hack-leti-acc-low
Browse files
README.md
CHANGED
|
@@ -16,8 +16,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [ai-forever/ruRoberta-large](https://huggingface.co/ai-forever/ruRoberta-large) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss:
|
| 20 |
-
- Accuracy: 0.
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
@@ -42,42 +42,16 @@ The following hyperparameters were used during training:
|
|
| 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 | Accuracy |
|
| 50 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 51 |
-
| No log | 1.0 | 95 |
|
| 52 |
-
| No log | 2.0 | 190 |
|
| 53 |
-
| No log | 3.0 | 285 |
|
| 54 |
-
| No log | 4.0 | 380 | 2.
|
| 55 |
-
| No log | 5.0 | 475 | 2.3186 | 0.4550 |
|
| 56 |
-
| 0.5693 | 6.0 | 570 | 2.3792 | 0.4180 |
|
| 57 |
-
| 0.5693 | 7.0 | 665 | 2.5442 | 0.4603 |
|
| 58 |
-
| 0.5693 | 8.0 | 760 | 2.6696 | 0.4444 |
|
| 59 |
-
| 0.5693 | 9.0 | 855 | 2.3096 | 0.4762 |
|
| 60 |
-
| 0.5693 | 10.0 | 950 | 2.8165 | 0.4550 |
|
| 61 |
-
| 0.4429 | 11.0 | 1045 | 3.2277 | 0.4233 |
|
| 62 |
-
| 0.4429 | 12.0 | 1140 | 3.0367 | 0.4233 |
|
| 63 |
-
| 0.4429 | 13.0 | 1235 | 3.3685 | 0.4127 |
|
| 64 |
-
| 0.4429 | 14.0 | 1330 | 3.3807 | 0.4180 |
|
| 65 |
-
| 0.4429 | 15.0 | 1425 | 3.5592 | 0.4021 |
|
| 66 |
-
| 0.3361 | 16.0 | 1520 | 3.8085 | 0.3915 |
|
| 67 |
-
| 0.3361 | 17.0 | 1615 | 3.3284 | 0.4233 |
|
| 68 |
-
| 0.3361 | 18.0 | 1710 | 3.2054 | 0.4339 |
|
| 69 |
-
| 0.3361 | 19.0 | 1805 | 3.6119 | 0.4180 |
|
| 70 |
-
| 0.3361 | 20.0 | 1900 | 3.5837 | 0.4339 |
|
| 71 |
-
| 0.3361 | 21.0 | 1995 | 3.4894 | 0.4180 |
|
| 72 |
-
| 0.2764 | 22.0 | 2090 | 3.3315 | 0.4074 |
|
| 73 |
-
| 0.2764 | 23.0 | 2185 | 3.5583 | 0.4127 |
|
| 74 |
-
| 0.2764 | 24.0 | 2280 | 3.4930 | 0.4127 |
|
| 75 |
-
| 0.2764 | 25.0 | 2375 | 3.5141 | 0.4339 |
|
| 76 |
-
| 0.2764 | 26.0 | 2470 | 3.5180 | 0.4339 |
|
| 77 |
-
| 0.2182 | 27.0 | 2565 | 3.6063 | 0.4233 |
|
| 78 |
-
| 0.2182 | 28.0 | 2660 | 3.5765 | 0.4233 |
|
| 79 |
-
| 0.2182 | 29.0 | 2755 | 3.5506 | 0.4233 |
|
| 80 |
-
| 0.2182 | 30.0 | 2850 | 3.5528 | 0.4286 |
|
| 81 |
|
| 82 |
|
| 83 |
### Framework versions
|
|
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [ai-forever/ruRoberta-large](https://huggingface.co/ai-forever/ruRoberta-large) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 2.8738
|
| 20 |
+
- Accuracy: 0.2116
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 42 |
- seed: 42
|
| 43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
- lr_scheduler_type: linear
|
| 45 |
+
- num_epochs: 4
|
| 46 |
|
| 47 |
### Training results
|
| 48 |
|
| 49 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 50 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 51 |
+
| No log | 1.0 | 95 | 4.0504 | 0.0 |
|
| 52 |
+
| No log | 2.0 | 190 | 3.5087 | 0.0529 |
|
| 53 |
+
| No log | 3.0 | 285 | 3.1335 | 0.1958 |
|
| 54 |
+
| No log | 4.0 | 380 | 2.8738 | 0.2116 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
|
| 56 |
|
| 57 |
### Framework versions
|