Test fixes

langdspy/prompt_strategies.py

@@ -42,8 +42,6 @@ def __init__(self, **kwargs):
         hints = {}  # New dictionary to hold hint fields
 
         for name, attribute in self.__class__.__fields__.items():
-            print(f"Type of attribute.type_: {type(attribute.type_)}")
-            print(f"Class of attribute.type_: {attribute.type_.__class__}")
             if issubclass(attribute.type_, InputField):
                 inputs[name] = attribute.default
             elif issubclass(attribute.type_, OutputField):

tests/test_model_train.py

@@ -7,7 +7,51 @@
 dotenv.load_dotenv()
 import pytest
 from unittest.mock import MagicMock
-from examples.amazon.generate_slugs import ProductSlugGenerator, slug_similarity, get_llm
+import langdspy
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.metrics.pairwise import cosine_similarity
+
+
+class GenerateSlug(langdspy.PromptSignature):
+    hint_slug = langdspy.HintField(desc="Generate a URL-friendly slug based on the provided H1, title, and product copy. The slug should be lowercase, use hyphens to separate words, and not exceed 50 characters.")
+
+    h1 = langdspy.InputField(name="H1", desc="The H1 heading of the product page")
+    title = langdspy.InputField(name="Title", desc="The title of the product page")
+    product_copy = langdspy.InputField(name="Product Copy", desc="The product description or copy")
+
+    slug = langdspy.OutputField(name="Slug", desc="The generated URL-friendly slug")
+
+class ProductSlugGenerator(langdspy.Model):
+    generate_slug = langdspy.PromptRunner(template_class=GenerateSlug, prompt_strategy=langdspy.DefaultPromptStrategy)
+
+    def invoke(self, input_dict, config):
+        h1 = input_dict['h1']
+        title = input_dict['title']
+        product_copy = input_dict['product_copy']
+
+        slug_res = self.generate_slug.invoke({'h1': h1, 'title': title, 'product_copy': product_copy}, config=config)
+
+        return slug_res.slug
+
+
+def cosine_similarity_tfidf(true_slugs, predicted_slugs):
+    # Convert slugs to lowercase
+    true_slugs = [slug.lower() for slug in true_slugs]
+    predicted_slugs = [slug.lower() for slug in predicted_slugs]
+
+    # for i in range(len(true_slugs)):
+    #     print(f"Actual Slug: {true_slugs[i]} Predicted: {predicted_slugs[i]}")
+
+    vectorizer = TfidfVectorizer()
+    true_vectors = vectorizer.fit_transform(true_slugs)
+    predicted_vectors = vectorizer.transform(predicted_slugs)
+    similarity_scores = cosine_similarity(true_vectors, predicted_vectors)
+    return similarity_scores.diagonal()
+
+def slug_similarity(X, true_slugs, predicted_slugs):
+    similarity_scores = cosine_similarity_tfidf(true_slugs, predicted_slugs)
+    average_similarity = sum(similarity_scores) / len(similarity_scores)
+    return average_similarity
 
 @pytest.fixture
 def model():