Release v1.4.0

src/layerlens/_version.py

@@ -1,4 +1,4 @@
-__version__ = "1.3.3"
+__version__ = "1.4.0"
 
 # Will be templated during the build
 __git_commit__ = "__GIT_COMMIT__"

src/layerlens/resources/comparisons/comparisons.py

@@ -82,6 +82,7 @@ def compare_models(
             sort_by="submitted_at",
             order="desc",
             page_size=1,
+            unique=True,
             timeout=timeout,
         )
         eval_id_1 = _find_evaluation_id(resp1, model_id_1, benchmark_id)
@@ -93,6 +94,7 @@ def compare_models(
             sort_by="submitted_at",
             order="desc",
             page_size=1,
+            unique=True,
             timeout=timeout,
         )
         eval_id_2 = _find_evaluation_id(resp2, model_id_2, benchmark_id)
@@ -172,6 +174,7 @@ async def compare_models(
             sort_by="submitted_at",
             order="desc",
             page_size=1,
+            unique=True,
             timeout=timeout,
         )
         eval_id_1 = _find_evaluation_id(resp1, model_id_1, benchmark_id)
@@ -183,6 +186,7 @@ async def compare_models(
             sort_by="submitted_at",
             order="desc",
             page_size=1,
+            unique=True,
             timeout=timeout,
         )
         eval_id_2 = _find_evaluation_id(resp2, model_id_2, benchmark_id)

src/layerlens/resources/evaluations/evaluations.py

@@ -86,6 +86,7 @@ def get_many(
         model_ids: Optional[List[str]] = None,
         benchmark_ids: Optional[List[str]] = None,
         status: Optional[EvaluationStatus] = None,
+        unique: bool = False,
         timeout: float | httpx.Timeout | None = DEFAULT_TIMEOUT,
     ) -> Optional[EvaluationsResponse]:
         """
@@ -99,6 +100,7 @@ def get_many(
             model_ids: Filter by model IDs
             benchmark_ids: Filter by benchmark/dataset IDs
             status: Filter by evaluation status
+            unique: If True, deduplicate by model+dataset keeping only the latest evaluation per pair
             timeout: Request timeout
 
         Returns:
@@ -125,6 +127,8 @@ def get_many(
             params["datasets"] = ",".join(benchmark_ids)
         if status:
             params["status"] = status.value
+        if unique:
+            params["unique"] = "true"
 
         resp = self._get(
             f"/evaluations",
@@ -241,6 +245,7 @@ async def get_many(
         model_ids: Optional[List[str]] = None,
         benchmark_ids: Optional[List[str]] = None,
         status: Optional[EvaluationStatus] = None,
+        unique: bool = False,
         timeout: float | httpx.Timeout | None = DEFAULT_TIMEOUT,
     ) -> Optional[EvaluationsResponse]:
         """
@@ -254,6 +259,7 @@ async def get_many(
             model_ids: Filter by model IDs
             benchmark_ids: Filter by benchmark/dataset IDs
             status: Filter by evaluation status
+            unique: If True, deduplicate by model+dataset keeping only the latest evaluation per pair
             timeout: Request timeout
 
         Returns:
@@ -280,6 +286,8 @@ async def get_many(
             params["datasets"] = ",".join(benchmark_ids)
         if status:
             params["status"] = status.value
+        if unique:
+            params["unique"] = "true"
 
         resp = await self._get(
             f"/evaluations",

src/layerlens/resources/public_evaluations/public_evaluations.py

@@ -44,6 +44,7 @@ def get_many(
         model_ids: Optional[List[str]] = None,
         benchmark_ids: Optional[List[str]] = None,
         status: Optional[EvaluationStatus] = None,
+        unique: bool = False,
         timeout: float | httpx.Timeout | None = DEFAULT_TIMEOUT,
     ) -> Optional[EvaluationsResponse]:
         """
@@ -57,6 +58,7 @@ def get_many(
             model_ids: Filter by model IDs
             benchmark_ids: Filter by benchmark/dataset IDs
             status: Filter by evaluation status
+            unique: If True, deduplicate by model+dataset keeping only the latest evaluation per pair
             timeout: Request timeout
 
         Returns:
@@ -80,6 +82,8 @@ def get_many(
             params["datasets"] = ",".join(benchmark_ids)
         if status:
             params["status"] = status.value
+        if unique:
+            params["unique"] = "true"
 
         resp = self._get(
             "/evaluations",
@@ -137,6 +141,7 @@ async def get_many(
         model_ids: Optional[List[str]] = None,
         benchmark_ids: Optional[List[str]] = None,
         status: Optional[EvaluationStatus] = None,
+        unique: bool = False,
         timeout: float | httpx.Timeout | None = DEFAULT_TIMEOUT,
     ) -> Optional[EvaluationsResponse]:
         """
@@ -150,6 +155,7 @@ async def get_many(
             model_ids: Filter by model IDs
             benchmark_ids: Filter by benchmark/dataset IDs
             status: Filter by evaluation status
+            unique: If True, deduplicate by model+dataset keeping only the latest evaluation per pair
             timeout: Request timeout
 
         Returns:
@@ -173,6 +179,8 @@ async def get_many(
             params["datasets"] = ",".join(benchmark_ids)
         if status:
             params["status"] = status.value
+        if unique:
+            params["unique"] = "true"
 
         resp = await self._get(
             "/evaluations",

tests/resources/test_comparisons.py

@@ -88,6 +88,7 @@ def test_compare_models_success(self, comparisons, mock_public_client):
         assert calls[0].kwargs["sort_by"] == "submitted_at"
         assert calls[0].kwargs["order"] == "desc"
         assert calls[0].kwargs["page_size"] == 1
+        assert calls[0].kwargs["unique"] is True
 
         assert calls[1].kwargs["model_ids"] == ["model-b"]
 
@@ -199,3 +200,4 @@ def test_compare_models_picks_most_recent(self, comparisons, mock_public_client)
             assert call.kwargs["order"] == "desc"
             assert call.kwargs["page_size"] == 1
             assert call.kwargs["status"] == EvaluationStatus.SUCCESS
+            assert call.kwargs["unique"] is True