Add files via upload

quality.py

@@ -0,0 +1,61 @@
+import pandas as pd
+import numpy as np
+from google.cloud import bigquery
+from google.oauth2 import service_account
+from evidently import ColumnMapping
+from evidently.report import Report
+from evidently.metrics import EmbeddingsDriftMetric
+from sentence_transformers import SentenceTransformer
+
+# Load credentials from the service account key file
+project = "legalqaapp"
+dataset_id = "legal_data"
+table_id = "qa_logs"
+key_path = 'C:/Users/a/Desktop/legalapp2/legalqaapp_key.json'  # Replace with the path to your key file
+credentials = service_account.Credentials.from_service_account_file(key_path,)
+
+# Define your BigQuery query
+bq_query = """
+SELECT timestamp, question, context, answer, confidence
+FROM legal_data.qa_logs
+"""
+
+# Load BigQuery data
+client = bigquery.Client(credentials=credentials, project=project)
+current_data = client.query(bq_query).to_dataframe()
+
+# Load Hugging Face dataset as reference data
+from datasets import load_dataset
+reference_dataset = load_dataset("umarbutler/open-australian-legal-qa")
+reference_data = reference_dataset['train'].to_pandas()
+
+# Select relevant columns
+reference_data = reference_data[['question', 'answer']]
+current_data = current_data[['question', 'answer']]
+
+# Use sentence transformer to convert the text into embeddings
+model_miniLM = SentenceTransformer('all-MiniLM-L6-v2')
+
+# Convert the text columns into embeddings
+reference_embeddings = model_miniLM.encode(reference_data['question'].tolist())
+current_embeddings = model_miniLM.encode(current_data['question'].tolist())
+
+# Convert embeddings into DataFrame
+reference_df = pd.DataFrame(reference_embeddings, columns=[f'col_{i}' for i in range(reference_embeddings.shape[1])])
+current_df = pd.DataFrame(current_embeddings, columns=[f'col_{i}' for i in range(current_embeddings.shape[1])])
+
+# Create a ColumnMapping to define embedding columns
+column_mapping = ColumnMapping(embeddings={'small_subset': reference_df.columns[:10]})
+
+# Initialize Evidently report for embeddings drift
+report = Report(metrics=[
+    EmbeddingsDriftMetric('small_subset')
+])
+
+# Run the report
+report.run(reference_data=reference_df[:500], current_data=current_df[:500], column_mapping=column_mapping)
+
+# Save the report as an HTML file
+report.save_html('C:/Users/a/Desktop/legalapp2/embeddings_drift_report.html')
+
+print("Embeddings drift report generated successfully.")