Commit
•
3e6196c
1
Parent(s):
12ecf7f
Update README.md
Browse files
README.md
CHANGED
|
@@ -17,17 +17,6 @@ model-index:
|
|
| 17 |
|
| 18 |
SAFE_100M is a cutting-edge transformer-based model developed for molecular generation tasks. Trained from scratch on the [ZINC dataset](https://huggingface.co/datasets/sagawa/ZINC-canonicalized) converted to the SAFE (SMILES Augmented For Encoding) format, SAFE_100M achieves a loss of **0.3887** on its evaluation set, demonstrating robust performance in generating valid and diverse molecular structures.
|
| 19 |
|
| 20 |
-
## Table of Contents
|
| 21 |
-
|
| 22 |
-
- [Model Description](#model-description)
|
| 23 |
-
- [Intended Uses & Limitations](#intended-uses--limitations)
|
| 24 |
-
- [Training and Evaluation Data](#training-and-evaluation-data)
|
| 25 |
-
- [Training Procedure](#training-procedure)
|
| 26 |
-
- [Training Hyperparameters](#training-hyperparameters)
|
| 27 |
-
- [Framework Versions](#framework-versions)
|
| 28 |
-
- [Acknowledgements](#acknowledgements)
|
| 29 |
-
- [References](#references)
|
| 30 |
-
|
| 31 |
## Model Description
|
| 32 |
|
| 33 |
SAFE_100M leverages the [SAFE framework](#references) to enhance molecular representation through the SMILES Augmented For Encoding format. By utilizing the comprehensive [ZINC dataset](https://huggingface.co/datasets/sagawa/ZINC-canonicalized), the model excels in navigating chemical space, making it highly effective for applications such as:
|
|
|
|
| 17 |
|
| 18 |
SAFE_100M is a cutting-edge transformer-based model developed for molecular generation tasks. Trained from scratch on the [ZINC dataset](https://huggingface.co/datasets/sagawa/ZINC-canonicalized) converted to the SAFE (SMILES Augmented For Encoding) format, SAFE_100M achieves a loss of **0.3887** on its evaluation set, demonstrating robust performance in generating valid and diverse molecular structures.
|
| 19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
## Model Description
|
| 21 |
|
| 22 |
SAFE_100M leverages the [SAFE framework](#references) to enhance molecular representation through the SMILES Augmented For Encoding format. By utilizing the comprehensive [ZINC dataset](https://huggingface.co/datasets/sagawa/ZINC-canonicalized), the model excels in navigating chemical space, making it highly effective for applications such as:
|