This repository contains code and materials for the Hugging Face Neuron workshop using AWS SageMaker. The goal is to demonstrate how to fine-tune and perform inference with models like TinyLlama and Mistral using AWS infrastructure and Hugging Face libraries optimized for the AWS Neuron SDK.
- 01_finetuning_tinyLlama/: Code for fine-tuning the TinyLlama model on AWS SageMaker.
- 02_inference_mistral/: Code for running inference with the Mistral model using the Neuron SDK.
- additional_content/: Additional resources, scripts, or content related to the workshop.
- AWS Account: Ensure you have access to AWS SageMaker services.
- Neuron SDK: Installed to work with AWS Inferentia hardware.
- Hugging Face Transformers: Required for model fine-tuning and inference.
To install dependencies:
pip install -r requirements.txtNavigate to the 01_finetuning_tinyLlama directory and follow the notebook to fine-tune the model on SageMaker.
In the 02_inference_mistral directory, run the provided scripts to load the Mistral model and perform inference on SageMaker.
This project is licensed under the MIT License. See the LICENSE file for more details.