Christopher McMaster
commited on
Commit
•
6105b09
1
Parent(s):
d29a21c
update model card README.md
Browse files
README.md
CHANGED
|
@@ -13,7 +13,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
This model was trained from scratch on the None dataset.
|
| 15 |
It achieves the following results on the evaluation set:
|
| 16 |
-
- Loss: 0.
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
|
@@ -32,20 +32,22 @@ More information needed
|
|
| 32 |
### Training hyperparameters
|
| 33 |
|
| 34 |
The following hyperparameters were used during training:
|
| 35 |
-
- learning_rate:
|
| 36 |
- train_batch_size: 72
|
| 37 |
- eval_batch_size: 72
|
| 38 |
- seed: 42
|
| 39 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 40 |
- lr_scheduler_type: linear
|
| 41 |
-
- num_epochs:
|
| 42 |
|
| 43 |
### Training results
|
| 44 |
|
| 45 |
-
| Training Loss | Epoch | Step
|
| 46 |
-
|
| 47 |
-
| 0.
|
| 48 |
-
| 0.
|
|
|
|
|
|
|
| 49 |
|
| 50 |
|
| 51 |
### Framework versions
|
|
|
|
| 13 |
|
| 14 |
This model was trained from scratch on the None dataset.
|
| 15 |
It achieves the following results on the evaluation set:
|
| 16 |
+
- Loss: 0.7111
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
|
|
|
| 32 |
### Training hyperparameters
|
| 33 |
|
| 34 |
The following hyperparameters were used during training:
|
| 35 |
+
- learning_rate: 7e-05
|
| 36 |
- train_batch_size: 72
|
| 37 |
- eval_batch_size: 72
|
| 38 |
- seed: 42
|
| 39 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 40 |
- lr_scheduler_type: linear
|
| 41 |
+
- num_epochs: 4
|
| 42 |
|
| 43 |
### Training results
|
| 44 |
|
| 45 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 46 |
+
|:-------------:|:-----:|:------:|:---------------:|
|
| 47 |
+
| 0.9168 | 0.96 | 40000 | 0.8666 |
|
| 48 |
+
| 0.8392 | 1.91 | 80000 | 0.7871 |
|
| 49 |
+
| 0.7867 | 2.87 | 120000 | 0.7432 |
|
| 50 |
+
| 0.7418 | 3.83 | 160000 | 0.7111 |
|
| 51 |
|
| 52 |
|
| 53 |
### Framework versions
|