feste
panml
feste | panml | |
---|---|---|
2 | 25 | |
37 | 109 | |
- | 0.0% | |
4.3 | 7.3 | |
about 1 year ago | 11 months ago | |
Python | Python | |
Apache License 2.0 | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
feste
-
[P] I got fed up with LangChain, so I made a simple open-source alternative for building Python AI apps as easy and intuitive as possible.
Agree, I do think though that for real deployment you need to reduce latency and be able to observe things and debug. This is what is missing, that's why I started Feste that build a lazy graph of calls and then optimize to use the batching API (when available) and also dispatch them in multiple processes.
-
[P] Feste, an open-source framework to optimize and parallelize NLP tasks
Github: https://github.com/perone/feste
panml
-
Best locally-runnable LLM
This is something I’m looking into as well. Currently built an open source kit to allow this experimentation easily: https://github.com/Pan-ML/panml
- PanML – a high lvl LLM library
- Article on general take of prompt engineering
- PanML - open source python library for LLM tuning and experimentation
-
Prompt Engineering
In terms of prompt approach and design, I’ve written some material on this in the context of sharing some of the directions people can think about. It’s using the open source library PanML, which sits on top of open source and commercial LLMs. GitHub: https://github.com/Pan-ML/panml
-
Alternative to LangChain for Open LLMs?
There are alternatives to langchain when it comes to working with LLMs, some have bit less abstraction involved. One such library is PanML where it's coming at it from a different angle where you can feed in direct HuggingFace parameters: https://github.com/Pan-ML/panml
Regarding your requirements of loading in 4-bit, as I created the library, it's something at we are looking to support - but also keen to get inputs and feedback on it to collaborate. Generally speaking, we use the model_arg to pass in any HuggingFace based parameters to the AutoModel instantiation.
- [P] I got fed up with LangChain, so I made a simple open-source alternative for building Python AI apps as easy and intuitive as possible.
- Python library for open source LLM experimentation and fine tuning
-
Writing direct python functions for programmatic control of prompting LLMs
It is related to this open source library: https://github.com/Pan-ML/panml
- High level library for interfacing with open source and OpenAI LLMs
What are some alternatives?
simpleAI - An easy way to host your own AI API and expose alternative models, while being compatible with "open" AI clients.
bcapps
lmql - A language for constraint-guided and efficient LLM programming.
guidance - A guidance language for controlling large language models. [Moved to: https://github.com/guidance-ai/guidance]
agency - Agency: Robust LLM Agent Management with Go
open-llms - 📋 A list of open LLMs available for commercial use.
transformers - 🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.