Instructions to run unit test locally

[arnaudgelas]

Feature description

Instructions and simple setup to run all unit tests locally

Your Feature

[voidking]

You can refer to unittest.yaml and run_install_deps.sh.

[arnaudgelas]

Can you comment on how to set these variables for a local setup, please? https://github.com/geekan/MetaGPT/blob/main/.github/workflows/unittest.yaml#L35-L39

Note: having instructions in the ReadMe or a Contributing page would be useful

[voidking]

OK, I will add it to the docs soon.

1. Install unittest dependencies

bash run_install_deps.sh

2. Prepare config file key.yaml

OPENAI_API_KEY: "sk-xxx"
OPENAI_BASE_URL: "https://api.openai.com/v1"
OPENAI_API_MODEL: "gpt-3.5-turbo-16k"
MAX_TOKENS: 4096
RPM: 10

OPENAI_API_VERSION: "2023-09-01-preview"
DEPLOYMENT_NAME: "gpt-4-32k"

### browser path for pyppeteer engine, support Chrome, Chromium,MS Edge
# linux
PYPPETEER_EXECUTABLE_PATH: "/usr/bin/chromium"
# macos
# PYPPETEER_EXECUTABLE_PATH: "/Applications/Google Chrome.app/Contents/MacOS/Google Chrome"

SEARCH_ENGINE: ddg
SERPAPI_API_KEY: "xxx"
GOOGLE_API_KEY: "xxx"
GOOGLE_CSE_ID: "xxx"
SERPER_API_KEY: "xxx"

AZURE_TTS_SUBSCRIPTION_KEY: "xxx"
AZURE_TTS_REGION: "eastus"

IFLYTEK_APP_ID: "xxx"
IFLYTEK_API_KEY: "xxx"
IFLYTEK_API_SECRET: "xxx"

S3_ACCESS_KEY: "MOCK_S3_ACCESS_KEY"
S3_SECRET_KEY: "MOCK_S3_SECRET_KEY"
S3_ENDPOINT_URL: "http://mock:9000"
S3_SECURE: false # true/false
S3_BUCKET: "mock"

METAGPT_TEXT_TO_IMAGE_MODEL_URL: "http://mock.com"

3. Run the unittest

pytest tests/ --doctest-modules --cov=./metagpt/ --cov-report=xml:cov.xml --cov-report=html:htmlcov --durations=20
coverage report -m

[arnaudgelas]

@voidking Thanks! I have tried following the above instructions, but it hangs indefinitely with an opened window showing a graph.

[shroud777]

OK, I will add it to the docs soon.

1. Install unittest dependencies

bash run_install_deps.sh

2. Prepare config file key.yaml

OPENAI_API_KEY: "sk-xxx"
OPENAI_BASE_URL: "https://api.openai.com/v1"
OPENAI_API_MODEL: "gpt-3.5-turbo-16k"
MAX_TOKENS: 4096
RPM: 10

OPENAI_API_VERSION: "2023-09-01-preview"
DEPLOYMENT_NAME: "gpt-4-32k"

### browser path for pyppeteer engine, support Chrome, Chromium,MS Edge
# linux
PYPPETEER_EXECUTABLE_PATH: "/usr/bin/chromium"
# macos
# PYPPETEER_EXECUTABLE_PATH: "/Applications/Google Chrome.app/Contents/MacOS/Google Chrome"

SEARCH_ENGINE: ddg
SERPAPI_API_KEY: "xxx"
GOOGLE_API_KEY: "xxx"
GOOGLE_CSE_ID: "xxx"
SERPER_API_KEY: "xxx"

AZURE_TTS_SUBSCRIPTION_KEY: "xxx"
AZURE_TTS_REGION: "eastus"

IFLYTEK_APP_ID: "xxx"
IFLYTEK_API_KEY: "xxx"
IFLYTEK_API_SECRET: "xxx"

S3_ACCESS_KEY: "MOCK_S3_ACCESS_KEY"
S3_SECRET_KEY: "MOCK_S3_SECRET_KEY"
S3_ENDPOINT_URL: "http://mock:9000"
S3_SECURE: false # true/false
S3_BUCKET: "mock"

METAGPT_TEXT_TO_IMAGE_MODEL_URL: "http://mock.com"

3. Run the unittest

pytest tests/ --doctest-modules --cov=./metagpt/ --cov-report=xml:cov.xml --cov-report=html:htmlcov --durations=20
coverage report -m

Hello,did anyone successfully run the test? Where should I put this key.yaml? My running tests keep showing

` rsp_cache = {'\n\nExtract relevant information from every preceding sentence and use it to succinctly supplement or rewrite the fo...nship between sepal length and sepal width, color-coded by species.",\n "task_type": "eda"\n }\n]\n```', ...} mocker = <pytest_mock.plugin.MockerFixture object at 0x7fc8049b4410>, request = <SubRequest 'llm_mock' for <Function test_product_manager>>

@pytest.fixture(scope="function", autouse=True)
def llm_mock(rsp_cache, mocker, request):

  llm = MockLLM(allow_open_api_call=ALLOW_OPENAI_API_CALL)

tests/conftest.py:67:

---

tests/mock/mock_llm.py:20: in init super().init(original_llm_config) metagpt/provider/openai_api.py:60: in init self._init_client() metagpt/provider/azure_openai_api.py:24: in _init_client kwargs = self._make_client_kwargs()

---

self = <tests.mock.mock_llm.MockLLM object at 0x7fc8049b7990>

def _make_client_kwargs(self) -> dict:
    kwargs = dict(

      api_key=self.config.api_key,

        api_version=self.config.api_version,
        azure_endpoint=self.config.base_url,
    )

E AttributeError: 'NoneType' object has no attribute 'api_key'

metagpt/provider/azure_openai_api.py:32: AttributeError`