Update README.md

README.md

@@ -34,16 +34,16 @@ queries = [
     "girls sandals",
     "backpacks",
     "shoes", 
-    "mustard blouse"
+    "mustard sleeveless gown"
 ]
 documents =  [
-        "Nike Air Max, with air cushion",
-        "Adidas Ultraboost, the best boost you can get",
-        "Women's sandals wide width 9",
-        "Girl's surf backpack",
-        "Fresh watermelon, all you can eat",
-        "Floral yellow dress with frills and lace"
-    ]
+    '{"title": "Nike Air Max, with air cushion", "description": "The best shoes you can get", "brand": "Nike", "color": "black"}',
+    '{"title": "Adidas Ultraboost, the best boost you can get", "description": "The shoes that represent the world", "brand": "Adidas", "color": "white"}',
+    '{"title": "Womens sandals", "description": "Sandals:  wide width 9", "brand": "Chacos", "color": "blue"}',
+    '{"title": "Girls surf backpack", "description": "The best backpack in town", "brand": "Roxy", "color": "pink"}',
+    '{"title": "Fresh watermelon", "description": "The best fruit in town, all you can eat", "brand": "Fruitsellers Inc.", "color": "green"}',
+    '{"title": "Floral yellow dress with frills and lace", "description": "Brighten up your summers with a gorgeous dress", "brand": "Dressmakers Inc.", "color": "bright yellow"}'
+]
 
 model = AutoModelForSequenceClassification.from_pretrained(model_name)
 tokenizer = AutoTokenizer.from_pretrained(model_name)
@@ -74,21 +74,25 @@ queries = [
     "girls sandals",
     "backpacks",
     "shoes", 
-    "mustard blouse"
+    "mustard sleeveless gown"
 ]
 documents =  [
-        "Nike Air Max, with air cushion",
-        "Adidas Ultraboost, the best boost you can get",
-        "Women's sandals wide width 9",
-        "Girl's surf backpack",
-        "Fresh watermelon, all you can eat",
-        "Floral yellow dress with frills and lace"
-    ]
+    '{"title": "Nike Air Max, with air cushion", "description": "The best shoes you can get", "brand": "Nike", "color": "black"}',
+    '{"title": "Adidas Ultraboost, the best boost you can get", "description": "The shoes that represent the world", "brand": "Adidas", "color": "white"}',
+    '{"title": "Womens sandals", "description": "Sandals:  wide width 9", "brand": "Chacos", "color": "blue"}',
+    '{"title": "Girls surf backpack", "description": "The best backpack in town", "brand": "Roxy", "color": "pink"}',
+    '{"title": "Fresh watermelon", "description": "The best fruit in town, all you can eat", "brand": "Fruitsellers Inc.", "color": "green"}',
+    '{"title": "Floral yellow dress with frills and lace", "description": "Brighten up your summers with a gorgeous dress", "brand": "Dressmakers Inc.", "color": "bright yellow"}'
+]
 model = CrossEncoder(model_name, max_length=512)
 scores = model.predict([(q, d) for q, d in zip(queries, documents)])
 print(scores)
 ```
 
+```md
+[ 1.1324936   0.51267356  1.039221    1.5969192  -0.8867093   0.5035825 ]
+```
+
 ## Training
 
 Trained using `CrossEntropyLoss` using `<query, document>` pairs with `grade` as the label.