https://github.com/build-on-aws/compilation-optimization-of-sagemaker

How to Speed Up Open Source LLMs Training and Cut Down Billing Time with Amazon SageMaker
https://github.com/build-on-aws/compilation-optimization-of-sagemaker

aws huggingface llm sagemaker

Last synced: about 1 month ago
JSON representation

How to Speed Up Open Source LLMs Training and Cut Down Billing Time with Amazon SageMaker

• Host: GitHub
• URL: https://github.com/build-on-aws/compilation-optimization-of-sagemaker
• Owner: build-on-aws
• License: mit-0
• Created: 2023-07-28T20:15:16.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2023-09-01T13:57:08.000Z (almost 2 years ago)
• Last Synced: 2025-02-18T01:39:28.310Z (4 months ago)
• Topics: aws, huggingface, llm, sagemaker
• Language: Python
• Homepage: https://community.aws/posts/llm-compiler-optimization-gpt2
• Size: 19.5 KB
• Stars: 2
• Watchers: 5
• Forks: 0
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE
  • Code of conduct: CODE_OF_CONDUCT.md

Awesome Lists containing this project

README

        
## Open Source LLMs Optimization with Amazon SageMaker

In this project, we'll explain how to use Amazon SageMaker to optimize the training and inference of open source large language models (LLMs).

1/ How to Speed Up Open Source LLMs Training and Cut Down Billing Time with Amazon SageMaker

Model: GPT-2

Corresponding subdirectory: /gpt2

2/ Easily Deploy Falcon-40B Open Source LLM with Amazon SageMaker JumpStart

Model: Falcon-40B

Corresponding subdirectory: TBC

3/ Fine-tune Falcon-40B Open Source LLM Using 4bit QLoRA and PEFT

Model: Falcon-40B

Corresponding subdirectory: TBC

## Security

See [CONTRIBUTING](CONTRIBUTING.md#security-issue-notifications) for more information.

## License

This library is licensed under the MIT-0 License. See the LICENSE file.