metadata
library_name: transformers
language:
- en
base_model: Hartunka/tiny_bert_rand_5_v2
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- accuracy
- f1
model-index:
- name: tiny_bert_rand_5_v2_qqp
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: GLUE QQP
type: glue
args: qqp
metrics:
- name: Accuracy
type: accuracy
value: 0.8146178580262181
- name: F1
type: f1
value: 0.7371558828686656
tiny_bert_rand_5_v2_qqp
This model is a fine-tuned version of Hartunka/tiny_bert_rand_5_v2 on the GLUE QQP dataset. It achieves the following results on the evaluation set:
- Loss: 0.4161
- Accuracy: 0.8146
- F1: 0.7372
- Combined Score: 0.7759
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: 256
- eval_batch_size: 256
- seed: 10
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Combined Score |
|---|---|---|---|---|---|---|
| 0.4948 | 1.0 | 1422 | 0.4481 | 0.7835 | 0.6860 | 0.7347 |
| 0.4052 | 2.0 | 2844 | 0.4211 | 0.8023 | 0.7129 | 0.7576 |
| 0.3464 | 3.0 | 4266 | 0.4161 | 0.8146 | 0.7372 | 0.7759 |
| 0.2996 | 4.0 | 5688 | 0.4293 | 0.8191 | 0.7423 | 0.7807 |
| 0.2623 | 5.0 | 7110 | 0.4363 | 0.8210 | 0.7572 | 0.7891 |
| 0.2317 | 6.0 | 8532 | 0.4542 | 0.8218 | 0.7602 | 0.7910 |
| 0.207 | 7.0 | 9954 | 0.4872 | 0.8216 | 0.7650 | 0.7933 |
| 0.184 | 8.0 | 11376 | 0.5373 | 0.8273 | 0.7644 | 0.7959 |
Framework versions
- Transformers 4.50.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.21.1