DeBERTa-v3-small fine-tuned on CoLA
This model is a fine-tuned version of microsoft/deberta-v3-small on the GLUE COLA dataset. It achieves the following results on the evaluation set:
- Loss: 0.4051
- Matthews Correlation: 0.6333
Model description
DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data.
Please check the official repository for more details and updates.
In DeBERTa V3, we replaced the MLM objective with the RTD(Replaced Token Detection) objective introduced by ELECTRA for pre-training, as well as some innovations to be introduced in our upcoming paper. Compared to DeBERTa-V2, our V3 version significantly improves the model performance in downstream tasks. You can find a simple introduction about the model from the appendix A11 in our original paper, but we will provide more details in a separate write-up.
The DeBERTa V3 small model comes with 6 layers and a hidden size of 768. Its total parameter number is 143M since we use a vocabulary containing 128K tokens which introduce 98M parameters in the Embedding layer. This model was trained using the 160GB data as DeBERTa V2.
Intended uses & limitations
More information needed
Training and evaluation data
The Corpus of Linguistic Acceptability (CoLA) in its full form consists of 10657 sentences from 23 linguistics publications, expertly annotated for acceptability (grammaticality) by their original authors. The public version provided here contains 9594 sentences belonging to training and development sets, and excludes 1063 sentences belonging to a held out test set.
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 3e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
Training results
Training Loss | Epoch | Step | Validation Loss | Matthews Correlation |
---|---|---|---|---|
No log | 1.0 | 535 | 0.4051 | 0.6333 |
0.3371 | 2.0 | 1070 | 0.4455 | 0.6531 |
0.3371 | 3.0 | 1605 | 0.5755 | 0.6499 |
0.1305 | 4.0 | 2140 | 0.7188 | 0.6553 |
0.1305 | 5.0 | 2675 | 0.8047 | 0.6700 |
Framework versions
- Transformers 4.13.0.dev0
- Pytorch 1.10.0+cu111
- Datasets 1.16.1
- Tokenizers 0.10.3
- Downloads last month
- 320
Dataset used to train mrm8488/deberta-v3-small-finetuned-cola
Evaluation results
- Matthews Correlation on GLUE COLAself-reported0.633
- Accuracy on gluevalidation set verified0.849
- Precision on gluevalidation set verified0.846
- Recall on gluevalidation set verified0.957
- AUC on gluevalidation set verified0.917
- F1 on gluevalidation set verified0.898
- loss on gluevalidation set verified0.405