blade
/
dc-ml-emea-ar
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

optimize prompts for TOR

This commit is contained in:
Blade He 2024-12-06 14:50:34 -06:00
parent a25991e2bb
commit d79b05885d
4 changed files with 12 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
configuration/datapoint_reported_name.json
View File

@ -176,7 +176,7 @@
"Rotatie van de portefeuille", "Rotatie van de portefeuille",
"POF", "POF",
"Portefeuille omloop factor", "Portefeuille omloop factor",
"Taux de rotation corrigé - Gecorrigeerde omloopsnelheid" "Taux de rotation - Omloopsnelheid"
], ],
"french": [ "french": [
"taux de rotation", "taux de rotation",

3
instructions/data_extraction_prompts_config.json
View File

@ -70,7 +70,8 @@
"special_rule": { "special_rule": {
"tor": [ "tor": [
"If there are multiple TOR reported names, here is the priority rules:", "If there are multiple TOR reported names, here is the priority rules:",
"- With \"Taux de rotation - Omloopsnelheid\" and \"Taux de rotation corrigé - Gecorrigeerde omloopsnelheid\", pick up the values from \"Taux de rotation corrigé - Gecorrigeerde omloopsnelheid\"." "- With \"Taux de rotation - Omloopsnelheid\" and \"Taux de rotation corrigé - Gecorrigeerde omloopsnelheid\", pick up the values from \"Taux de rotation - Omloopsnelheid\".",
"- With \"Omloopsnelheid\" and \"Gecorrigeerde omloopsnelheid\", pick up the values from \"Omloopsnelheid\"."
], ],
"ter": [ "ter": [
"If there are multiple TER value columns, here is the priority rules:", "If there are multiple TER value columns, here is the priority rules:",

9
main.py
View File

@ -1122,7 +1122,6 @@ def batch_run_documents():
"407275419", "407275419",
"337937633", "337937633",
"337293427", "337293427",
"334584772",
"404712928", "404712928",
"451063582", "451063582",
"451878128", "451878128",
@ -1151,7 +1150,7 @@ def batch_run_documents():
"534535767" "534535767"
] ]
special_doc_id_list = check_db_mapping_doc_id_list special_doc_id_list = check_db_mapping_doc_id_list
special_doc_id_list = ["422760156"] # special_doc_id_list = ["407275419", "425595958", "451063582", "451878128"]
pdf_folder = r"/data/emea_ar/pdf/" pdf_folder = r"/data/emea_ar/pdf/"
page_filter_ground_truth_file = ( page_filter_ground_truth_file = (
r"/data/emea_ar/ground_truth/page_filter/datapoint_page_info_88_documents.xlsx" r"/data/emea_ar/ground_truth/page_filter/datapoint_page_info_88_documents.xlsx"
@ -1160,9 +1159,9 @@ def batch_run_documents():
output_extract_data_total_folder = r"/data/emea_ar/output/extract_data/total/" output_extract_data_total_folder = r"/data/emea_ar/output/extract_data/total/"
output_mapping_child_folder = r"/data/emea_ar/output/mapping_data/docs/" output_mapping_child_folder = r"/data/emea_ar/output/mapping_data/docs/"
output_mapping_total_folder = r"/data/emea_ar/output/mapping_data/total/" output_mapping_total_folder = r"/data/emea_ar/output/mapping_data/total/"
re_run_extract_data = True re_run_extract_data = False
re_run_mapping_data = True re_run_mapping_data = False
force_save_total_data = False force_save_total_data = True
calculate_metrics = False calculate_metrics = False
extract_ways = ["text"] extract_ways = ["text"]

8
specific_calc_metrics.py
View File

@ -62,9 +62,11 @@ def calculate_complex_document_metrics(verify_file_path: str, document_list: lis
document_count = len(document_list) \ document_count = len(document_list) \
if document_list is not None and len(document_list) > 0 \ if document_list is not None and len(document_list) > 0 \
else len(data_df["doc_id"].unique()) else len(data_df["doc_id"].unique())
verify_file_name = os.path.basename(verify_file_path).replace(".xlsx", "")
output_metrics_file = os.path.join(output_folder, output_metrics_file = os.path.join(output_folder,
f"complex_document_{document_count}_metrics.xlsx") f"complex_{verify_file_name}_metrics.xlsx")
with pd.ExcelWriter(output_metrics_file) as writer: with pd.ExcelWriter(output_metrics_file) as writer:
metrics_df.to_excel(writer, index=False, sheet_name="metrics") metrics_df.to_excel(writer, index=False, sheet_name="metrics")
@ -111,7 +113,7 @@ def get_sub_metrics(data_df: pd.DataFrame, data_point: str) -> dict:
if __name__ == "__main__": if __name__ == "__main__":
file_folder = r"/data/emea_ar/ground_truth/data_extraction/verify/complex/" file_folder = r"/data/emea_ar/ground_truth/data_extraction/verify/complex/"
verify_file = "mapping_data_info_31_documents_by_text_first_round.xlsx" verify_file = "mapping_data_info_31_documents_by_text_second_round.xlsx"
verify_file_path = os.path.join(file_folder, verify_file) verify_file_path = os.path.join(file_folder, verify_file)
document_list = [ document_list = [
"334584772", "334584772",