks(dgdr, aic_supported, search_strategy, backend)
 
         (

components/src/dynamo/profiler/rapid.py

@@ -44,12 +44,19 @@ def _generate_dgd_from_pick(
     if best_config_df is None or best_config_df.empty:
         return None
 
+    row = best_config_df.iloc[0]
+
     tc = task_configs.get(chosen_exp)
+    # TODO: temporary workaround — when backend="auto", AIC's
+    # merge_experiment_results_by_mode collapses e.g. "agg_vllm" into "agg",
+    # but task_configs retains the original keys. Reconstruct the key from
+    # the winning row's backend column. Proper fix: AIC should return the
+    # original task config key alongside the merged chosen experiment name.
+    if tc is None and "backend" in row.index:
+        tc = task_configs.get(f"{chosen_exp}_{row['backend']}")
     if tc is None:
         return None
 
-    row = best_config_df.iloc[0]
-
     original_total_gpus = tc.total_gpus
     if "total_gpus_needed" in row.index and row["total_gpus_needed"] > 0:
         tc.total_gpus = int(row["total_gpus_needed"])
@@ -87,6 +94,10 @@ def _generate_dgd_from_pick(
     return None
 
 
+# in naive mode, use vllm as the default backend
+_DEFAULT_NAIVE_BACKEND = "vllm"
+
+
 def _run_naive_fallback(
     dgdr: DynamoGraphDeploymentRequestSpec,
     model: str,
@@ -95,6 +106,12 @@ def _run_naive_fallback(
     backend: str,
 ) -> dict:
     """Handle the AIC-unsupported path via naive config generation."""
+    if backend == "auto":
+        backend = _DEFAULT_NAIVE_BACKEND
+        logger.info(
+            "Auto backend resolved to '%s' for naive fallback.",
+            backend,
+        )
     logger.info(
         "AIC does not support this combo — falling back to naive config generation."
     )

tests/profiler/configs/11_auto_rapid_no_planner_no_load.yaml

@@ -0,0 +1,10 @@
+# SPDX-FileCopyrightText: Copyright (c) 2025-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+
+# Case 11: Auto backend, rapid, without planner, no input load
+model: "Qwen/Qwen3-32B"
+image: "hongkuanz196/trtllm-runtime:hzhou-0224"
+hardware:
+  gpuSku: h200_sxm
+  totalGpus: 8
+  numGpusPerNode: 8

tests/profiler/test_profile_sla_dgdr.py

@@ -251,6 +251,25 @@ def test_thorough_auto_backend_rejected(self, tmp_path):
             asyncio.run(run_profile(dgdr, ops))
 
 
+class TestAutoBackend:
+    """Rapid strategy with auto backend resolution."""
+
+    @pytest.mark.pre_merge
+    @pytest.mark.gpu_0
+    def test_no_planner_no_load(self, tmp_path):
+        """Case 11: auto backend, rapid, no planner, no target load."""
+        dgdr = _load_dgdr(CONFIGS_DIR / "11_auto_rapid_no_planner_no_load.yaml")
+        assert dgdr.backend == BackendType.Auto
+        ops = _make_ops(tmp_path)
+        asyncio.run(run_profile(dgdr, ops))
+
+        output = tmp_path / "profiling_results" / "final_config.yaml"
+        assert output.exists()
+        config = yaml.safe_load(output.read_text())
+        assert config, "final_config.yaml should not be empty"
+        assert "spec" in config
+
+
 class TestThoroughEdgeCases:
     """Edge cases for thorough mode."""