|
--- |
|
language: |
|
- en |
|
license: mit |
|
base_model: microsoft/deberta-v2-xlarge |
|
tags: |
|
- nycu-112-2-datamining-hw2 |
|
- generated_from_trainer |
|
datasets: |
|
- DandinPower/review_onlytitleandtext |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: deberta-v2-xlarge-otat |
|
results: |
|
- task: |
|
name: Text Classification |
|
type: text-classification |
|
dataset: |
|
name: DandinPower/review_onlytitleandtext |
|
type: DandinPower/review_onlytitleandtext |
|
metrics: |
|
- name: Accuracy |
|
type: accuracy |
|
value: 0.20114285714285715 |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# deberta-v2-xlarge-otat |
|
|
|
This model is a fine-tuned version of [microsoft/deberta-v2-xlarge](https://huggingface.co/microsoft/deberta-v2-xlarge) on the DandinPower/review_onlytitleandtext dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.6316 |
|
- Accuracy: 0.2011 |
|
- Macro F1: 0.0670 |
|
|
|
## 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: 4.5e-05 |
|
- train_batch_size: 4 |
|
- eval_batch_size: 4 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 8 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 1500 |
|
- num_epochs: 5 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Macro F1 | |
|
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:--------:| |
|
| 1.1994 | 0.14 | 500 | 1.6893 | 0.4029 | 0.3240 | |
|
| 1.6344 | 0.29 | 1000 | 1.6403 | 0.2011 | 0.0670 | |
|
| 1.6413 | 0.43 | 1500 | 1.6270 | 0.2 | 0.0667 | |
|
| 1.6326 | 0.57 | 2000 | 1.6375 | 0.1971 | 0.0659 | |
|
| 1.6128 | 0.71 | 2500 | 1.6604 | 0.2011 | 0.0670 | |
|
| 1.6213 | 0.86 | 3000 | 1.6161 | 0.2 | 0.0667 | |
|
| 1.6199 | 1.0 | 3500 | 1.6132 | 0.2017 | 0.0671 | |
|
| 1.6177 | 1.14 | 4000 | 1.6142 | 0.2011 | 0.0670 | |
|
| 1.6183 | 1.29 | 4500 | 1.6213 | 0.2 | 0.0667 | |
|
| 1.6211 | 1.43 | 5000 | 1.6136 | 0.1971 | 0.0659 | |
|
| 1.6145 | 1.57 | 5500 | 1.6169 | 0.1971 | 0.0659 | |
|
| 1.6187 | 1.71 | 6000 | 1.6160 | 0.2011 | 0.0670 | |
|
| 1.6174 | 1.86 | 6500 | 1.6146 | 0.2 | 0.0667 | |
|
| 1.6164 | 2.0 | 7000 | 1.6181 | 0.2 | 0.0667 | |
|
| 1.6184 | 2.14 | 7500 | 1.6109 | 0.1971 | 0.0659 | |
|
| 1.6152 | 2.29 | 8000 | 1.6189 | 0.2 | 0.0667 | |
|
| 1.6175 | 2.43 | 8500 | 1.6146 | 0.1971 | 0.0659 | |
|
| 1.6134 | 2.57 | 9000 | 1.6160 | 0.1971 | 0.0659 | |
|
| 1.6144 | 2.71 | 9500 | 1.6167 | 0.2011 | 0.0670 | |
|
| 1.6141 | 2.86 | 10000 | 1.6106 | 0.2017 | 0.0671 | |
|
| 1.6128 | 3.0 | 10500 | 1.6139 | 0.1971 | 0.0659 | |
|
| 1.6179 | 3.14 | 11000 | 1.6112 | 0.2 | 0.0667 | |
|
| 1.6096 | 3.29 | 11500 | 1.6127 | 0.2 | 0.0667 | |
|
| 1.6132 | 3.43 | 12000 | 1.6135 | 0.2011 | 0.0670 | |
|
| 1.6053 | 3.57 | 12500 | 1.6186 | 0.2 | 0.0667 | |
|
| 1.6049 | 3.71 | 13000 | 1.6277 | 0.2011 | 0.0670 | |
|
| 1.6044 | 3.86 | 13500 | 1.6271 | 0.2011 | 0.0670 | |
|
| 1.6017 | 4.0 | 14000 | 1.6275 | 0.2011 | 0.0670 | |
|
| 1.608 | 4.14 | 14500 | 1.6192 | 0.2011 | 0.0670 | |
|
| 1.6075 | 4.29 | 15000 | 1.6259 | 0.2011 | 0.0670 | |
|
| 1.601 | 4.43 | 15500 | 1.6267 | 0.2011 | 0.0670 | |
|
| 1.6086 | 4.57 | 16000 | 1.6339 | 0.2011 | 0.0670 | |
|
| 1.5955 | 4.71 | 16500 | 1.6340 | 0.2011 | 0.0670 | |
|
| 1.6013 | 4.86 | 17000 | 1.6322 | 0.2011 | 0.0670 | |
|
| 1.5976 | 5.0 | 17500 | 1.6316 | 0.2011 | 0.0670 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.39.3 |
|
- Pytorch 2.2.2+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
