Commit
•
eeacd1a
1
Parent(s):
d33d866
Update README.md
Browse files
README.md
CHANGED
|
@@ -18,10 +18,14 @@ This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base)
|
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
- Loss: 1.2337
|
| 20 |
|
| 21 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
|
|
|
|
| 23 |
### Training hyperparameters
|
| 24 |
-
|
| 25 |
The following hyperparameters were used during training:
|
| 26 |
- learning_rate: 0.0001
|
| 27 |
- train_batch_size: 4
|
|
@@ -66,11 +70,3 @@ The following hyperparameters were used during training:
|
|
| 66 |
| 1.4272 | 9.06 | 2700 | 1.2386 |
|
| 67 |
| 1.4104 | 9.4 | 2800 | 1.2323 |
|
| 68 |
| 1.4179 | 9.73 | 2900 | 1.2337 |
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
### Framework versions
|
| 72 |
-
|
| 73 |
-
- Transformers 4.21.0
|
| 74 |
-
- Pytorch 1.12.0+cu113
|
| 75 |
-
- Datasets 2.4.0
|
| 76 |
-
- Tokenizers 0.12.1
|
|
|
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
- Loss: 1.2337
|
| 20 |
|
| 21 |
+
### Framework versions
|
| 22 |
+
- Transformers 4.21.0
|
| 23 |
+
- Pytorch 1.12.0+cu113
|
| 24 |
+
- Datasets 2.4.0
|
| 25 |
+
- Tokenizers 0.12.1
|
| 26 |
|
| 27 |
+
## Training procedure
|
| 28 |
### Training hyperparameters
|
|
|
|
| 29 |
The following hyperparameters were used during training:
|
| 30 |
- learning_rate: 0.0001
|
| 31 |
- train_batch_size: 4
|
|
|
|
| 70 |
| 1.4272 | 9.06 | 2700 | 1.2386 |
|
| 71 |
| 1.4104 | 9.4 | 2800 | 1.2323 |
|
| 72 |
| 1.4179 | 9.73 | 2900 | 1.2337 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|