|
--- |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- f1 |
|
- precision |
|
- recall |
|
model-index: |
|
- name: Roberta-base-Rewritten-commit_messages_v2 |
|
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. --> |
|
|
|
# Roberta-base-Rewritten-commit_messages_v2 |
|
|
|
This model was trained from scratch on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 3.4196 |
|
- Accuracy: 0.7704 |
|
- F1: 0.7707 |
|
- Precision: 0.7811 |
|
- Recall: 0.7704 |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 64 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 32 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 500 |
|
- num_epochs: 30 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:| |
|
| 0.2288 | 0.09 | 100 | 0.7615 | 0.6716 | 0.6578 | 0.7461 | 0.6716 | |
|
| 0.1103 | 0.17 | 200 | 0.6396 | 0.7453 | 0.7452 | 0.7609 | 0.7453 | |
|
| 0.2204 | 0.26 | 300 | 1.3317 | 0.6596 | 0.6334 | 0.7863 | 0.6596 | |
|
| 0.1872 | 0.34 | 400 | 0.8661 | 0.6333 | 0.5996 | 0.7712 | 0.6333 | |
|
| 0.1261 | 0.43 | 500 | 1.9369 | 0.7130 | 0.7023 | 0.7906 | 0.7130 | |
|
| 0.1902 | 0.52 | 600 | 2.2998 | 0.6823 | 0.6695 | 0.7573 | 0.6823 | |
|
| 0.3639 | 0.6 | 700 | 4.0162 | 0.6915 | 0.6815 | 0.7562 | 0.6915 | |
|
| 0.1655 | 0.69 | 800 | 2.2680 | 0.6859 | 0.6715 | 0.7705 | 0.6859 | |
|
| 0.1534 | 0.77 | 900 | 2.7909 | 0.7951 | 0.7937 | 0.7959 | 0.7951 | |
|
| 0.288 | 0.86 | 1000 | 2.9443 | 0.7752 | 0.7751 | 0.7920 | 0.7752 | |
|
| 0.2261 | 0.95 | 1100 | 2.9976 | 0.7318 | 0.7267 | 0.7810 | 0.7318 | |
|
| 0.162 | 1.03 | 1200 | 2.4699 | 0.8063 | 0.8067 | 0.8096 | 0.8063 | |
|
| 0.0379 | 1.12 | 1300 | 2.6939 | 0.8051 | 0.8051 | 0.8051 | 0.8051 | |
|
| 0.1852 | 1.2 | 1400 | 3.9005 | 0.7031 | 0.6940 | 0.7669 | 0.7031 | |
|
| 0.1258 | 1.29 | 1500 | 2.6666 | 0.8023 | 0.8027 | 0.8042 | 0.8023 | |
|
| 0.1707 | 1.38 | 1600 | 2.8308 | 0.7892 | 0.7892 | 0.7892 | 0.7892 | |
|
| 0.0817 | 1.46 | 1700 | 3.6049 | 0.7573 | 0.7497 | 0.7700 | 0.7573 | |
|
| 0.3516 | 1.55 | 1800 | 2.6816 | 0.7772 | 0.7777 | 0.7846 | 0.7772 | |
|
| 0.5502 | 1.63 | 1900 | 2.2493 | 0.8131 | 0.8099 | 0.8203 | 0.8131 | |
|
| 0.1531 | 1.72 | 2000 | 3.2802 | 0.7417 | 0.7407 | 0.7645 | 0.7417 | |
|
| 0.1112 | 1.81 | 2100 | 1.9678 | 0.7748 | 0.7737 | 0.8010 | 0.7748 | |
|
| 0.1617 | 1.89 | 2200 | 3.0694 | 0.7501 | 0.7490 | 0.7746 | 0.7501 | |
|
| 0.1912 | 1.98 | 2300 | 3.2285 | 0.7529 | 0.7530 | 0.7659 | 0.7529 | |
|
| 0.2725 | 2.06 | 2400 | 3.0008 | 0.7800 | 0.7805 | 0.7826 | 0.7800 | |
|
| 0.1694 | 2.15 | 2500 | 3.5542 | 0.7290 | 0.7286 | 0.7459 | 0.7290 | |
|
| 0.1283 | 2.24 | 2600 | 4.4577 | 0.7003 | 0.6944 | 0.7466 | 0.7003 | |
|
| 0.1321 | 2.32 | 2700 | 3.1128 | 0.7350 | 0.7356 | 0.7411 | 0.7350 | |
|
| 0.0 | 2.41 | 2800 | 4.2938 | 0.7222 | 0.7149 | 0.7828 | 0.7222 | |
|
| 0.0871 | 2.49 | 2900 | 3.9750 | 0.7266 | 0.7237 | 0.7607 | 0.7266 | |
|
| 0.0952 | 2.58 | 3000 | 3.7697 | 0.7437 | 0.7424 | 0.7690 | 0.7437 | |
|
| 0.1034 | 2.67 | 3100 | 3.7283 | 0.7350 | 0.7312 | 0.7764 | 0.7350 | |
|
| 0.2425 | 2.75 | 3200 | 3.4196 | 0.7704 | 0.7707 | 0.7811 | 0.7704 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.39.3 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.16.1 |
|
- Tokenizers 0.15.1 |
|
|
