def test_extract_hs2():
md = open_file("examples/nlu_retrieval_intents.md")
result = _extract_h2s(md)
assert result == [
"## intent: chitchat/what_should_i_buy",
"## intent: how_should_i_spend_money",
"## intent: chitchat/tell_me_a_joke",
]
reverse=True)
if args.similarity:
similar: List[List] = [[t, s] for t, s in outputs_with_sim
if s >= float(args.similarity)]
less_similar: List[List] = [[t, s] for t, s in outputs_with_sim
if s < float(args.similarity)]
print("Similar: ")
print(tabulate(similar))
print("\nLess Similar: ")
print(tabulate(less_similar))
if args.csv:
# csv will always include similarity
df = pd.DataFrame(outputs_with_sim, columns=["text", "similarity"])
df.to_csv(args.csv)
if args.nlu:
expanded_md = expand_nlu(open_file(args.nlu),
params,
model=model,
verbose=args.verbose)
print(expanded_md)
# Stop the clock
stop_perf = perf_counter()
if args.verbose:
print("Elapsed time:", abs(stop_perf - start_perf))
def test_open_file():
result = open_file("examples/nlu_no_entities.md")
assert result is not None
def test_extract_nlu():
result = extract_nlu(open_file("examples/nlu_no_entities.md"))
assert result is not None
assert isinstance(result[0], tuple)
assert isinstance(result[0][0], str)
assert isinstance(result[0][1], list)
def test_extract_nlu_leading_line_breaks():
md = open_file("examples/nlu_leading_breaks.md")
result = extract_nlu(md)
assert any(result[0][1])