Commit
•
d0b87b5
1
Parent(s):
6c132db
Model save
Browse files
README.md
CHANGED
@@ -17,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
|
18 |
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the squad dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss:
|
21 |
|
22 |
## Model description
|
23 |
|
@@ -36,21 +36,28 @@ More information needed
|
|
36 |
### Training hyperparameters
|
37 |
|
38 |
The following hyperparameters were used during training:
|
39 |
-
- learning_rate:
|
40 |
- train_batch_size: 40
|
41 |
- eval_batch_size: 40
|
42 |
- seed: 42
|
43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
44 |
- lr_scheduler_type: linear
|
45 |
-
- num_epochs:
|
46 |
|
47 |
### Training results
|
48 |
|
49 |
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
-
| No log | 1.0 | 100 |
|
52 |
-
| No log | 2.0 | 200 |
|
53 |
-
| No log | 3.0 | 300 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
54 |
|
55 |
|
56 |
### Framework versions
|
|
|
17 |
|
18 |
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the squad dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 3.5143
|
21 |
|
22 |
## Model description
|
23 |
|
|
|
36 |
### Training hyperparameters
|
37 |
|
38 |
The following hyperparameters were used during training:
|
39 |
+
- learning_rate: 5e-05
|
40 |
- train_batch_size: 40
|
41 |
- eval_batch_size: 40
|
42 |
- seed: 42
|
43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
44 |
- lr_scheduler_type: linear
|
45 |
+
- num_epochs: 10
|
46 |
|
47 |
### Training results
|
48 |
|
49 |
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
+
| No log | 1.0 | 100 | 4.7435 |
|
52 |
+
| No log | 2.0 | 200 | 4.3343 |
|
53 |
+
| No log | 3.0 | 300 | 4.0804 |
|
54 |
+
| No log | 4.0 | 400 | 3.8983 |
|
55 |
+
| 4.3932 | 5.0 | 500 | 3.7642 |
|
56 |
+
| 4.3932 | 6.0 | 600 | 3.6649 |
|
57 |
+
| 4.3932 | 7.0 | 700 | 3.5978 |
|
58 |
+
| 4.3932 | 8.0 | 800 | 3.5499 |
|
59 |
+
| 4.3932 | 9.0 | 900 | 3.5216 |
|
60 |
+
| 3.7318 | 10.0 | 1000 | 3.5143 |
|
61 |
|
62 |
|
63 |
### Framework versions
|