metadata
library_name: transformers
language:
- en
base_model: Hartunka/tiny_bert_rand_50_v2
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- spearmanr
model-index:
- name: tiny_bert_rand_50_v2_stsb
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: GLUE STSB
type: glue
args: stsb
metrics:
- name: Spearmanr
type: spearmanr
value: 0.25802571079319986
tiny_bert_rand_50_v2_stsb
This model is a fine-tuned version of Hartunka/tiny_bert_rand_50_v2 on the GLUE STSB dataset. It achieves the following results on the evaluation set:
- Loss: 2.2828
- Pearson: 0.2634
- Spearmanr: 0.2580
- Combined Score: 0.2607
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 | Pearson | Spearmanr | Combined Score |
|---|---|---|---|---|---|---|
| 3.577 | 1.0 | 23 | 2.3197 | 0.1206 | 0.1077 | 0.1142 |
| 2.0557 | 2.0 | 46 | 2.4031 | 0.1291 | 0.1249 | 0.1270 |
| 1.8854 | 3.0 | 69 | 2.3713 | 0.2039 | 0.1988 | 0.2013 |
| 1.7118 | 4.0 | 92 | 2.3258 | 0.2474 | 0.2463 | 0.2469 |
| 1.4486 | 5.0 | 115 | 2.2828 | 0.2634 | 0.2580 | 0.2607 |
| 1.2898 | 6.0 | 138 | 2.7080 | 0.2622 | 0.2744 | 0.2683 |
| 1.0578 | 7.0 | 161 | 2.6507 | 0.2815 | 0.2900 | 0.2857 |
| 0.8953 | 8.0 | 184 | 2.8633 | 0.2585 | 0.2633 | 0.2609 |
| 0.7584 | 9.0 | 207 | 3.1760 | 0.2421 | 0.2473 | 0.2447 |
| 0.6589 | 10.0 | 230 | 3.0019 | 0.2613 | 0.2697 | 0.2655 |
Framework versions
- Transformers 4.50.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.21.1