metadata

language:
  - en
license: mit
tags:
  - generated_from_trainer
datasets:
  - glue
metrics:
  - matthews_correlation
widget:
  - text: They represented seriously to the dean Mary as a genuine linguist.
model-index:
  - name: deberta-v3-small
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: GLUE COLA
          type: glue
          args: cola
        metrics:
          - type: matthews_correlation
            value: 0.6333205721749096
            name: Matthews Correlation
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: glue
          type: glue
          config: cola
          split: validation
        metrics:
          - type: accuracy
            value: 0.8494726749760306
            name: Accuracy
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMjJjOTM0MTEzMzBlZWJlMWYwNzgzZmI3M2NiZWVjMDQ5ZDA1MWY0NGY3NjU1NTlmZWE3N2JjZWEzODE0ZTNkNSIsInZlcnNpb24iOjF9.Kt-3jnDTp3-Te5zMHVgG_5hpB5UMCkAMP7fmjx46QDWJfFHpyRgBlf-qz_fw5saFPAQ5G6QNq3bjEJ6mY2lhAw
          - type: precision
            value: 0.8455882352941176
            name: Precision
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiODAxMzNkZGEwNGNmYjk4NWRhZDk4OWE4MzA5Y2NiNjQyNTdkOWRmYjU0ZjY0YzQzYmE4ZmI3MjQ4OTk4OWIwNCIsInZlcnNpb24iOjF9.YBFnePtD5-HX15aST39xpPLroFYBgqEn5iLyVaClh62j0M7HQbB8aaGEbgaTIUIr-qz12gVfIQ7UZZIHxby_BQ
          - type: recall
            value: 0.957004160887656
            name: Recall
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMjRjMTVhN2E4YjNlOWY2MWRhODZiM2FhZDVjNzYwMjIyNWUyYTMxMWFlZjkwNzVhYjNmMjQxYjk2MTFmMzYyYiIsInZlcnNpb24iOjF9.40GYlU9Do74Y_gLmbIKR2WM8okz5fm-QUwJAsoIyM1UtQ71lKd-FV5Yr9CdAh3fyQYa3SMYe6tm9OByNMMw_AA
          - type: auc
            value: 0.9167413271767129
            name: AUC
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiYzVjYmMyZDkyMzM0ZTQ1MTk0ZmY4MWUwZmIxMGRlOWMyMjJmNDRiZGNkMGZlZDZmY2I5OWI2NDYzMGQ2YzhiNSIsInZlcnNpb24iOjF9.setZF_g9x-aknFXM1k0NxrOWMJcmpNi6z7QlyfL0i6fTPJOj6SbKJ1WQb3J1zTuabgx9cOc5xgHtBH3IA7fkDQ
          - type: f1
            value: 0.8978529603122967
            name: F1
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNmQ1NmNiMDhmNTU2Y2UxMzU0ODRmYmZmZTFkYjI4MzczMWUwYWQ4OTk2NGJlY2MzNmViYTA4MTRkODJhMTU1MyIsInZlcnNpb24iOjF9.GUIRxsYKgjYK63JS2rd9vCLHHmCiB4H68Xo5GxMaITfyzcUcdNc6l62njmQGrOoUidlTt1F7DzGP2Cu_Gz8HDg
          - type: loss
            value: 0.4050811529159546
            name: loss
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNjBjNjg0OTFjOTc5Mzc2MWQ1ZDIyYmM5MmIzZDVlY2JjYzBlZjMyN2IwOWU4YzNlMDcwZmM0NTMxYjExY2I0MiIsInZlcnNpb24iOjF9.xayLZc97iUW0zNqG65TiW9BXoqzV-tqF8g9qGCYQ1ZGuSDSjLlK7Y4og7-wqPEiME8JtNyVxl6-ZcWnF1t8cDg

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