Add evaluation results on the qqp config of glue

6ac1079 over 3 years ago

3.34 kB

language:
  - en
license: mit
tags:
  - generated_from_trainer
datasets:
  - glue
metrics:
  - accuracy
  - f1
model-index:
  - name: roberta-base-qqp
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: GLUE QQP
          type: glue
          args: qqp
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9152609448429384
          - name: F1
            type: f1
            value: 0.8867138416771377
      - task:
          type: natural-language-inference
          name: Natural Language Inference
        dataset:
          name: glue
          type: glue
          config: qqp
          split: validation
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9153104130596093
            verified: true
          - name: Precision
            type: precision
            value: 0.8732009117551286
            verified: true
          - name: Recall
            type: recall
            value: 0.9007725898555593
            verified: true
          - name: AUC
            type: auc
            value: 0.9685235648551861
            verified: true
          - name: F1
            type: f1
            value: 0.8867724867724867
            verified: true
          - name: loss
            type: loss
            value: 0.4435121417045593
            verified: true

roberta-base-qqp

This model is a fine-tuned version of roberta-base on the GLUE QQP dataset. It achieves the following results on the evaluation set:

Loss: 0.4435
Accuracy: 0.9153
F1: 0.8867
Combined Score: 0.9010

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: 2e-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
lr_scheduler_warmup_ratio: 0.06
num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Combined Score
0.2751	1.0	22741	0.3057	0.8905	0.8512	0.8709
0.2443	2.0	45482	0.2530	0.9005	0.8710	0.8857
0.2157	3.0	68223	0.2643	0.9070	0.8769	0.8919
0.1838	4.0	90964	0.2806	0.9109	0.8815	0.8962
0.146	5.0	113705	0.3277	0.9113	0.8809	0.8961
0.1262	6.0	136446	0.3939	0.9113	0.8812	0.8962
0.0867	7.0	159187	0.4435	0.9153	0.8867	0.9010
0.0757	8.0	181928	0.4812	0.9147	0.8844	0.8996
0.0479	9.0	204669	0.5081	0.9151	0.8871	0.9011
0.0379	10.0	227410	0.5647	0.9149	0.8858	0.9003

Framework versions

Transformers 4.20.0.dev0
Pytorch 1.11.0+cu113
Datasets 2.1.0
Tokenizers 0.12.1