From bcb2c7b25ea8c3d93ad923831c5f71a01da38595 Mon Sep 17 00:00:00 2001
From: nan <zjn32202153@gmail.com>
Date: Wed, 2 Sep 2026 16:04:39 +0800
Subject: [PATCH] fix: include all evaluation score components

---
 VLABench/evaluation/utils.py   |  9 +++++--
 tests/test_evaluation_utils.py | 47 ++++++++++++++++++++++++++++++++++
 2 files changed, 54 insertions(+), 2 deletions(-)
 create mode 100644 tests/test_evaluation_utils.py

diff --git a/VLABench/evaluation/utils.py b/VLABench/evaluation/utils.py
index 0f5fef6..4715135 100644
--- a/VLABench/evaluation/utils.py
+++ b/VLABench/evaluation/utils.py
@@ -328,7 +328,12 @@ def get_final_score(standard_skill_sequence, model_skill_sequence, dependency):
         "skill_with_entity_match_score": 0.1,
         "exact_match_score": 0.1
     }
-    total_score = sum(score_weight[key] * value for key, value in skill_entity_scores.items())
+    all_scores = {
+        **skill_entity_scores,
+        **skill_with_entity_scores,
+        "exact_match_score": exact_match_score,
+    }
+    total_score = sum(score_weight[key] * all_scores[key] for key in score_weight)
     
     return {
         "skill_match_score": skill_entity_scores["skill_match_score"],
@@ -336,4 +341,4 @@ def get_final_score(standard_skill_sequence, model_skill_sequence, dependency):
         "skill_with_entity_match_score": skill_with_entity_scores["skill_with_entity_match_score"],
         "exact_match_score": exact_match_score,
         "total_score": total_score
-    }
\ No newline at end of file
+    }
diff --git a/tests/test_evaluation_utils.py b/tests/test_evaluation_utils.py
new file mode 100644
index 0000000..c7ecf83
--- /dev/null
+++ b/tests/test_evaluation_utils.py
@@ -0,0 +1,47 @@
+from importlib.util import module_from_spec, spec_from_file_location
+from pathlib import Path
+
+
+_UTILS_PATH = Path(__file__).parents[1] / "VLABench" / "evaluation" / "utils.py"
+_SPEC = spec_from_file_location("vlabench_evaluation_utils", _UTILS_PATH)
+utils = module_from_spec(_SPEC)
+assert _SPEC.loader is not None
+_SPEC.loader.exec_module(utils)
+
+
+def test_get_final_score_includes_all_weighted_metrics(monkeypatch):
+    monkeypatch.setattr(
+        utils,
+        "calculate_skill_and_entity_scores",
+        lambda standard, model: {"skill_match_score": 100.0, "entity_match_score": 0.0},
+    )
+    monkeypatch.setattr(
+        utils,
+        "calculate_skill_with_entity_scores",
+        lambda standard, model: {"skill_with_entity_match_score": 50.0},
+    )
+    monkeypatch.setattr(utils, "get_exact_match", lambda standard, model, dependency: 0.0)
+
+    result = utils.get_final_score([], [], "Sequential")
+
+    assert result["skill_match_score"] == 100.0
+    assert result["entity_match_score"] == 0.0
+    assert result["skill_with_entity_match_score"] == 50.0
+    assert result["exact_match_score"] == 0.0
+    assert result["total_score"] == 45.0
+
+
+def test_get_final_score_reaches_100_when_every_metric_is_perfect(monkeypatch):
+    monkeypatch.setattr(
+        utils,
+        "calculate_skill_and_entity_scores",
+        lambda standard, model: {"skill_match_score": 100.0, "entity_match_score": 100.0},
+    )
+    monkeypatch.setattr(
+        utils,
+        "calculate_skill_with_entity_scores",
+        lambda standard, model: {"skill_with_entity_match_score": 100.0},
+    )
+    monkeypatch.setattr(utils, "get_exact_match", lambda standard, model, dependency: 100.0)
+
+    assert utils.get_final_score([], [], "Sequential")["total_score"] == 100.0