QWEN_FACT_updates
This model is a fine-tuned version of
Qwen/Qwen2-7B on the None dataset.
It achieves the following results on the evaluation set:
-
Loss: 0.5144
-
Balanced Accuracy: 0.7801
-
Accuracy: 0.7998
-
Micro F1: 0.7998
-
Macro F1: 0.7392
-
Weighted F1: 0.8114
-
Classification Report: precision recall f1-score support
0 0.92 0.81 0.86 857
1 0.52 0.75 0.61 232
accuracy 0.80 1089
macro avg 0.72 0.78 0.74 1089
weighted avg 0.84 0.80 0.81 1089
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: 8
- seed: 42
- 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 | Accuracy | Balanced Accuracy | Classification Report | Validation Loss | Macro F1 | Micro F1 | Weighted F1 |
|---|
| 0.6846 | 1.0 | 391 | 0.7980 | 0.7553 | precision recall f1-score support | | | | |
0 0.91 0.83 0.87 857
1 0.52 0.68 0.59 232
accuracy 0.80 1089
macro avg 0.71 0.76 0.73 1089
weighted avg 0.82 0.80 0.81 1089
| 0.5173 | 0.7278 | 0.7980 | 0.8071 |
| 0.5021 | 2.0 | 782 | 0.8044 | 0.7673 | precision recall f1-score support
0 0.91 0.83 0.87 857
1 0.53 0.70 0.60 232
accuracy 0.80 1089
macro avg 0.72 0.77 0.74 1089
weighted avg 0.83 0.80 0.81 1089
| 0.4834 | 0.7374 | 0.8044 | 0.8135 |
| 0.408 | 3.0 | 1173 | 0.8356 | 0.7667 | precision recall f1-score support
0 0.90 0.89 0.89 857
1 0.61 0.65 0.63 232
accuracy 0.84 1089
macro avg 0.75 0.77 0.76 1089
weighted avg 0.84 0.84 0.84 1089
| 0.4296 | 0.7605 | 0.8356 | 0.8375 |
| 0.3032 | 4.0 | 1564 | 0.7511 | 0.7712 | precision recall f1-score support
0 0.93 0.74 0.82 857
1 0.45 0.81 0.58 232
accuracy 0.75 1089
macro avg 0.69 0.77 0.70 1089
weighted avg 0.83 0.75 0.77 1089
| 0.5927 | 0.7015 | 0.7511 | 0.7714 |
| 0.234 | 5.0 | 1955 | 0.5144 | 0.7801 | 0.7998 | 0.7998 | 0.7392 | 0.8114 | precision recall f1-score support
0 0.92 0.81 0.86 857
1 0.52 0.75 0.61 232
accuracy 0.80 1089
macro avg 0.72 0.78 0.74 1089
weighted avg 0.84 0.80 0.81 1089
|
Framework versions
- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
