<!-- 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. -->
hing-roberta-finetuned-combined-DS
This model is a fine-tuned version of l3cube-pune/hing-roberta on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.0005
- Accuracy: 0.6840
- Precision: 0.6568
- Recall: 0.6579
- F1: 0.6570
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: 3.927975767245621e-05
- train_batch_size: 8
- eval_batch_size: 16
- seed: 43
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
---|---|---|---|---|---|---|---|
0.8684 | 1.0 | 1423 | 0.8762 | 0.6643 | 0.6561 | 0.6209 | 0.6215 |
0.6545 | 2.0 | 2846 | 0.8043 | 0.6805 | 0.6497 | 0.6522 | 0.6502 |
0.4267 | 3.0 | 4269 | 1.1337 | 0.6966 | 0.6668 | 0.6699 | 0.6680 |
0.2762 | 4.0 | 5692 | 1.6520 | 0.6784 | 0.6558 | 0.6571 | 0.6553 |
0.1535 | 5.0 | 7115 | 2.0005 | 0.6840 | 0.6568 | 0.6579 | 0.6570 |
Framework versions
- Transformers 4.20.1
- Pytorch 1.10.1+cu111
- Datasets 2.3.2
- Tokenizers 0.12.1