Commit
•
162291b
1
Parent(s):
bbf4aee
Model save
Browse files
README.md
CHANGED
|
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [Twitter/twhin-bert-large](https://huggingface.co/Twitter/twhin-bert-large) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss:
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
@@ -40,14 +40,22 @@ The following hyperparameters were used during training:
|
|
| 40 |
- seed: 42
|
| 41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
- lr_scheduler_type: linear
|
| 43 |
-
- num_epochs:
|
| 44 |
|
| 45 |
### Training results
|
| 46 |
|
| 47 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 48 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 49 |
-
| No log | 1.0 | 300 | 2.
|
| 50 |
-
| 2.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
|
| 52 |
|
| 53 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [Twitter/twhin-bert-large](https://huggingface.co/Twitter/twhin-bert-large) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 1.8996
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 40 |
- seed: 42
|
| 41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
- lr_scheduler_type: linear
|
| 43 |
+
- num_epochs: 10
|
| 44 |
|
| 45 |
### Training results
|
| 46 |
|
| 47 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 48 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 49 |
+
| No log | 1.0 | 300 | 2.1878 |
|
| 50 |
+
| 2.4077 | 2.0 | 600 | 2.0959 |
|
| 51 |
+
| 2.4077 | 3.0 | 900 | 2.1126 |
|
| 52 |
+
| 2.2053 | 4.0 | 1200 | 2.0066 |
|
| 53 |
+
| 2.0736 | 5.0 | 1500 | 1.9590 |
|
| 54 |
+
| 2.0736 | 6.0 | 1800 | 1.9668 |
|
| 55 |
+
| 2.0221 | 7.0 | 2100 | 1.9509 |
|
| 56 |
+
| 2.0221 | 8.0 | 2400 | 1.9274 |
|
| 57 |
+
| 1.9679 | 9.0 | 2700 | 1.8871 |
|
| 58 |
+
| 1.9687 | 10.0 | 3000 | 1.8996 |
|
| 59 |
|
| 60 |
|
| 61 |
### Framework versions
|