metadata

license: mit
base_model: roberta-base
tags:
  - generated_from_trainer
datasets:
  - ag_news
metrics:
  - accuracy
model-index:
  - name: N_roberta_agnews_padding90model
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: ag_news
          type: ag_news
          config: default
          split: test
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9481578947368421

N_roberta_agnews_padding90model

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

Loss: 0.5490
Accuracy: 0.9482

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: 16
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.1972	1.0	7500	0.2055	0.9412
0.1723	2.0	15000	0.1951	0.9463
0.1529	3.0	22500	0.2150	0.9461
0.1256	4.0	30000	0.2472	0.9457
0.1092	5.0	37500	0.2550	0.9459
0.0729	6.0	45000	0.2972	0.9479
0.0801	7.0	52500	0.3123	0.9453
0.058	8.0	60000	0.3705	0.9463
0.0463	9.0	67500	0.3698	0.9438
0.0387	10.0	75000	0.3702	0.9495
0.0299	11.0	82500	0.4177	0.9474
0.0235	12.0	90000	0.4637	0.9432
0.0172	13.0	97500	0.4843	0.9464
0.0144	14.0	105000	0.4647	0.9483
0.0147	15.0	112500	0.4965	0.9468
0.0124	16.0	120000	0.5153	0.9470
0.006	17.0	127500	0.5196	0.9483
0.0047	18.0	135000	0.5287	0.9480
0.0056	19.0	142500	0.5443	0.9480
0.0012	20.0	150000	0.5490	0.9482

Framework versions

Transformers 4.33.2
Pytorch 2.0.1+cu117
Datasets 2.14.5
Tokenizers 0.13.3