Update app.py

app.py

@@ -75,12 +75,15 @@ from datasets import load_dataset
 import huggingface_hub
 from huggingface_hub import Repository
 from datetime import datetime
+import pathlib as Path
 
 
-DATASET_REPO_URL = "https://huggingface.co/datasets/Seetha/visual_files/blob/main"
+DATASET_REPO_URL = "https://huggingface.co/datasets/Seetha/visual_files/raw/main"
 DATA_FILENAME = "level2.json"
 DATA_FILE = os.path.join(DATASET_REPO_URL, DATA_FILENAME)
-st.write(DATA_FILE)
+
+feedback_file = Path("https://huggingface.co/datasets/Seetha/visual_files/raw/main") / f"level2.json"
+st.write(feedback_file)
 
 HF_TOKEN = os.environ.get("HF_TOKEN")
 
@@ -167,18 +170,9 @@ def main():
       entity_list.append(i['entity_group'])
   
   filename = 'Checkpoint-classification.sav'
-  #filename = 'model.bin'
-  # count_vect = CountVectorizer(ngram_range=(1,3))
-  # tfidf_transformer=TfidfTransformer()
   loaded_model = pickle.load(open(filename, 'rb'))
-  #loaded_model = pickle.load(open(filename, 'rb'))
-  #loaded_model = joblib.load(filename)
-  #loaded_vectorizer = dill.load(open('vectorizefile_classification.pickle', 'rb'))
   loaded_vectorizer = pickle.load(open('vectorizefile_classification.pickle', 'rb'))
 
-  # from sklearn.pipeline import Pipeline
-  # pipeline1 = Pipeline([('count_vect',count_vect),('tfidf_transformer',tfidf_transformer)])
-  # pipeline_test_output = pipeline1.fit_transform(class_list)
   pipeline_test_output = loaded_vectorizer.transform(class_list)
   predicted = loaded_model.predict(pipeline_test_output)
   pred1 = predicted