Skip to content

Latest commit

 

History

History
65 lines (29 loc) · 3.22 KB

README.md

File metadata and controls

65 lines (29 loc) · 3.22 KB

Overview

We might agree that the operation of LLMs will embed in daily programming in the future. So, we use these notebooks to familiarize ourselves with the LLMs tools ecosystem and quantization techniques. I believe that Cloud quantum computing is needed for the future of LLMs. Maybe somthing Qubernetes.

All these notebooks have been completed running on the Kaggle platform. With the free GPUs. Some of the notebooks use a single GPU P100, some of notebooks use double GPU T4x2, others use CPUs.

Note: Some of the large size notebooks like Topic Modeling with BERTopic may not be able to show the complete version on the preview of Github. You can open it in the Kaggle platform by clicking the link in the notebook's title.

This project is interested in

I want to use of deep neural networks to do GenAI on consumer-grade hardware for researching.

the field of AI in layers

The ML tasks are covered in this project

This project's notebooks are covering the following some of the following tasks:

ML Tasks

The LLMs are used in this project

The some of LLMs are used in this project are as follows:

LLMs ecosystem LLMs

Metrics of fine-tuning

Note: All the fine-tuning here is under the limited computing resource, so the metrics are not the best. Most of reasons are num_train_epochs is not enough. However, the fine-tuning process is the same as the normal process.

And you can check the metrics of the fine-tuning in wandb.ai. It includes many of useful metrics, like: training, evaling, system power usage, like below:

The metrics of fine-tuning

LLMs tools are used in this project

The tools we covered in this project are as follows:

LLMs ecosystems-LLMs ecosystem

Quantization techniques are used in this project

The quantization techniques we used in this project are as follows:

quantization techniques

The Concepts From Papers We Should Know

the concepts from papers

License

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. See the LICENSE file for details.

Credits

Many of the notebooks are based on articles on Medium, TheNewStack, Huggingface and other open-source projects etc. Thanks for these great works.