ArabicNewSplits5_FineTuningAraBERT_run2_AugV5_k1_task1_organization

This model is a fine-tuned version of aubmindlab/bert-base-arabertv02 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8108
• Qwk: 0.6625
• Mse: 0.8108
• Rmse: 0.9005

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: 8
• eval_batch_size: 8
• 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	Qwk	Mse	Rmse
No log	0.2	2	5.0351	-0.0123	5.0351	2.2439
No log	0.4	4	2.9189	0.0825	2.9189	1.7085
No log	0.6	6	1.7659	0.1067	1.7659	1.3289
No log	0.8	8	1.1997	0.3106	1.1997	1.0953
No log	1.0	10	1.2579	0.1923	1.2579	1.1216
No log	1.2	12	1.3240	0.2229	1.3240	1.1506
No log	1.4	14	1.4189	0.1871	1.4189	1.1912
No log	1.6	16	1.1322	0.3957	1.1322	1.0640
No log	1.8	18	1.0731	0.3447	1.0731	1.0359
No log	2.0	20	1.0890	0.3504	1.0890	1.0435
No log	2.2	22	1.0719	0.3725	1.0719	1.0353
No log	2.4	24	1.1932	0.2594	1.1932	1.0924
No log	2.6	26	1.2032	0.3361	1.2032	1.0969
No log	2.8	28	1.2319	0.3907	1.2319	1.1099
No log	3.0	30	1.0915	0.4277	1.0915	1.0447
No log	3.2	32	0.9320	0.4739	0.9320	0.9654
No log	3.4	34	0.8798	0.5045	0.8798	0.9380
No log	3.6	36	0.9182	0.5247	0.9182	0.9582
No log	3.8	38	0.9707	0.5294	0.9707	0.9853
No log	4.0	40	0.8344	0.5791	0.8344	0.9135
No log	4.2	42	0.8074	0.6388	0.8074	0.8985
No log	4.4	44	0.8281	0.6075	0.8281	0.9100
No log	4.6	46	0.8086	0.6299	0.8086	0.8992
No log	4.8	48	0.8186	0.5752	0.8186	0.9048
No log	5.0	50	1.0059	0.5563	1.0059	1.0030
No log	5.2	52	1.1557	0.5321	1.1557	1.0750
No log	5.4	54	1.0344	0.5739	1.0344	1.0170
No log	5.6	56	0.8369	0.6009	0.8369	0.9148
No log	5.8	58	0.7614	0.5831	0.7614	0.8726
No log	6.0	60	0.7683	0.6572	0.7683	0.8765
No log	6.2	62	0.7654	0.6966	0.7654	0.8749
No log	6.4	64	0.7572	0.6812	0.7572	0.8702
No log	6.6	66	0.7588	0.6097	0.7588	0.8711
No log	6.8	68	0.8309	0.6110	0.8309	0.9115
No log	7.0	70	0.8817	0.5807	0.8817	0.9390
No log	7.2	72	0.8659	0.5693	0.8659	0.9306
No log	7.4	74	0.7926	0.5975	0.7926	0.8903
No log	7.6	76	0.7699	0.6414	0.7699	0.8774
No log	7.8	78	0.7918	0.6964	0.7918	0.8898
No log	8.0	80	0.8038	0.6919	0.8038	0.8965
No log	8.2	82	0.8007	0.7081	0.8007	0.8948
No log	8.4	84	0.8056	0.6893	0.8056	0.8975
No log	8.6	86	0.8019	0.6938	0.8019	0.8955
No log	8.8	88	0.7989	0.6945	0.7989	0.8938
No log	9.0	90	0.8017	0.6714	0.8017	0.8954
No log	9.2	92	0.8060	0.6671	0.8060	0.8978
No log	9.4	94	0.8075	0.6625	0.8075	0.8986
No log	9.6	96	0.8085	0.6448	0.8085	0.8991
No log	9.8	98	0.8102	0.6448	0.8102	0.9001
No log	10.0	100	0.8108	0.6625	0.8108	0.9005

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu118
• Datasets 2.21.0
• Tokenizers 0.19.1