Update README.md
Browse files
README.md
CHANGED
|
@@ -6,10 +6,13 @@ datasets:
|
|
| 6 |
---
|
| 7 |
# Indonesian T5 Small
|
| 8 |
|
|
|
|
| 9 |
T5 (Text-to-Text Transfer Transformer) model pretrained on Indonesian mC4 with [extra filtering](https://github.com/Wikidepia/indonesian_datasets/tree/master/dump/mc4). This model is pre-trained only and needs to be fine-tuned to be used for specific tasks.
|
| 10 |
|
| 11 |
## Pretraining Details
|
| 12 |
|
|
|
|
|
|
|
| 13 |
Trained for 1M steps following [`google/t5-v1_1-small`](https://huggingface.co/google/t5-v1_1-small).
|
| 14 |
|
| 15 |
## Model Performance
|
|
|
|
| 6 |
---
|
| 7 |
# Indonesian T5 Small
|
| 8 |
|
| 9 |
+
|
| 10 |
T5 (Text-to-Text Transfer Transformer) model pretrained on Indonesian mC4 with [extra filtering](https://github.com/Wikidepia/indonesian_datasets/tree/master/dump/mc4). This model is pre-trained only and needs to be fine-tuned to be used for specific tasks.
|
| 11 |
|
| 12 |
## Pretraining Details
|
| 13 |
|
| 14 |
+
**Apparently I do not use the same learning rate as the original T5-smallv1.1, I will retrain it soon :)**
|
| 15 |
+
|
| 16 |
Trained for 1M steps following [`google/t5-v1_1-small`](https://huggingface.co/google/t5-v1_1-small).
|
| 17 |
|
| 18 |
## Model Performance
|