Spaces:
Running
Running
| title: "Training PyThia-160M from scratch on AWS Sagemaker" | |
| emoji: 📈 | |
| colorFrom: indigo | |
| colorTo: yellow | |
| sdk: gradio | |
| sdk_version: 4.9.0 | |
| app_file: app.py | |
| pinned: false | |
| license: mit | |
| # Training PyThia-160M from scratch on AWS Sagemaker | |
| [**Repo Link**](https://github.com/RaviNaik/ERA-SESSION22/) | |
| This is an implementation of Pythia-160M [LitGPT](https://github.com/Lightning-AI/lit-gpt) by LightningAI. | |
| Dataset used to train: [RedPajama](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T). | |
| ### Tasks: | |
| 1. :heavy_check_mark: Once done, share: screenshot of the logs showing your submission result and clearly showing PageMaker in the URL [800 pts] | |
| 2. :heavy_check_mark: 4 sample predictions [200 pts] | |
| 3. ✔️ The model moved to HuggingFace Spaces as an app. [1000 pts] | |
| ### Training Log on AWS Sagemaker: | |
|  | |
| ### Sample Results: | |
|  | |
|  | |
|  | |
|  | |