fix: bug in image quality + prompt alignment metrics

wandb · Oct 30, 2024 · a2a4331 · a2a4331
1 parent 342ba02
commit a2a4331
Show file tree

Hide file tree

Showing 6 changed files with 6 additions and 6 deletions.
diff --git a/hemm/metrics/image_quality/lpips.py b/hemm/metrics/image_quality/lpips.py
@@ -85,7 +85,7 @@ def compute_metric(
         }
 
     @weave.op()
-    def evaluate(
+    def score(
         self, prompt: str, ground_truth_image: Image.Image, model_output: Dict[str, Any]
     ) -> Union[float, Dict[str, float]]:
         return super().evaluate(prompt, ground_truth_image, model_output)
diff --git a/hemm/metrics/image_quality/psnr.py b/hemm/metrics/image_quality/psnr.py
@@ -62,7 +62,7 @@ def compute_metric(
         }
 
     @weave.op()
-    def evaluate(
+    def score(
         self, prompt: str, ground_truth_image: Image.Image, model_output: Dict[str, Any]
     ) -> Union[float, Dict[str, float]]:
         _ = "PSNRMetric"

diff --git a/hemm/metrics/image_quality/ssim.py b/hemm/metrics/image_quality/ssim.py
@@ -106,7 +106,7 @@ def compute_metric(
         }
 
     @weave.op()
-    def evaluate(
+    def score(
         self, prompt: str, ground_truth_image: Image.Image, model_output: Dict[str, Any]
     ) -> Union[float, Dict[str, float]]:
         return super().evaluate(prompt, ground_truth_image, model_output)
diff --git a/hemm/metrics/prompt_alignment/blip_score.py b/hemm/metrics/prompt_alignment/blip_score.py
@@ -51,5 +51,5 @@ def compute_metric(
         )
 
     @weave.op()
-    def evaluate(self, prompt: str, model_output: Dict[str, Any]) -> Dict[str, float]:
+    def score(self, prompt: str, model_output: Dict[str, Any]) -> Dict[str, float]:
         return super().evaluate(prompt, model_output)
diff --git a/hemm/metrics/prompt_alignment/clip_iqa_score.py b/hemm/metrics/prompt_alignment/clip_iqa_score.py
@@ -79,5 +79,5 @@ def compute_metric(
         return score_dict
 
     @weave.op()
-    def evaluate(self, prompt: str, model_output: Dict[str, Any]) -> Dict[str, float]:
+    def score(self, prompt: str, model_output: Dict[str, Any]) -> Dict[str, float]:
         return super().evaluate(prompt, model_output)
diff --git a/hemm/metrics/prompt_alignment/clip_score.py b/hemm/metrics/prompt_alignment/clip_score.py
@@ -39,5 +39,5 @@ def compute_metric(
         )
 
     @weave.op()
-    def evaluate(self, prompt: str, model_output: Dict[str, Any]) -> Dict[str, float]:
+    def score(self, prompt: str, model_output: Dict[str, Any]) -> Dict[str, float]:
         return super().evaluate(prompt, model_output)