Comments (2)
Good suggestion on adding this question to our FAQ!
Perhaps the important distinction to make here is that you don't need to use WhyLabs when using LangKit + whylogs.
LangKit defines UDFs which extract features from text (focused on language model scenarios). The LangKit UDFs are whylogs UDFs, and they execute as part of whylogs log() call, but whylogs can be used as an open source standalone tool, on premise.
whylogs executes UDFs to augment raw features and create additional extracted features useful for monitoring or data quality validation scenarios. The output of a whylogs log call contains a profile of this combined set of features to produce a compact statistical representation of your data. You can send these profiles to WhyLabs for storage and monitoring, but you don't have to use WhyLabs, check out these examples showing how to write/read profiles to disk or other locations: https://github.com/whylabs/whylogs/blob/mainline/python/examples/basic/Getting_Started.ipynb https://github.com/whylabs/whylogs/blob/mainline/python/examples/integrations/writers/Writing_Profiles.ipynb
The profiles themselves are good at giving you a compact statistical view of your data that can support a number of scenarios: profiles can compared and used to calculate drift, they can be validated against conditions, or used for visualization without retaining the raw data.
- https://github.com/whylabs/whylogs/blob/mainline/python/examples/advanced/Drift_Algorithm_Configuration.ipynb
- https://github.com/whylabs/whylogs/blob/mainline/python/examples/advanced/Metric_Constraints.ipynb
- https://github.com/whylabs/whylogs/blob/mainline/python/examples/basic/Notebook_Profile_Visualizer.ipynb
from langkit.
Thanks A LOT! :)
from langkit.
Related Issues (20)
- response_hallucination: remove llm requirement for consistency check
- Response Hallucination: decouple LLMs for sample generation and consistency checking
- text-davinci-003 OpenAI model deprecated
- Add feature to use bedrock agent to do the evaluation
- faiss-cpu - installation through pip not supported
- Original data HOT 1
- Jupyter kernel crashes on running injections module in Mac HOT 3
- Need example of prompt column name override
- Importing metrics issue since there is not a way to pass the model path if stored locally HOT 3
- Need env variable to avoid attempted download of nltk artifacts
- TypeError: unsupported operand type(s) for +=: 'int' and 'NoneType' HOT 2
- Better error message if OpenAI key is not provided in response_hallucination
- Explain refusal similarity HOT 1
- Specify Python version compatibility HOT 1
- consistency_check in response_hallucination HOT 1
- Use custom prompt in response_hallucination HOT 3
- add support for detoxify local models
- reduce latency in lazy initialization
- SWE-Agent: Implement the hallucinations metric from the main branch in the workflow branch
- Possible merge issue with 0.0.31 release HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from langkit.