Downloads · 30 days
7
3% of all-time downloads
hBERTv2_new_pretrain_48_emb_com_qqp
gokuls/hBERTv2_new_pretrain_48_emb_com_qqp
hBERTv2_new_pretrain_48_emb_com_qqp is a text classification model from gokuls. Use it when you need a label for a piece of text. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
218
Public
Repo size
6 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin397 MB · 100%
At a glance
- Task
- Text Classification
- Library
- transformers
- Model type
- hybridbert
- Access
- Public
- Created
- Jun 14, 2023
- Updated
- Jun 15, 2023
- SHA
- cff90da2
Try a prompt
Datasets
Model card
From the Hugging Face model README
<!-- 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. -->
hBERTv2_new_pretrain_48_emb_com_qqp
This model is a fine-tuned version of gokuls/bert_12_layer_model_v2_complete_training_new_emb_compress_48 on the GLUE QQP dataset. It achieves the following results on the evaluation set:
- Loss: 0.4578
- Accuracy: 0.7937
- F1: 0.7323
- Combined Score: 0.7630
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 | F1 | Combined Score |
|---|---|---|---|---|---|---|
| 0.5521 | 1.0 | 2843 | 0.5375 | 0.6937 | 0.6680 | 0.6808 |
| 0.5006 | 2.0 | 5686 | 0.4939 | 0.7551 | 0.6743 | 0.7147 |
| 0.469 | 3.0 | 8529 | 0.4866 | 0.7714 | 0.6698 | 0.7206 |
| 0.4561 | 4.0 | 11372 | 0.5096 | 0.7633 | 0.6588 | 0.7110 |
| 0.44 | 5.0 | 14215 | 0.4706 | 0.7813 | 0.6987 | 0.7400 |
| 0.4168 | 6.0 | 17058 | 0.4598 | 0.7867 | 0.7061 | 0.7464 |
| 0.4 | 7.0 | 19901 | 0.4769 | 0.7797 | 0.7154 | 0.7476 |
| 0.3843 | 8.0 | 22744 | 0.4907 | 0.7829 | 0.7122 | 0.7476 |
| 0.3686 | 9.0 | 25587 | 0.4590 | 0.7844 | 0.7303 | 0.7574 |
| 0.3457 | 10.0 | 28430 | 0.4578 | 0.7937 | 0.7323 | 0.7630 |
| 0.3278 | 11.0 | 31273 | 0.4756 | 0.8034 | 0.7251 | 0.7643 |
| 0.3124 | 12.0 | 34116 | 0.4793 | 0.8026 | 0.7349 | 0.7688 |
| 0.2975 | 13.0 | 36959 | 0.4680 | 0.8009 | 0.7392 | 0.7701 |
| 0.2851 | 14.0 | 39802 | 0.4649 | 0.8061 | 0.7328 | 0.7695 |
| 0.273 | 15.0 | 42645 | 0.4699 | 0.7990 | 0.7379 | 0.7685 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.12.0
- Tokenizers 0.13.3
- Task
- Text Classification
- Library
- transformers
- Type
- hybridbert
- Languages
- en
- Created
- Jun 14, 2023
- Updated
- Jun 15, 2023