DSPy: A New Framework - Program Your Foundation Models, Not Just Prompting
Welcome to the?latest edition of the AI in 5 newsletter with Clarifai!
Every week, we bring you the models, tools and tips to build production-ready AI!
This week, we bring you: ??
DSPy ?? Clarifai ??
DSPy is a framework that offers the flexibility to algorithmically optimize language model prompts and weights.?
Instead of writing detailed prompts, you can simply define your task and the metrics you want to maximize, and prepare a few example inputs. DSPy will optimize the language model weights and instructions for you.
With the new Clarifai DSPy integration, you can now access LLMs from the Clarifai community and the vector database to build your applications.
The following video guides you through what DSPy is and how you can build a simple Retrieval Augmented Generation (RAG) system. ??
New models in the Clarifai platform ???
Gemini Pro Vision: Gemini Pro Vision is a Gemini large language vision model that understands input from text and visual modalities (image and video) to generate relevant text responses.?
You can now access the model with an API which opens up a lot of possible use cases such as object recognition, reasoning, digital content understanding and much more.
Qwen1.5-72B-chat: The Qwen1.5-72B-chat model is a transformer-based decoder-only language model pre-trained on a large amount of data.
This is the largest in the Qwen1.5 series, designed for language understanding, reasoning, and multilingual capabilities and also supports a context length of up to 32,768 tokens.?
领英推荐
DeepSeek-Coder-33B-Instruct: DeepSeek-Coder-33B-Instruct model is a SOTA 33 billion parameter code generation model, fine-tuned on 2 billion tokens of instruction data.
The model offers superior performance in code completion and infilling tasks across more than 80 programming languages.
Few Shot Learning in Production ??
Few-show learning is an approach focused on learning from only a few examples and is designed for situations where labeled data is scarce and hard to create.
With the advent of visual language models (VLMs), large models that connect text and language data, few-shot learning has become more tractable.?
We’ve paired up with the University of Toronto Engineering Science (Machine Intelligence) students to take a first step in productionizing a few-shot learning system. ??
AI tip of the week: ??
Smart Image Search in Python
Clarifai's Smart Search feature leverages vector search capabilities to power the search experience.?
Instead of traditional keyword-based search, where exact matches are sought, vector search allows for searching based on visual and semantic similarity by calculating distances between vector embedding representations of the data.
The following code shows how to use vector search to find similar images using the Python SDK.?Check out the code here.
Want to learn more from Clarifai? “Subscribe” to make sure you don’t miss the latest news, tutorials, educational materials, and tips. Thanks for reading!