data2 = parse_sgm_file(args.file2)
assert(len(data1) == len(data2))
assert(len(data1) % len(contexts) == 0)
// deduplication
data = []
for context, inst1, inst2 in zip(cycle(contexts), data1, data2):
After Change
ap = ArgumentParser()
ap.add_argument("-n", "--num-samples", type=int, default=1000, help="number of samples to create")
ap.add_argument("-t", "--test-questions", action="store_true",
help="create test questions (uses just the 1st input file + contexts)")
ap.add_argument("context_file", type=str, help="file with contexts")
ap.add_argument("file1", type=str, help="1st file with outputs for comparison")
// TODO make optional if test questions are generated
ap.add_argument("file2", type=str, help="2nd file with outputs for comparison")
