search

Giskard-AI / giskard

🐢 Open-Source Evaluation & Testing for ML & LLM systems
https://docs.giskard.ai
Apache License 2.0
4.05k stars 266 forks source link

LLMHarmfulContentDetector Failure/Crash #2027

Closed ClarkKentIsSuperman closed 1 month ago

ClarkKentIsSuperman commented 1 month ago

Issue Type

Bug

Source

source

Giskard Library Version

2.14.4

OS Platform and Distribution

Mac Command line, python venv

Python version

3.11.9

Installed python packages

aiohappyeyeballs==2.4.0
aiohttp==3.10.5
aiosignal==1.3.1
annotated-types==0.7.0
anyio==4.4.0
attrs==24.2.0
bert-score==0.3.13
bokeh==3.5.2
cachetools==5.5.0
certifi==2024.7.4
chardet==5.2.0
charset-normalizer==3.3.2
click==8.1.7
cloudpickle==3.0.0
colorama==0.4.6
contourpy==1.2.1
cycler==0.12.1
databricks-sdk==0.30.0
datasets==2.21.0
Deprecated==1.2.14
dill==0.3.8
distro==1.9.0
docopt==0.6.2
entrypoints==0.4
evaluate==0.4.2
faiss-cpu==1.8.0
filelock==3.15.4
fonttools==4.53.1
frozenlist==1.4.1
fsspec==2024.6.1
giskard==2.14.4
gitdb==4.0.11
GitPython==3.1.43
google-auth==2.34.0
griffe==0.48.0
h11==0.14.0
httpcore==1.0.5
httpx==0.27.0
huggingface-hub==0.24.6
idna==3.8
importlib_metadata==7.2.1
Jinja2==3.1.4
jiter==0.5.0
joblib==1.4.2
kiwisolver==1.4.5
langdetect==1.0.9
llvmlite==0.43.0
Markdown==3.7
MarkupSafe==2.1.5
matplotlib==3.9.2
mixpanel==4.10.1
mlflow-skinny==2.15.1
mpmath==1.3.0
multidict==6.0.5
multiprocess==0.70.16
networkx==3.3
num2words==0.5.13
numba==0.60.0
numpy==1.26.4
openai==1.42.0
opentelemetry-api==1.26.0
opentelemetry-sdk==1.26.0
opentelemetry-semantic-conventions==0.47b0
packaging==24.1
pandas==2.2.2
pillow==10.4.0
pip==24.2
protobuf==5.27.3
pyarrow==17.0.0
pyasn1==0.6.0
pyasn1_modules==0.4.0
pydantic==2.8.2
pydantic_core==2.20.1
pynndescent==0.5.13
pyparsing==3.1.2
python-dateutil==2.9.0.post0
pytz==2024.1
PyYAML==6.0.2
regex==2024.7.24
requests==2.32.3
requests-toolbelt==1.0.0
rsa==4.9
safetensors==0.4.4
scikit-learn==1.5.1
scipy==1.11.4
sentry-sdk==2.13.0
setuptools==72.1.0
six==1.16.0
smmap==5.0.1
sniffio==1.3.1
sqlparse==0.5.1
sympy==1.13.2
tenacity==9.0.0
threadpoolctl==3.5.0
tokenizers==0.19.1
torch==2.4.0
tornado==6.4.1
tqdm==4.66.5
transformers==4.44.2
typing_extensions==4.12.2
tzdata==2024.1
umap-learn==0.5.6
urllib3==2.2.2
wheel==0.43.0
wrapt==1.16.0
xxhash==3.5.0
xyzservices==2024.6.0
yarl==1.9.4
zipp==3.20.0
zstandard==0.23.0

Current Behaviour?

Testing the sample code against a Azure OpenAI gpt-3.5 instance - I get this error, after a handful of tests seem to successfully run:

2024-09-18 11:35:09,227 pid:4610 MainThread giskard.scanner.logger ERROR    Detector LLMHarmfulContentDetector failed with error: the JSON object must be str, bytes or bytearray, not NoneType
Traceback (most recent call last):
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/scanner/scanner.py", line 162, in _run_detectors
    detected_issues = detector.run(model, dataset, features=features)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/scanner/llm/base.py", line 77, in run
    eval_dataset = dg.generate_dataset(model, self.num_samples)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/llm/generators/base.py", line 72, in generate_dataset
    generated = self._parse_output(out)
                ^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/llm/generators/base.py", line 85, in _parse_output
    data = json.loads(raw_output.content)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/json/__init__.py", line 339, in loads
    raise TypeError(f'the JSON object must be str, bytes or bytearray, '
TypeError: the JSON object must be str, bytes or bytearray, not NoneType
Traceback (most recent call last):
  File "/Users/labuser/giskard_azure_openai.py", line 82, in <module>
    scan_results = giskard.scan(giskard_model, raise_exceptions=True)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/scanner/__init__.py", line 67, in scan
    return scanner.analyze(
           ^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/scanner/scanner.py", line 126, in analyze
    issues, errors = self._run_detectors(
                     ^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/scanner/scanner.py", line 176, in _run_detectors
    raise err
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/scanner/scanner.py", line 162, in _run_detectors
    detected_issues = detector.run(model, dataset, features=features)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/scanner/llm/base.py", line 77, in run
    eval_dataset = dg.generate_dataset(model, self.num_samples)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/llm/generators/base.py", line 72, in generate_dataset
    generated = self._parse_output(out)
                ^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/site-packages/giskard/llm/generators/base.py", line 85, in _parse_output
    data = json.loads(raw_output.content)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/giskard-env/lib/python3.11/json/__init__.py", line 339, in loads
    raise TypeError(f'the JSON object must be str, bytes or bytearray, '
TypeError: the JSON object must be str, bytes or bytearray, not NoneType

Standalone code OR list down the steps to reproduce the issue

I'm trying to do the stand-alone LLM wrapping based on this doc: https://docs.giskard.ai/en/stable/open_source/scan/scan_llm/index.html

and I wrote my own llm_api() since it appears thats not included in the sample/doc, which simply does a POST to the Azure OpenAI endpoint passing the "question" properly to the chat endpoint.

Relevant log output

I'm unsure where the log files are?
ClarkKentIsSuperman commented 1 month ago

I guess I can get past it by changing raise_exceptions=False good enough for now.