|
--- |
|
base_model: cardiffnlp/twitter-xlm-roberta-base-sentiment |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
model-index: |
|
- name: xlm-roberta-meta4types-ft |
|
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. --> |
|
|
|
# xlm-roberta-meta4types-ft |
|
|
|
This model is a fine-tuned version of [cardiffnlp/twitter-xlm-roberta-base-sentiment](https://huggingface.co/cardiffnlp/twitter-xlm-roberta-base-sentiment) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.8324 |
|
- Roc Auc: 0.7122 |
|
- Hamming Loss: 0.2261 |
|
- F1 Score: 0.6089 |
|
- Accuracy: 0.5528 |
|
- Precision: 0.6081 |
|
- Recall: 0.6436 |
|
- Per Label: {'f1_score': 0.608905822183525, 'precision': 0.6080571799870046, 'recall': 0.6435841440010588, 'support': 235} |
|
|
|
## 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: 2e-05 |
|
- train_batch_size: 4 |
|
- eval_batch_size: 4 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 500 |
|
- num_epochs: 10 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Roc Auc | Hamming Loss | F1 Score | Accuracy | Precision | Recall | Per Label | |
|
|:-------------:|:-----:|:----:|:---------------:|:-------:|:------------:|:--------:|:--------:|:---------:|:------:|:-----------------------------------------------------------------------------------------------------------------:| |
|
| 0.4279 | 1.0 | 199 | 0.5287 | 0.4967 | 0.2496 | 0.3209 | 0.5276 | 0.6759 | 0.3575 | {'f1_score': 0.3208852937872149, 'precision': 0.6759286629224553, 'recall': 0.35748792270531404, 'support': 235} | |
|
| 0.4609 | 2.0 | 398 | 0.5076 | 0.5276 | 0.2245 | 0.3757 | 0.5779 | 0.8026 | 0.3913 | {'f1_score': 0.3757246741060956, 'precision': 0.8025944726452341, 'recall': 0.3913043478260869, 'support': 235} | |
|
| 0.5875 | 3.0 | 597 | 0.5463 | 0.5557 | 0.2127 | 0.4232 | 0.6080 | 0.6653 | 0.4153 | {'f1_score': 0.42320834457332973, 'precision': 0.6653348029760265, 'recall': 0.41534974521871487, 'support': 235} | |
|
| 0.493 | 4.0 | 796 | 0.5526 | 0.6428 | 0.2077 | 0.5744 | 0.6080 | 0.6577 | 0.5455 | {'f1_score': 0.5744086944086945, 'precision': 0.6577216876443267, 'recall': 0.5455495996294091, 'support': 235} | |
|
| 0.3519 | 5.0 | 995 | 0.6760 | 0.6795 | 0.2161 | 0.5809 | 0.5879 | 0.6192 | 0.5961 | {'f1_score': 0.5809003977320809, 'precision': 0.6191632544737641, 'recall': 0.5960790152868771, 'support': 235} | |
|
| 0.2451 | 6.0 | 1194 | 0.7729 | 0.7046 | 0.2312 | 0.6045 | 0.5578 | 0.6161 | 0.6045 | {'f1_score': 0.6045152483631816, 'precision': 0.6161038489469862, 'recall': 0.6044603269141685, 'support': 235} | |
|
| 0.0608 | 7.0 | 1393 | 0.7616 | 0.6942 | 0.2127 | 0.6060 | 0.5779 | 0.6221 | 0.6095 | {'f1_score': 0.6060266030810951, 'precision': 0.6220689655172414, 'recall': 0.6094566871815233, 'support': 235} | |
|
| 0.0859 | 8.0 | 1592 | 0.8324 | 0.7122 | 0.2261 | 0.6089 | 0.5528 | 0.6081 | 0.6436 | {'f1_score': 0.608905822183525, 'precision': 0.6080571799870046, 'recall': 0.6435841440010588, 'support': 235} | |
|
| 0.0767 | 9.0 | 1791 | 0.8192 | 0.6950 | 0.2127 | 0.6004 | 0.5578 | 0.6086 | 0.6073 | {'f1_score': 0.6003549503292779, 'precision': 0.6086247086247086, 'recall': 0.6072827741380452, 'support': 235} | |
|
| 0.0221 | 10.0 | 1990 | 0.8094 | 0.6975 | 0.2077 | 0.6135 | 0.5578 | 0.6116 | 0.6215 | {'f1_score': 0.6135398054397458, 'precision': 0.6116043923140263, 'recall': 0.6215108199324995, 'support': 235} | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.43.1 |
|
- Pytorch 1.13.1+cu116 |
|
- Datasets 2.20.0 |
|
- Tokenizers 0.19.1 |
|
|
