Fine-Tune Llama 3.1 with Your Data [No-Code] ??

Welcome to the?AI in 5?newsletter with Clarifai!

Every week we bring you new models, tools, and tips to build production-ready AI!

Here is the summary of what we will be covering this week: ??

• Tutorial: Fine-Tune Llama 3.1 with Your Data.
• Notebook: Building RAG Apps with Clarifai and Unstructured IO.
• Community: Popular Text and Image Embedding Models
• How-To: Image Search Using the Python SDK
• Tip of the Week: Bulk Label Your Image Data

Fine-Tuning Llama 3.1 Made Easy ???

You can now easily fine-tune the Llama 3.1 8B instruct model using the new training template in the Platform UI.

Whether you need extended context, enhanced instruction-following, or advanced capabilities for tasks like text generation and classification, it’s all achievable.

Everything in this video is no-code using the Platform. All you need to do is upload your data, select the pre-built Llama 3.1 fine-tuning template, customize parameters such as the number of epochs, quantization, and other settings, and train your model.

Check out the tutorial: Fine-Tune Llama 3.1 with Your Data

Building RAG Applications with Clarifai and Unstructured IO ?

Unstructured IO provides tools to ingest and process unstructured documents.?

With the Clarifai and Unstructured IO integration, you can now retrieve data from local machines, cloud storage, or databases, and easily build RAG applications.

The below notebook provides a step-by-step guide on how to use Clarifai as your destination connector with Unstructured IO. It shows how to import raw data from a GitHub repository, process it, and send it to your Clarifai app, which serves as your vector database.

Once the data is ingested, you can leverage LLMs from the Clarifai community and interact with your data using the Python SDK.??

RAG using Clarifai <> Unstructured IO

Access Popular Embedding Models from the Community ??

Mxbai-Embed-Large-v1: Mxbai-Embed-Large-v1 is a versatile sentence embedding model trained on a unique dataset, offering superior performance across a wide range of NLP tasks, including text classification, semantic textual similarity, information retrieval, and more.

Image-Embedder-Clip: This is a CLIP based image embedding model that can be used for generating embeddings from images and can be used for use cases such as filtering, indexing, ranking, and organizing images according to visual similarity and transfer learning tasks.

Text-Embedding-3-Large: Text-Embedding-3-Large model is a larger text-embedding model designed to represent concepts within content such as natural language or code. It generates embeddings with up to 3072 dimensions.

You can also access various embedding models from the Clarifai Community for different use cases. Check them out here.

Image Search Using the Python SDK???

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/or semantic similarity by calculating distances between vector embedding representations of the data.?

Below is an example of how to use vector search to find similar images from your Clarifai App. Check out the code here.

Tip of the week: ??

Bulk Label Your Image Data at Upload Time!

You can bulk label your image inputs as they are uploaded to the platform. To do so, go to your app's page and select the Inputs option on the left sidebar.

Next, upload your inputs. Once they are uploaded, you can easily add labels by selecting the inputs you want to label by clicking the checkmark. You can even select multiple inputs by holding down the "Shift" key.

Then, label them all in bulk by clicking the Label as button, which saves you a ton of time. Check out the complete guide 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!