File size: 4,654 Bytes
9346a11 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 |
---
base_model: ai-forever/ruRoberta-large
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: ruRoberta-large_pos
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# ruRoberta-large_pos
This model is a fine-tuned version of [ai-forever/ruRoberta-large](https://huggingface.co/ai-forever/ruRoberta-large) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5140
- Precision: 0.5566
- Recall: 0.5871
- F1: 0.5714
- Accuracy: 0.8981
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 100
### Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| No log | 1.0 | 50 | 0.6582 | 0.0 | 0.0 | 0.0 | 0.7628 |
| No log | 2.0 | 100 | 0.5705 | 0.0118 | 0.0173 | 0.0140 | 0.7783 |
| No log | 3.0 | 150 | 0.4784 | 0.0277 | 0.0501 | 0.0356 | 0.8028 |
| No log | 4.0 | 200 | 0.4043 | 0.0784 | 0.1329 | 0.0986 | 0.8323 |
| No log | 5.0 | 250 | 0.3553 | 0.1545 | 0.2697 | 0.1965 | 0.8523 |
| No log | 6.0 | 300 | 0.4051 | 0.2312 | 0.2601 | 0.2448 | 0.8692 |
| No log | 7.0 | 350 | 0.3351 | 0.3456 | 0.3796 | 0.3618 | 0.8901 |
| No log | 8.0 | 400 | 0.2774 | 0.3344 | 0.3911 | 0.3606 | 0.8974 |
| No log | 9.0 | 450 | 0.3010 | 0.3819 | 0.5048 | 0.4349 | 0.9022 |
| 0.3753 | 10.0 | 500 | 0.2892 | 0.4114 | 0.4875 | 0.4462 | 0.9051 |
| 0.3753 | 11.0 | 550 | 0.2773 | 0.3707 | 0.5222 | 0.4336 | 0.9076 |
| 0.3753 | 12.0 | 600 | 0.3447 | 0.4706 | 0.5549 | 0.5093 | 0.9076 |
| 0.3753 | 13.0 | 650 | 0.3312 | 0.4317 | 0.5356 | 0.4781 | 0.9073 |
| 0.3753 | 14.0 | 700 | 0.2870 | 0.4818 | 0.6378 | 0.5489 | 0.9132 |
| 0.3753 | 15.0 | 750 | 0.3944 | 0.4443 | 0.5992 | 0.5103 | 0.9024 |
| 0.3753 | 16.0 | 800 | 0.3599 | 0.4319 | 0.6416 | 0.5163 | 0.9018 |
| 0.3753 | 17.0 | 850 | 0.3568 | 0.4560 | 0.6397 | 0.5325 | 0.9042 |
| 0.3753 | 18.0 | 900 | 0.4296 | 0.4674 | 0.5241 | 0.4941 | 0.9106 |
| 0.3753 | 19.0 | 950 | 0.3939 | 0.4617 | 0.5453 | 0.5 | 0.9137 |
| 0.0842 | 20.0 | 1000 | 0.3882 | 0.5109 | 0.5434 | 0.5266 | 0.9066 |
| 0.0842 | 21.0 | 1050 | 0.3870 | 0.5311 | 0.6243 | 0.5740 | 0.9075 |
| 0.0842 | 22.0 | 1100 | 0.4163 | 0.4252 | 0.6628 | 0.5181 | 0.8925 |
| 0.0842 | 23.0 | 1150 | 0.4097 | 0.4577 | 0.5010 | 0.4784 | 0.9004 |
| 0.0842 | 24.0 | 1200 | 0.3709 | 0.5482 | 0.6031 | 0.5743 | 0.9161 |
| 0.0842 | 25.0 | 1250 | 0.3366 | 0.5088 | 0.6647 | 0.5764 | 0.9141 |
| 0.0842 | 26.0 | 1300 | 0.4558 | 0.6132 | 0.6108 | 0.6120 | 0.9171 |
| 0.0842 | 27.0 | 1350 | 0.4982 | 0.5720 | 0.5896 | 0.5806 | 0.9102 |
| 0.0842 | 28.0 | 1400 | 0.3998 | 0.5615 | 0.6513 | 0.6030 | 0.9178 |
| 0.0842 | 29.0 | 1450 | 0.5028 | 0.5620 | 0.6551 | 0.6050 | 0.9108 |
| 0.0476 | 30.0 | 1500 | 0.3672 | 0.5739 | 0.6435 | 0.6067 | 0.9117 |
| 0.0476 | 31.0 | 1550 | 0.4520 | 0.5330 | 0.6532 | 0.5870 | 0.9084 |
| 0.0476 | 32.0 | 1600 | 0.5027 | 0.5628 | 0.6127 | 0.5867 | 0.9101 |
| 0.0476 | 33.0 | 1650 | 0.4461 | 0.4581 | 0.6108 | 0.5235 | 0.9087 |
| 0.0476 | 34.0 | 1700 | 0.4407 | 0.4726 | 0.5992 | 0.5285 | 0.9070 |
| 0.0476 | 35.0 | 1750 | 0.4512 | 0.5211 | 0.5241 | 0.5226 | 0.9082 |
### Framework versions
- Transformers 4.38.2
- Pytorch 2.1.2
- Datasets 2.1.0
- Tokenizers 0.15.2
|