Improved-Arabert-twitter-sentiment2

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

• Loss: 0.4308
• Accuracy: 0.8759

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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.07	50	0.4102	0.8130
No log	0.14	100	0.3141	0.8769
No log	0.21	150	0.2981	0.8806
No log	0.27	200	0.3297	0.8769
No log	0.34	250	0.2998	0.8796
No log	0.41	300	0.3312	0.8630
No log	0.48	350	0.3615	0.8491
No log	0.55	400	0.3695	0.8481
No log	0.62	450	0.3094	0.8778
0.316	0.68	500	0.2784	0.8907
0.316	0.75	550	0.3404	0.8759
0.316	0.82	600	0.3045	0.8806
0.316	0.89	650	0.3435	0.8731
0.316	0.96	700	0.2849	0.9
0.316	1.03	750	0.2846	0.8963
0.316	1.1	800	0.3034	0.8926
0.316	1.16	850	0.3801	0.8787
0.316	1.23	900	0.3525	0.8898
0.316	1.3	950	0.3388	0.8889
0.2119	1.37	1000	0.3823	0.8843
0.2119	1.44	1050	0.3621	0.8935
0.2119	1.51	1100	0.4106	0.8843
0.2119	1.58	1150	0.3820	0.8870
0.2119	1.64	1200	0.3770	0.8796
0.2119	1.71	1250	0.4199	0.8824
0.2119	1.78	1300	0.4308	0.8759

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.7
• Tokenizers 0.14.1