ArabicNewSplits6_FineTuningAraBERT_run3_AugV5_k2_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: 0.6671
• Qwk: 0.6974
• Mse: 0.6671
• Rmse: 0.8167

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	0.1429	2	5.4036	-0.0607	5.4036	2.3246
No log	0.2857	4	3.4226	0.0699	3.4226	1.8500
No log	0.4286	6	2.0746	0.0686	2.0746	1.4403
No log	0.5714	8	2.1220	-0.1206	2.1220	1.4567
No log	0.7143	10	1.7488	-0.0557	1.7488	1.3224
No log	0.8571	12	1.1309	0.2244	1.1309	1.0634
No log	1.0	14	1.0081	0.3512	1.0081	1.0040
No log	1.1429	16	1.1122	0.4034	1.1122	1.0546
No log	1.2857	18	1.2652	0.3733	1.2652	1.1248
No log	1.4286	20	1.0840	0.4516	1.0840	1.0411
No log	1.5714	22	0.8305	0.4377	0.8305	0.9113
No log	1.7143	24	0.8954	0.5642	0.8954	0.9462
No log	1.8571	26	0.7952	0.5460	0.7952	0.8917
No log	2.0	28	0.7532	0.4901	0.7532	0.8679
No log	2.1429	30	1.0572	0.4589	1.0572	1.0282
No log	2.2857	32	1.4316	0.3599	1.4316	1.1965
No log	2.4286	34	1.2943	0.4359	1.2943	1.1377
No log	2.5714	36	0.8761	0.5258	0.8761	0.9360
No log	2.7143	38	0.6940	0.5978	0.6940	0.8331
No log	2.8571	40	0.6180	0.6794	0.6180	0.7861
No log	3.0	42	0.6043	0.6840	0.6043	0.7774
No log	3.1429	44	0.5957	0.6787	0.5957	0.7718
No log	3.2857	46	0.6139	0.7249	0.6139	0.7835
No log	3.4286	48	0.7543	0.6925	0.7543	0.8685
No log	3.5714	50	1.0967	0.5348	1.0967	1.0472
No log	3.7143	52	1.0532	0.5603	1.0532	1.0263
No log	3.8571	54	0.7502	0.7004	0.7502	0.8661
No log	4.0	56	0.6011	0.7488	0.6011	0.7753
No log	4.1429	58	0.7006	0.7249	0.7006	0.8370
No log	4.2857	60	0.6896	0.7231	0.6896	0.8304
No log	4.4286	62	0.5953	0.7253	0.5953	0.7716
No log	4.5714	64	0.6813	0.6865	0.6813	0.8254
No log	4.7143	66	1.1222	0.5641	1.1222	1.0593
No log	4.8571	68	1.2716	0.5026	1.2716	1.1277
No log	5.0	70	1.0454	0.6121	1.0454	1.0224
No log	5.1429	72	0.7051	0.6638	0.7051	0.8397
No log	5.2857	74	0.5653	0.7759	0.5653	0.7519
No log	5.4286	76	0.6266	0.7387	0.6266	0.7916
No log	5.5714	78	0.7251	0.7164	0.7251	0.8515
No log	5.7143	80	0.7244	0.7144	0.7244	0.8511
No log	5.8571	82	0.6263	0.7467	0.6263	0.7914
No log	6.0	84	0.5819	0.7670	0.5819	0.7628
No log	6.1429	86	0.6471	0.7182	0.6471	0.8044
No log	6.2857	88	0.7067	0.6879	0.7067	0.8406
No log	6.4286	90	0.6720	0.7055	0.6720	0.8198
No log	6.5714	92	0.5988	0.7410	0.5988	0.7738
No log	6.7143	94	0.5634	0.7398	0.5634	0.7506
No log	6.8571	96	0.5652	0.7591	0.5652	0.7518
No log	7.0	98	0.5678	0.7497	0.5678	0.7535
No log	7.1429	100	0.6251	0.7158	0.6251	0.7906
No log	7.2857	102	0.6673	0.7093	0.6673	0.8169
No log	7.4286	104	0.6510	0.7093	0.6510	0.8068
No log	7.5714	106	0.6540	0.7128	0.6540	0.8087
No log	7.7143	108	0.6296	0.7314	0.6296	0.7935
No log	7.8571	110	0.6195	0.7370	0.6195	0.7871
No log	8.0	112	0.6149	0.7350	0.6149	0.7842
No log	8.1429	114	0.6015	0.7359	0.6015	0.7756
No log	8.2857	116	0.5851	0.7494	0.5851	0.7649
No log	8.4286	118	0.5916	0.7494	0.5916	0.7692
No log	8.5714	120	0.5958	0.7494	0.5958	0.7719
No log	8.7143	122	0.5948	0.7439	0.5948	0.7712
No log	8.8571	124	0.6006	0.7446	0.6006	0.7750
No log	9.0	126	0.6219	0.7370	0.6219	0.7886
No log	9.1429	128	0.6589	0.6974	0.6589	0.8117
No log	9.2857	130	0.6896	0.7017	0.6896	0.8304
No log	9.4286	132	0.6983	0.6921	0.6983	0.8357
No log	9.5714	134	0.6906	0.6957	0.6906	0.8310
No log	9.7143	136	0.6796	0.6974	0.6796	0.8244
No log	9.8571	138	0.6696	0.6974	0.6696	0.8183
No log	10.0	140	0.6671	0.6974	0.6671	0.8167

Framework versions

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