File size: 597 Bytes
1c6caef c4d4455 1c6caef c4d4455 1c6caef c4d4455 1c6caef c4d4455 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 |
---
library_name: transformers
tags:
- orpo
- llama
- llama 3
- fine tune
license: apache-2.0
datasets:
- mlabonne/orpo-dpo-mix-40k
---
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
Quick Llama 3 8B finetune with ORPO. Demontration that it can be fine tune in 2 hours only.
Thanks to Maxime Labonne's notebook:
https://colab.research.google.com/drive/1eHNWg9gnaXErdAa8_mcvjMupbSS6rDvi?usp=sharing
- **Number of training samples from the dataset**: 1500 out of 40K
- **Hardware Type:** L4
- **Hours of training:** 2
- **Cloud Provider:** google colab
|