update model card README.md
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 [](https://huggingface.co/) on the generator dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 5.
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
@@ -41,7 +41,7 @@ The following hyperparameters were used during training:
|
|
| 41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
- lr_scheduler_type: cosine
|
| 43 |
- lr_scheduler_warmup_steps: 1000
|
| 44 |
-
- num_epochs:
|
| 45 |
- mixed_precision_training: Native AMP
|
| 46 |
|
| 47 |
### Training results
|
|
@@ -69,6 +69,11 @@ The following hyperparameters were used during training:
|
|
| 69 |
| 5.2684 | 35.92 | 19000 | 5.6687 |
|
| 70 |
| 5.2609 | 37.81 | 20000 | 5.6684 |
|
| 71 |
| 5.2606 | 39.7 | 21000 | 5.6719 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
|
| 73 |
|
| 74 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [](https://huggingface.co/) on the generator dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 5.6691
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
- lr_scheduler_type: cosine
|
| 43 |
- lr_scheduler_warmup_steps: 1000
|
| 44 |
+
- num_epochs: 50
|
| 45 |
- mixed_precision_training: Native AMP
|
| 46 |
|
| 47 |
### Training results
|
|
|
|
| 69 |
| 5.2684 | 35.92 | 19000 | 5.6687 |
|
| 70 |
| 5.2609 | 37.81 | 20000 | 5.6684 |
|
| 71 |
| 5.2606 | 39.7 | 21000 | 5.6719 |
|
| 72 |
+
| 5.2624 | 41.59 | 22000 | 5.6697 |
|
| 73 |
+
| 5.2551 | 43.48 | 23000 | 5.6718 |
|
| 74 |
+
| 5.2461 | 45.37 | 24000 | 5.6699 |
|
| 75 |
+
| 5.2431 | 47.26 | 25000 | 5.6692 |
|
| 76 |
+
| 5.2414 | 49.15 | 26000 | 5.6691 |
|
| 77 |
|
| 78 |
|
| 79 |
### Framework versions
|