diff --git a/textract-pipeline/src/lambda/prompt-orchestrator/index.py b/textract-pipeline/src/lambda/prompt-orchestrator/index.py index 2f2698a..8995915 100644 --- a/textract-pipeline/src/lambda/prompt-orchestrator/index.py +++ b/textract-pipeline/src/lambda/prompt-orchestrator/index.py @@ -26,7 +26,7 @@ env = os.environ.get('ENVIRONMENT') # Get first character from env to get env name abbreviation env = env[0] # Get the bucket name based on the env that is linked with the storage integration -snowflake_ingestion_bucket = f' doczyai-use2-{env}-infra-s3-raw-data-ingestion' +snowflake_ingestion_bucket = f'doczyai-use2-{env}-infra-s3-raw-data-ingestion' # Configure logging logger = logging.getLogger() @@ -423,6 +423,10 @@ def main(bucket, object_key, field_groups): # df['Revised Prompt'] = [question] * len(contract_list_f) # Need to add batch_id to the df df['Batch_id'] = batch_id + + # Remove column confidence level and revised prompt from df + df = df.drop(columns=['Confidence Level']) + df = df.drop(columns=['Revised Prompt']) # TODO: uncomment accuracy function after testing # df, history = compare_with_actuals(df)