<!-- 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. -->
MedRuRobertaLarge
This model is a fine-tuned version of DmitryPogrebnoy/MedRuRobertaLarge on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4535
- Precision: 0.6121
- Recall: 0.6961
- F1: 0.6514
- Accuracy: 0.9283
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: 3
- 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 | 2.0 | 50 | 0.6334 | 0.0105 | 0.0147 | 0.0122 | 0.7491 |
| No log | 4.0 | 100 | 0.4261 | 0.1629 | 0.2451 | 0.1957 | 0.8234 |
| No log | 6.0 | 150 | 0.2987 | 0.3151 | 0.4510 | 0.3710 | 0.8850 |
| No log | 8.0 | 200 | 0.2382 | 0.4635 | 0.5294 | 0.4943 | 0.9166 |
| No log | 10.0 | 250 | 0.2304 | 0.5066 | 0.5686 | 0.5358 | 0.9232 |
| No log | 12.0 | 300 | 0.2574 | 0.5781 | 0.6716 | 0.6213 | 0.9192 |
| No log | 14.0 | 350 | 0.2437 | 0.5582 | 0.6814 | 0.6137 | 0.9298 |
| No log | 16.0 | 400 | 0.2504 | 0.6287 | 0.6225 | 0.6256 | 0.9361 |
| No log | 18.0 | 450 | 0.3189 | 0.5983 | 0.6716 | 0.6328 | 0.9252 |
| 0.2581 | 20.0 | 500 | 0.2555 | 0.5508 | 0.6912 | 0.6130 | 0.9300 |
| 0.2581 | 22.0 | 550 | 0.3072 | 0.5731 | 0.7108 | 0.6346 | 0.9384 |
| 0.2581 | 24.0 | 600 | 0.3128 | 0.6184 | 0.6912 | 0.6528 | 0.9450 |
| 0.2581 | 26.0 | 650 | 0.4012 | 0.5805 | 0.6716 | 0.6227 | 0.9272 |
| 0.2581 | 28.0 | 700 | 0.3723 | 0.5622 | 0.6863 | 0.6181 | 0.9295 |
| 0.2581 | 30.0 | 750 | 0.4157 | 0.592 | 0.7255 | 0.6520 | 0.9300 |
| 0.2581 | 32.0 | 800 | 0.2628 | 0.6278 | 0.6863 | 0.6557 | 0.9464 |
| 0.2581 | 34.0 | 850 | 0.3904 | 0.5660 | 0.6520 | 0.6059 | 0.9286 |
| 0.2581 | 36.0 | 900 | 0.2846 | 0.5739 | 0.6471 | 0.6083 | 0.9372 |
| 0.2581 | 38.0 | 950 | 0.3801 | 0.5992 | 0.7255 | 0.6563 | 0.9341 |
| 0.0241 | 40.0 | 1000 | 0.4299 | 0.5581 | 0.7304 | 0.6327 | 0.9272 |
| 0.0241 | 42.0 | 1050 | 0.3921 | 0.6272 | 0.7010 | 0.6620 | 0.9415 |
| 0.0241 | 44.0 | 1100 | 0.4305 | 0.6092 | 0.7108 | 0.6561 | 0.9418 |
| 0.0241 | 46.0 | 1150 | 0.3073 | 0.6376 | 0.7157 | 0.6744 | 0.9495 |
| 0.0241 | 48.0 | 1200 | 0.3380 | 0.6562 | 0.7206 | 0.6869 | 0.9427 |
| 0.0241 | 50.0 | 1250 | 0.4763 | 0.6151 | 0.7206 | 0.6637 | 0.9214 |
| 0.0241 | 52.0 | 1300 | 0.3092 | 0.6244 | 0.6765 | 0.6494 | 0.9409 |
| 0.0241 | 54.0 | 1350 | 0.3842 | 0.5521 | 0.7010 | 0.6177 | 0.9255 |
| 0.0241 | 56.0 | 1400 | 0.2719 | 0.5146 | 0.6912 | 0.5900 | 0.9280 |
| 0.0241 | 58.0 | 1450 | 0.2923 | 0.6824 | 0.7794 | 0.7277 | 0.9498 |
| 0.0227 | 60.0 | 1500 | 0.3172 | 0.6565 | 0.7402 | 0.6959 | 0.9475 |
| 0.0227 | 62.0 | 1550 | 0.4124 | 0.5845 | 0.6275 | 0.6052 | 0.9309 |
| 0.0227 | 64.0 | 1600 | 0.3563 | 0.7081 | 0.7255 | 0.7167 | 0.9438 |
| 0.0227 | 66.0 | 1650 | 0.3153 | 0.6016 | 0.7402 | 0.6637 | 0.9409 |
| 0.0227 | 68.0 | 1700 | 0.3588 | 0.6808 | 0.7108 | 0.6954 | 0.9427 |
| 0.0227 | 70.0 | 1750 | 0.4535 | 0.6121 | 0.6961 | 0.6514 | 0.9283 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1
