fix(profiling): normalize timing metrics before export

src/lerobot/scripts/lerobot_train.py

@@ -173,7 +173,7 @@ def update_policy(
             forward_s=forward_s,
             backward_s=backward_s,
             optimizer_s=optimizer_s,
-            total_update_s=train_metrics.update_s,
+            total_update_s=train_metrics.update_s.val,
         )
     return train_metrics, output_dict
 

src/lerobot/utils/profiling_utils.py

@@ -22,6 +22,7 @@ import io
 import json
 import pstats
 import statistics
+from numbers import Real
 from collections.abc import Callable
 from dataclasses import dataclass, field
 from pathlib import Path
@@ -244,6 +245,14 @@ def _summary(values: list[float]) -> dict[str, float] | dict[str, None]:
     }
 
 
+def _as_float(value: Any) -> float:
+    if isinstance(value, Real):
+        return float(value)
+    if hasattr(value, "val"):
+        return float(value.val)
+    raise TypeError(f"Expected a real-valued metric, got {type(value).__name__}")
+
+
 @dataclass
 class StepTimingCollector:
     forward_s: list[float] = field(default_factory=list)
@@ -261,13 +270,13 @@ class StepTimingCollector:
         optimizer_s: float,
         total_update_s: float,
     ) -> None:
-        self.forward_s.append(forward_s)
+        self.forward_s.append(_as_float(forward_s))
-        self.backward_s.append(backward_s)
+        self.backward_s.append(_as_float(backward_s))
-        self.optimizer_s.append(optimizer_s)
+        self.optimizer_s.append(_as_float(optimizer_s))
-        self.total_update_s.append(total_update_s)
+        self.total_update_s.append(_as_float(total_update_s))
 
     def record_dataloading(self, dataloading_s: float) -> None:
-        self.dataloading_s.append(dataloading_s)
+        self.dataloading_s.append(_as_float(dataloading_s))
 
     def record_memory(self, *, step: int, allocated_bytes: int, reserved_bytes: int) -> None:
         self.memory_timeline.append(

tests/scripts/test_model_profiling.py

@@ -220,3 +220,25 @@ def test_deterministic_forward_artifacts_preserve_policy_mode(tmp_path):
     assert payload["reference_batch_size"] == 2
     assert "operator_fingerprint" in payload
     assert payload["outputs"]["loss"]["numel"] == 1
+
+
+def test_step_timing_collector_accepts_metric_like_values(tmp_path):
+    from lerobot.utils.profiling_utils import StepTimingCollector
+
+    class _MetricLike:
+        def __init__(self, val):
+            self.val = val
+
+    collector = StepTimingCollector()
+    collector.record(
+        forward_s=0.1,
+        backward_s=0.2,
+        optimizer_s=0.3,
+        total_update_s=_MetricLike(0.6),
+    )
+    collector.record_dataloading(_MetricLike(0.05))
+    collector.write_json(tmp_path / "step_timing_summary.json")
+
+    payload = json.loads((tmp_path / "step_timing_summary.json").read_text())
+    assert payload["total_update_s"]["mean"] == 0.6
+    assert payload["dataloading_s"]["mean"] == 0.05