metadata
language:
- en
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- accuracy
model-index:
- name: hBERTv1_new_pretrain_48_KD_qnli
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: GLUE QNLI
type: glue
config: qnli
split: validation
args: qnli
metrics:
- name: Accuracy
type: accuracy
value: 0.6009518579535054
hBERTv1_new_pretrain_48_KD_qnli
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_48_KD on the GLUE QNLI dataset. It achieves the following results on the evaluation set:
- Loss: 0.6648
- Accuracy: 0.6010
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: 4e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 10
- distributed_type: multi-GPU
- 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 | Accuracy |
|---|---|---|---|---|
| 0.6818 | 1.0 | 819 | 0.6669 | 0.5966 |
| 0.6689 | 2.0 | 1638 | 0.6732 | 0.5858 |
| 0.6675 | 3.0 | 2457 | 0.6721 | 0.5810 |
| 0.663 | 4.0 | 3276 | 0.6793 | 0.5832 |
| 0.66 | 5.0 | 4095 | 0.6663 | 0.5999 |
| 0.6574 | 6.0 | 4914 | 0.6648 | 0.6010 |
| 0.6591 | 7.0 | 5733 | 0.6781 | 0.5731 |
| 0.659 | 8.0 | 6552 | 0.6685 | 0.5951 |
| 0.6697 | 9.0 | 7371 | 0.6793 | 0.5792 |
| 0.6755 | 10.0 | 8190 | 0.6829 | 0.5698 |
| 0.6794 | 11.0 | 9009 | 0.6780 | 0.5773 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.12.0
- Tokenizers 0.13.3