Open redwyd opened 1 month ago
Evaluation not works here too I tried to solve the problems. But evaluation is virtually impossible
It seems because the developer didn't update the evaluation.py file. In the commit history of generate.py, information is split by '\n===================\n in txt files, so when running evaluation.py, the splitter cannot split the str into list by '\n===================\n' and zero division error occurs
When I evaluated the generated results after running the third step, the following error occurred: Traceback (most recent call last): File "/RAG-privacy/evaluation_results.py", line 767, in
eval_results(settings, title_table, table_list)
File "/RAG-privacy/evaluation_results.py", line 603, in evalresults
sources, outputs, contexts = getdata(path, model, tem, p, seq, gen)
File "/RAG-privacy/evaluation_results.py", line 120, in get_data
k = len(sources) // len(outputs)
ZeroDivisionError: integer division or modulo by zero
The error occurs in this code. Is there something wrong?: outputs = output.split('\n===================\n')[:-1]