|
|
--- |
|
|
license: apache-2.0 |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
metrics: |
|
|
- accuracy |
|
|
model-index: |
|
|
- name: SST2_DistilBERT_5E |
|
|
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. --> |
|
|
|
|
|
# SST2_DistilBERT_5E |
|
|
|
|
|
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.4125 |
|
|
- Accuracy: 0.8933 |
|
|
|
|
|
## 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: 1e-05 |
|
|
- 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 | Validation Loss | Accuracy | |
|
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
|
| 0.6744 | 0.12 | 50 | 0.6094 | 0.66 | |
|
|
| 0.4942 | 0.23 | 100 | 0.3772 | 0.8667 | |
|
|
| 0.3857 | 0.35 | 150 | 0.3256 | 0.8867 | |
|
|
| 0.3483 | 0.46 | 200 | 0.3634 | 0.84 | |
|
|
| 0.3235 | 0.58 | 250 | 0.3338 | 0.8733 | |
|
|
| 0.3129 | 0.69 | 300 | 0.3482 | 0.8667 | |
|
|
| 0.3573 | 0.81 | 350 | 0.3632 | 0.8333 | |
|
|
| 0.3266 | 0.92 | 400 | 0.3274 | 0.86 | |
|
|
| 0.2615 | 1.04 | 450 | 0.3400 | 0.8667 | |
|
|
| 0.2409 | 1.15 | 500 | 0.3541 | 0.8467 | |
|
|
| 0.2508 | 1.27 | 550 | 0.2997 | 0.88 | |
|
|
| 0.2442 | 1.39 | 600 | 0.3654 | 0.86 | |
|
|
| 0.2625 | 1.5 | 650 | 0.3302 | 0.8667 | |
|
|
| 0.1983 | 1.62 | 700 | 0.3184 | 0.8867 | |
|
|
| 0.2356 | 1.73 | 750 | 0.3239 | 0.8867 | |
|
|
| 0.2078 | 1.85 | 800 | 0.2968 | 0.9 | |
|
|
| 0.2343 | 1.96 | 850 | 0.3148 | 0.8933 | |
|
|
| 0.1544 | 2.08 | 900 | 0.3535 | 0.9 | |
|
|
| 0.1407 | 2.19 | 950 | 0.3603 | 0.8733 | |
|
|
| 0.187 | 2.31 | 1000 | 0.3843 | 0.88 | |
|
|
| 0.144 | 2.42 | 1050 | 0.4546 | 0.8467 | |
|
|
| 0.1786 | 2.54 | 1100 | 0.3681 | 0.88 | |
|
|
| 0.1315 | 2.66 | 1150 | 0.3806 | 0.8867 | |
|
|
| 0.1399 | 2.77 | 1200 | 0.3880 | 0.8867 | |
|
|
| 0.1905 | 2.89 | 1250 | 0.3944 | 0.8733 | |
|
|
| 0.2043 | 3.0 | 1300 | 0.3974 | 0.8733 | |
|
|
| 0.1081 | 3.12 | 1350 | 0.3731 | 0.9067 | |
|
|
| 0.1055 | 3.23 | 1400 | 0.3809 | 0.8867 | |
|
|
| 0.1092 | 3.35 | 1450 | 0.3568 | 0.9 | |
|
|
| 0.0981 | 3.46 | 1500 | 0.3610 | 0.9133 | |
|
|
| 0.109 | 3.58 | 1550 | 0.4126 | 0.8867 | |
|
|
| 0.1001 | 3.7 | 1600 | 0.3831 | 0.9 | |
|
|
| 0.1027 | 3.81 | 1650 | 0.4064 | 0.9 | |
|
|
| 0.133 | 3.93 | 1700 | 0.3845 | 0.9 | |
|
|
| 0.1031 | 4.04 | 1750 | 0.3915 | 0.9 | |
|
|
| 0.0772 | 4.16 | 1800 | 0.3988 | 0.8867 | |
|
|
| 0.0785 | 4.27 | 1850 | 0.3962 | 0.9 | |
|
|
| 0.1059 | 4.39 | 1900 | 0.3969 | 0.9 | |
|
|
| 0.0668 | 4.5 | 1950 | 0.4095 | 0.8933 | |
|
|
| 0.0915 | 4.62 | 2000 | 0.4077 | 0.8933 | |
|
|
| 0.1413 | 4.73 | 2050 | 0.4004 | 0.9067 | |
|
|
| 0.0727 | 4.85 | 2100 | 0.4100 | 0.8933 | |
|
|
| 0.0724 | 4.97 | 2150 | 0.4125 | 0.8933 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.24.0 |
|
|
- Pytorch 1.12.1+cu113 |
|
|
- Datasets 2.7.0 |
|
|
- Tokenizers 0.13.2 |
|
|
|
