metadata

license: mit
base_model: MoritzLaurer/DeBERTa-v3-base-mnli-fever-anli
tags:
  - generated_from_trainer
datasets:
  - sem_eval_2024_task_2
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: results2
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: sem_eval_2024_task_2
          type: sem_eval_2024_task_2
          config: sem_eval_2024_task_2_source
          split: validation
          args: sem_eval_2024_task_2_source
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.76
          - name: Precision
            type: precision
            value: 0.7601040416166467
          - name: Recall
            type: recall
            value: 0.76
          - name: F1
            type: f1
            value: 0.75997599759976

results2

This model is a fine-tuned version of MoritzLaurer/DeBERTa-v3-base-mnli-fever-anli on the sem_eval_2024_task_2 dataset. It achieves the following results on the evaluation set:

Loss: 2.1827
Accuracy: 0.76
Precision: 0.7601
Recall: 0.76
F1: 0.7600

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: 5e-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
lr_scheduler_warmup_steps: 500
num_epochs: 50
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.6925	1.0	107	0.6665	0.6	0.6457	0.6	0.5660
0.6729	2.0	214	0.6025	0.69	0.6964	0.69	0.6875
0.6857	3.0	321	0.6071	0.665	0.7531	0.665	0.6331
0.6667	4.0	428	0.5650	0.695	0.7157	0.6950	0.6875
0.6168	5.0	535	0.5036	0.75	0.7504	0.75	0.7499
0.5165	6.0	642	0.6248	0.67	0.6701	0.67	0.6700
0.4087	7.0	749	0.5246	0.735	0.7379	0.7350	0.7342
0.3083	8.0	856	0.6130	0.7	0.7	0.7	0.7
0.2909	9.0	963	0.7584	0.735	0.7723	0.7350	0.7256
0.319	10.0	1070	0.7350	0.72	0.7360	0.72	0.7152
0.1812	11.0	1177	0.9320	0.715	0.7176	0.7150	0.7141
0.2824	12.0	1284	0.9723	0.705	0.7336	0.7050	0.6957
0.2662	13.0	1391	0.8676	0.72	0.7222	0.72	0.7193
0.1641	14.0	1498	0.9450	0.71	0.7103	0.71	0.7099
0.2264	15.0	1605	1.1613	0.675	0.6764	0.675	0.6743
0.2077	16.0	1712	1.3497	0.715	0.7214	0.7150	0.7129
0.1767	17.0	1819	1.4154	0.705	0.7075	0.7050	0.7041
0.1751	18.0	1926	1.2369	0.735	0.7350	0.735	0.7350
0.1195	19.0	2033	1.1152	0.72	0.7334	0.72	0.7159
0.0507	20.0	2140	1.4853	0.715	0.7152	0.715	0.7149
0.0544	21.0	2247	1.7174	0.725	0.7302	0.7250	0.7234
0.0648	22.0	2354	1.7327	0.71	0.7121	0.71	0.7093
0.0039	23.0	2461	1.8211	0.725	0.7268	0.7250	0.7244
0.0153	24.0	2568	1.8315	0.715	0.7176	0.7150	0.7141
0.0017	25.0	2675	1.7446	0.72	0.7232	0.72	0.7190
0.0188	26.0	2782	1.6413	0.72	0.7274	0.72	0.7177
0.0168	27.0	2889	1.8013	0.73	0.7315	0.73	0.7296
0.0355	28.0	2996	2.0405	0.725	0.7354	0.725	0.7219
0.0168	29.0	3103	1.5087	0.735	0.7350	0.735	0.7350
0.0409	30.0	3210	1.5272	0.72	0.7244	0.72	0.7186
0.004	31.0	3317	1.9978	0.715	0.7214	0.7150	0.7129
0.0002	32.0	3424	1.9760	0.72	0.7244	0.72	0.7186
0.0111	33.0	3531	1.9985	0.74	0.7409	0.74	0.7398
0.052	34.0	3638	1.9607	0.73	0.7334	0.73	0.7290
0.0263	35.0	3745	1.7118	0.75	0.7525	0.75	0.7494
0.0101	36.0	3852	1.9553	0.755	0.7571	0.755	0.7545
0.0001	37.0	3959	2.0064	0.75	0.7537	0.75	0.7491
0.0186	38.0	4066	2.1726	0.74	0.7404	0.74	0.7399
0.0046	39.0	4173	2.1083	0.755	0.7550	0.755	0.7550
0.0042	40.0	4280	1.9944	0.76	0.7609	0.76	0.7598
0.0178	41.0	4387	2.0096	0.76	0.7604	0.76	0.7599
0.0089	42.0	4494	2.0431	0.765	0.7652	0.765	0.7649
0.0095	43.0	4601	2.0662	0.76	0.7604	0.76	0.7599
0.0162	44.0	4708	2.1703	0.745	0.7450	0.745	0.7450
0.0001	45.0	4815	2.1525	0.76	0.7601	0.76	0.7600
0.0001	46.0	4922	2.1581	0.76	0.7601	0.76	0.7600
0.0086	47.0	5029	2.1665	0.76	0.7601	0.76	0.7600
0.0088	48.0	5136	2.1747	0.76	0.7601	0.76	0.7600
0.0044	49.0	5243	2.1812	0.76	0.7601	0.76	0.7600
0.0043	50.0	5350	2.1827	0.76	0.7601	0.76	0.7600

Framework versions

Transformers 4.35.2
Pytorch 2.1.0+cu118
Datasets 2.15.0
Tokenizers 0.15.0