|
|
--- |
|
|
library_name: transformers |
|
|
license: mit |
|
|
base_model: microsoft/deberta-v3-large |
|
|
tags: |
|
|
- multi-label |
|
|
- question-answering |
|
|
- text-classification |
|
|
- generated_from_trainer |
|
|
datasets: |
|
|
- saiteki-kai/BeaverTails-it |
|
|
metrics: |
|
|
- f1 |
|
|
- accuracy |
|
|
- precision |
|
|
- recall |
|
|
language: |
|
|
- it |
|
|
- en |
|
|
--- |
|
|
|
|
|
<!-- 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. --> |
|
|
|
|
|
# QA-DeBERTa-v3-large |
|
|
|
|
|
This model is a fine-tuned version of [microsoft/deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) on the [saiteki-kai/BeaverTails-it](https://huggingface.co/datasets/saiteki-kai/BeaverTails-it) dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.0808 |
|
|
- Accuracy: 0.6938 |
|
|
- Macro F1: 0.6484 |
|
|
- Macro Precision: 0.7149 |
|
|
- Macro Recall: 0.6176 |
|
|
- Micro F1: 0.7545 |
|
|
- Micro Precision: 0.7874 |
|
|
- Micro Recall: 0.7242 |
|
|
- Flagged/accuracy: 0.8566 |
|
|
- Flagged/precision: 0.8975 |
|
|
- Flagged/recall: 0.8380 |
|
|
- Flagged/f1: 0.8667 |
|
|
|
|
|
## 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: 3.85e-06 |
|
|
- train_batch_size: 16 |
|
|
- eval_batch_size: 64 |
|
|
- seed: 42 |
|
|
- distributed_type: multi-GPU |
|
|
- 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_ratio: 0.1 |
|
|
- num_epochs: 10 |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Macro F1 | Macro Precision | Macro Recall | Micro F1 | Micro Precision | Micro Recall | Flagged/accuracy | Flagged/precision | Flagged/recall | Flagged/f1 | |
|
|
|:-------------:|:-----:|:------:|:---------------:|:--------:|:--------:|:---------------:|:------------:|:--------:|:---------------:|:------------:|:----------------:|:-----------------:|:--------------:|:----------:| |
|
|
| 0.0985 | 1.0 | 33814 | 0.0877 | 0.6750 | 0.6102 | 0.6629 | 0.5948 | 0.7406 | 0.7705 | 0.7129 | 0.8447 | 0.8701 | 0.8475 | 0.8586 | |
|
|
| 0.0867 | 2.0 | 67628 | 0.0817 | 0.6910 | 0.6185 | 0.7559 | 0.5598 | 0.7446 | 0.8165 | 0.6842 | 0.8465 | 0.9093 | 0.8043 | 0.8536 | |
|
|
| 0.0561 | 3.0 | 101442 | 0.0808 | 0.6938 | 0.6484 | 0.7149 | 0.6177 | 0.7545 | 0.7875 | 0.7242 | 0.8566 | 0.8975 | 0.8380 | 0.8667 | |
|
|
| 0.0913 | 4.0 | 135256 | 0.0812 | 0.6877 | 0.6412 | 0.7136 | 0.6144 | 0.7516 | 0.7796 | 0.7255 | 0.8546 | 0.8902 | 0.8428 | 0.8658 | |
|
|
| 0.0709 | 5.0 | 169070 | 0.0826 | 0.6911 | 0.6376 | 0.7306 | 0.5982 | 0.7500 | 0.7911 | 0.7129 | 0.8538 | 0.8936 | 0.8370 | 0.8643 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.51.3 |
|
|
- Pytorch 2.7.0+cu118 |
|
|
- Datasets 3.5.1 |
|
|
- Tokenizers 0.21.1 |
