|
--- |
|
license: mit |
|
base_model: VietAI/vit5-base |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- f1 |
|
- accuracy |
|
model-index: |
|
- name: sentiment_oversampling_25_12 |
|
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. --> |
|
|
|
# sentiment_oversampling_25_12 |
|
|
|
This model is a fine-tuned version of [VietAI/vit5-base](https://huggingface.co/VietAI/vit5-base) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.1955 |
|
- F1: 0.6240 |
|
- Accuracy: 0.7911 |
|
|
|
## Model description |
|
|
|
Sentiment analysis for user's review on Meta application |
|
|
|
## 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: 80 |
|
- eval_batch_size: 80 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 3 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | F1 | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:--------:| |
|
| 1.2607 | 0.37 | 200 | 0.2220 | 0.5931 | 0.7885 | |
|
| 0.2748 | 0.75 | 400 | 0.2099 | 0.6012 | 0.7954 | |
|
| 0.2591 | 1.12 | 600 | 0.1922 | 0.6114 | 0.8117 | |
|
| 0.2519 | 1.5 | 800 | 0.1993 | 0.6203 | 0.7975 | |
|
| 0.2442 | 1.87 | 1000 | 0.2153 | 0.6092 | 0.7555 | |
|
| 0.2408 | 2.24 | 1200 | 0.1915 | 0.6182 | 0.7992 | |
|
| 0.2309 | 2.62 | 1400 | 0.2064 | 0.6109 | 0.7703 | |
|
| 0.2322 | 2.99 | 1600 | 0.1972 | 0.6169 | 0.7881 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.2 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.16.0 |
|
- Tokenizers 0.15.0 |
|
|
