Arabic_FineTuningAraBERT_AugV0_k1_task2_organization_fold1

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: 0.5695
• Qwk: 0.5185
• Mse: 0.5695
• Rmse: 0.7546

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	0.1333	2	3.8687	0.0	3.8687	1.9669
No log	0.2667	4	1.7810	0.0792	1.7810	1.3345
No log	0.4	6	0.9379	0.0548	0.9379	0.9685
No log	0.5333	8	0.6310	0.3636	0.6310	0.7943
No log	0.6667	10	0.7225	0.25	0.7225	0.8500
No log	0.8	12	0.5897	0.4474	0.5897	0.7679
No log	0.9333	14	0.6534	0.3200	0.6534	0.8083
No log	1.0667	16	0.7022	0.5610	0.7022	0.8380
No log	1.2	18	0.8412	0.4375	0.8412	0.9172
No log	1.3333	20	0.9771	0.3306	0.9771	0.9885
No log	1.4667	22	0.7839	0.4000	0.7839	0.8854
No log	1.6	24	0.6065	0.3478	0.6065	0.7788
No log	1.7333	26	0.5811	0.3000	0.5811	0.7623
No log	1.8667	28	0.6142	0.4407	0.6142	0.7837
No log	2.0	30	0.8109	0.3182	0.8109	0.9005
No log	2.1333	32	0.8199	0.3368	0.8199	0.9055
No log	2.2667	34	0.7494	0.4545	0.7494	0.8657
No log	2.4	36	0.7815	0.4828	0.7815	0.8840
No log	2.5333	38	0.6163	0.2500	0.6163	0.7851
No log	2.6667	40	0.5584	0.1429	0.5584	0.7473
No log	2.8	42	0.6019	0.4000	0.6019	0.7758
No log	2.9333	44	0.8420	0.4571	0.8420	0.9176
No log	3.0667	46	1.3814	0.3137	1.3814	1.1753
No log	3.2	48	1.3002	0.2657	1.3002	1.1403
No log	3.3333	50	0.7816	0.4950	0.7816	0.8841
No log	3.4667	52	0.5511	0.475	0.5511	0.7424
No log	3.6	54	0.5443	0.3544	0.5443	0.7377
No log	3.7333	56	0.5036	0.4156	0.5036	0.7097
No log	3.8667	58	0.6528	0.475	0.6528	0.8080
No log	4.0	60	0.7194	0.4444	0.7194	0.8482
No log	4.1333	62	0.5702	0.4545	0.5702	0.7551
No log	4.2667	64	0.4660	0.56	0.4660	0.6826
No log	4.4	66	0.4382	0.48	0.4382	0.6620
No log	4.5333	68	0.4401	0.48	0.4401	0.6634
No log	4.6667	70	0.4499	0.5185	0.4499	0.6708
No log	4.8	72	0.5654	0.5352	0.5654	0.7519
No log	4.9333	74	0.6775	0.4658	0.6775	0.8231
No log	5.0667	76	0.5957	0.5352	0.5957	0.7718
No log	5.2	78	0.4905	0.5185	0.4905	0.7004
No log	5.3333	80	0.4941	0.5618	0.4941	0.7029
No log	5.4667	82	0.5341	0.5161	0.5341	0.7308
No log	5.6	84	0.5716	0.5161	0.5716	0.7561
No log	5.7333	86	0.5986	0.6	0.5986	0.7737
No log	5.8667	88	0.6613	0.6	0.6613	0.8132
No log	6.0	90	0.7771	0.5545	0.7771	0.8815
No log	6.1333	92	0.8015	0.4915	0.8015	0.8953
No log	6.2667	94	0.6725	0.56	0.6725	0.8201
No log	6.4	96	0.5343	0.6	0.5343	0.7310
No log	6.5333	98	0.4953	0.5185	0.4953	0.7037
No log	6.6667	100	0.4732	0.5185	0.4732	0.6879
No log	6.8	102	0.4524	0.5385	0.4524	0.6726
No log	6.9333	104	0.4691	0.5135	0.4691	0.6849
No log	7.0667	106	0.5222	0.5217	0.5222	0.7226
No log	7.2	108	0.5494	0.5185	0.5494	0.7412
No log	7.3333	110	0.5864	0.5263	0.5864	0.7657
No log	7.4667	112	0.6200	0.4179	0.6200	0.7874
No log	7.6	114	0.5929	0.5263	0.5929	0.7700
No log	7.7333	116	0.5386	0.5185	0.5386	0.7339
No log	7.8667	118	0.5065	0.5714	0.5065	0.7117
No log	8.0	120	0.5039	0.5610	0.5039	0.7098
No log	8.1333	122	0.5134	0.5610	0.5134	0.7165
No log	8.2667	124	0.5088	0.5610	0.5088	0.7133
No log	8.4	126	0.5132	0.6024	0.5132	0.7164
No log	8.5333	128	0.5256	0.5714	0.5256	0.7250
No log	8.6667	130	0.5409	0.5714	0.5409	0.7355
No log	8.8	132	0.5475	0.5714	0.5475	0.7400
No log	8.9333	134	0.5537	0.5714	0.5537	0.7441
No log	9.0667	136	0.5629	0.5185	0.5629	0.7503
No log	9.2	138	0.5778	0.5185	0.5778	0.7601
No log	9.3333	140	0.5890	0.4878	0.5890	0.7674
No log	9.4667	142	0.5907	0.4878	0.5907	0.7686
No log	9.6	144	0.5858	0.5185	0.5858	0.7654
No log	9.7333	146	0.5778	0.5185	0.5778	0.7601
No log	9.8667	148	0.5722	0.5185	0.5722	0.7564
No log	10.0	150	0.5695	0.5185	0.5695	0.7546

Framework versions

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