metadata
language:
- en
base_model: Hartunka/tiny_bert_rand_100_v1
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- spearmanr
model-index:
- name: tiny_bert_rand_100_v1_stsb
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: GLUE STSB
type: glue
args: stsb
metrics:
- name: Spearmanr
type: spearmanr
value: 0.2763743043991294
tiny_bert_rand_100_v1_stsb
This model is a fine-tuned version of Hartunka/tiny_bert_rand_100_v1 on the GLUE STSB dataset. It achieves the following results on the evaluation set:
- Loss: 2.2768
- Pearson: 0.2798
- Spearmanr: 0.2764
- Combined Score: 0.2781
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Pearson | Spearmanr | Combined Score |
|---|---|---|---|---|---|---|
| 3.4483 | 1.0 | 23 | 2.3146 | 0.1677 | 0.1464 | 0.1571 |
| 2.0255 | 2.0 | 46 | 2.5450 | 0.1168 | 0.1085 | 0.1126 |
| 1.8523 | 3.0 | 69 | 2.3148 | 0.2202 | 0.2082 | 0.2142 |
| 1.6156 | 4.0 | 92 | 2.3427 | 0.2703 | 0.2679 | 0.2691 |
| 1.3454 | 5.0 | 115 | 2.2768 | 0.2798 | 0.2764 | 0.2781 |
| 1.1616 | 6.0 | 138 | 2.6384 | 0.2686 | 0.2783 | 0.2734 |
| 0.9734 | 7.0 | 161 | 2.4772 | 0.2823 | 0.2840 | 0.2831 |
| 0.8406 | 8.0 | 184 | 2.8826 | 0.2435 | 0.2540 | 0.2487 |
| 0.7077 | 9.0 | 207 | 2.9091 | 0.2461 | 0.2524 | 0.2493 |
| 0.6149 | 10.0 | 230 | 2.8235 | 0.2652 | 0.2718 | 0.2685 |
Framework versions
- Transformers 4.40.0
- Pytorch 2.6.0+cu124
- Datasets 3.5.0
- Tokenizers 0.19.1