Robeczech-CERED2
This model is a fine-tuned version of ufal/robeczech-base on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.9004
- Accuracy: 0.8819
- Micro Precision: 0.8819
- Micro Recall: 0.8819
- Micro F1: 0.8819
- Macro Precision: 0.8488
- Macro Recall: 0.8326
- Macro F1: 0.8379
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: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Micro Precision | Micro Recall | Micro F1 | Macro Precision | Macro Recall | Macro F1 |
---|---|---|---|---|---|---|---|---|---|---|
0.5518 | 1.0000 | 11305 | 0.5227 | 0.8496 | 0.8496 | 0.8496 | 0.8496 | 0.8293 | 0.7648 | 0.7779 |
0.4797 | 2.0 | 22611 | 0.4742 | 0.8623 | 0.8623 | 0.8623 | 0.8623 | 0.8191 | 0.8141 | 0.8052 |
0.369 | 3.0000 | 33916 | 0.4886 | 0.8684 | 0.8684 | 0.8684 | 0.8684 | 0.8493 | 0.8094 | 0.8198 |
0.308 | 4.0 | 45222 | 0.4829 | 0.8685 | 0.8685 | 0.8685 | 0.8685 | 0.8347 | 0.8231 | 0.8228 |
0.2395 | 5.0000 | 56527 | 0.4928 | 0.8755 | 0.8755 | 0.8755 | 0.8755 | 0.8300 | 0.8326 | 0.8265 |
0.1852 | 6.0 | 67833 | 0.5186 | 0.8799 | 0.8799 | 0.8799 | 0.8799 | 0.8528 | 0.8385 | 0.8401 |
0.1353 | 7.0000 | 79138 | 0.5951 | 0.8809 | 0.8809 | 0.8809 | 0.8809 | 0.8419 | 0.8419 | 0.8377 |
0.0945 | 8.0 | 90444 | 0.6848 | 0.8847 | 0.8847 | 0.8847 | 0.8847 | 0.8510 | 0.8478 | 0.8438 |
0.0551 | 9.0000 | 101749 | 0.7723 | 0.8867 | 0.8867 | 0.8867 | 0.8867 | 0.8469 | 0.8440 | 0.8405 |
0.0319 | 9.9996 | 113050 | 0.8430 | 0.8882 | 0.8882 | 0.8882 | 0.8882 | 0.8492 | 0.8487 | 0.8448 |
Framework versions
- Transformers 4.46.2
- Pytorch 2.5.1+cu124
- Datasets 3.1.0
- Tokenizers 0.20.3
- Downloads last month
- 4
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
HF Inference deployability: The model has no pipeline_tag.
Model tree for stulcrad/Robeczech-CERED2
Base model
ufal/robeczech-base