README.md

metadata

tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - accuracy
model-index:
  - name: train
    results: []

train

This model is a fine-tuned version of aubmindlab/bert-base-arabertv02 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9723
• Macro F1: 0.7972
• Precision: 0.8006
• Recall: 0.8116
• Kappa: 0.7148
• Accuracy: 0.8116

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: 128
• seed: 25
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Macro F1	Precision	Recall	Kappa	Accuracy
No log	1.0	101	1.2223	0.5901	0.5634	0.6884	0.4728	0.6884
No log	2.0	203	1.0235	0.6570	0.6224	0.7303	0.5496	0.7303
No log	3.0	304	0.9087	0.7101	0.7166	0.7623	0.6226	0.7623
No log	4.0	406	0.8119	0.7500	0.7346	0.7808	0.6711	0.7808
0.9826	5.0	507	0.8113	0.7821	0.7832	0.8030	0.6989	0.8030
0.9826	6.0	609	0.8290	0.7765	0.7719	0.7943	0.6919	0.7943
0.9826	7.0	710	0.8481	0.7756	0.7696	0.7980	0.6907	0.7980
0.9826	8.0	812	0.8620	0.7820	0.7747	0.8030	0.6974	0.8030
0.9826	9.0	913	0.8717	0.7891	0.7878	0.8042	0.7055	0.8042
0.3153	10.0	1015	0.9027	0.7872	0.7879	0.8067	0.7070	0.8067
0.3153	11.0	1116	0.9492	0.7902	0.7915	0.8067	0.7066	0.8067
0.3153	12.0	1218	0.9462	0.7877	0.7850	0.8042	0.7037	0.8042
0.3153	13.0	1319	0.9696	0.7881	0.7892	0.8030	0.7022	0.8030
0.3153	14.0	1421	0.9658	0.7931	0.7975	0.8067	0.7090	0.8067
0.1362	14.93	1515	0.9723	0.7972	0.8006	0.8116	0.7148	0.8116

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Tokenizers 0.13.3