model update

README.md

@@ -91,7 +91,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.5934065934065934
+      value: 0.3196308724832215
   - task:
       name: Analogy Questions (TREX Analogy)
       type: multiple-choice-qa
@@ -102,7 +102,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.7872023809523809
+      value: 0.45901639344262296
   - task:
       name: Lexical Relation Classification (BLESS)
       type: classification
@@ -186,8 +186,8 @@ This model achieves the following results on the relation understanding tasks:
     - Accuracy on U2: 0.6710526315789473
     - Accuracy on U4: 0.6643518518518519
     - Accuracy on Google: 0.92
-    - Accuracy on ConceptNet Analogy: 0.5934065934065934
-    - Accuracy on T-Rex Analogy: 0.7872023809523809
+    - Accuracy on ConceptNet Analogy: 0.3196308724832215
+    - Accuracy on T-Rex Analogy: 0.45901639344262296
 - Lexical Relation Classification ([dataset](https://huggingface.co/datasets/relbert/lexical_relation_classification), [full result](https://huggingface.co/relbert/relbert-roberta-large-triplet-c-semeval2012/raw/main/classification.json)):
     - Micro F1 score on BLESS: 0.9025161970769926
     - Micro F1 score on CogALexV: 0.8798122065727699