类库
› agent-evaluation
awslabs/agent-evaluation
Agent Evaluation 是一个生成式AI驱动的框架，用于测试虚拟AI助手。它通过内置的LLM评估器与待测目标agent进行多轮对话，并实时评估其响应质量，支持集成到CI/CD流水线中，内置了对亚马逊Bedrock、Q Business等服务的支持。
技术栈
根目录 python
查看全部依赖 (7)
依赖
Pydantic
boto3
click
jinja2
jsonpath-ng
pyyaml
rich
samples/aws_step_functions_deployment/layers/agent-evaluation python
查看全部依赖 (1)
依赖
agent-evaluation
samples/aws_step_functions_deployment/layers/aws-lambda-powertools python
查看全部依赖 (1)
依赖
aws-lambda-powertools
samples/aws_step_functions_deployment python
查看全部依赖 (3)
依赖
aws-cdk-lib
constructs
pathlib
samples/streamlit_app python
框架
Tornado
网络
Requests
查看全部依赖 (81)
依赖
Faker
GitPython
Jinja2
Markdown
MarkupSafe
NumPy
Pandas
PyYAML
Pydantic
Pygments
SQLAlchemy
agent-evaluation
altair
annotated-types
attrs
beautifulsoup4
blinker
boto3
botocore
cachetools
certifi
charset-normalizer
click
contourpy
cycler
entrypoints
favicon
fonttools
gitdb
htbuilder
idna
jmespath
jsonpath-ng
jsonschema
jsonschema-specifications
kiwisolver
lxml
markdown-it-py
markdownlit
matplotlib
mdurl
more-itertools
narwhals
packaging
pillow
plotly
ply
prometheus_client
protobuf
pyarrow
pydantic_core
pydeck
pymdown-extensions
pyparsing
python-dateutil
pytz
referencing
rich
rpds-py
s3transfer
six
smmap
soupsieve
st-annotated-text
st-theme
streamlit
streamlit-camera-input-live
streamlit-card
streamlit-embedcode
streamlit-extras
streamlit-faker
streamlit-image-coordinates
streamlit-keyup
streamlit-toggle-switch
streamlit-vertical-slider
tenacity
toml
typing_extensions
tzdata
urllib3
validators
