ArabicNewSplits5_FineTuningAraBERT_run3_AugV5_k2_task3_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: 0.7080
• Qwk: 0.2850
• Mse: 0.7080
• Rmse: 0.8414

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	3.3385	-0.0258	3.3385	1.8272
No log	0.2667	4	1.6909	-0.0070	1.6909	1.3003
No log	0.4	6	1.2433	0.0294	1.2433	1.1150
No log	0.5333	8	1.4011	0.0588	1.4011	1.1837
No log	0.6667	10	0.9876	0.1496	0.9876	0.9938
No log	0.8	12	0.6144	-0.0370	0.6144	0.7839
No log	0.9333	14	0.5942	-0.0233	0.5942	0.7708
No log	1.0667	16	0.5885	-0.0794	0.5885	0.7672
No log	1.2	18	0.7927	0.1453	0.7927	0.8904
No log	1.3333	20	1.1767	0.0698	1.1767	1.0848
No log	1.4667	22	0.8037	0.1392	0.8037	0.8965
No log	1.6	24	0.7612	0.2072	0.7612	0.8725
No log	1.7333	26	0.6545	-0.0133	0.6545	0.8090
No log	1.8667	28	0.6236	-0.0853	0.6236	0.7897
No log	2.0	30	0.6344	0.0	0.6344	0.7965
No log	2.1333	32	0.6441	0.0	0.6441	0.8026
No log	2.2667	34	0.6139	-0.0081	0.6139	0.7835
No log	2.4	36	0.6671	0.0850	0.6671	0.8168
No log	2.5333	38	0.8028	0.0647	0.8028	0.8960
No log	2.6667	40	0.7143	0.1111	0.7143	0.8452
No log	2.8	42	0.6205	0.1111	0.6205	0.7877
No log	2.9333	44	0.5843	0.1020	0.5843	0.7644
No log	3.0667	46	0.5976	0.0850	0.5976	0.7730
No log	3.2	48	0.5941	0.1678	0.5941	0.7708
No log	3.3333	50	0.7213	0.1529	0.7213	0.8493
No log	3.4667	52	0.8263	0.1158	0.8263	0.9090
No log	3.6	54	0.7427	0.2265	0.7427	0.8618
No log	3.7333	56	0.6162	0.2809	0.6162	0.7850
No log	3.8667	58	0.8994	0.2640	0.8994	0.9484
No log	4.0	60	0.8501	0.2275	0.8501	0.9220
No log	4.1333	62	0.6342	0.2174	0.6342	0.7964
No log	4.2667	64	0.7981	0.1402	0.7981	0.8934
No log	4.4	66	0.8773	0.0049	0.8773	0.9367
No log	4.5333	68	0.7387	0.1304	0.7387	0.8595
No log	4.6667	70	0.6596	0.1739	0.6596	0.8122
No log	4.8	72	0.9061	0.1131	0.9061	0.9519
No log	4.9333	74	0.8600	0.1336	0.8600	0.9274
No log	5.0667	76	0.6936	0.1913	0.6936	0.8328
No log	5.2	78	0.7603	0.1675	0.7603	0.8719
No log	5.3333	80	0.9750	0.0279	0.9750	0.9874
No log	5.4667	82	1.0119	0.1515	1.0119	1.0060
No log	5.6	84	0.8151	0.1718	0.8151	0.9028
No log	5.7333	86	0.7504	0.2423	0.7504	0.8662
No log	5.8667	88	0.8756	0.2469	0.8756	0.9358
No log	6.0	90	0.7823	0.2348	0.7823	0.8845
No log	6.1333	92	0.7036	0.3200	0.7036	0.8388
No log	6.2667	94	0.9075	0.1506	0.9075	0.9526
No log	6.4	96	0.9565	0.1818	0.9565	0.9780
No log	6.5333	98	0.8154	0.2143	0.8154	0.9030
No log	6.6667	100	0.6892	0.3663	0.6892	0.8302
No log	6.8	102	0.6979	0.2475	0.6979	0.8354
No log	6.9333	104	0.7541	0.1848	0.7541	0.8684
No log	7.0667	106	0.8253	0.1570	0.8253	0.9084
No log	7.2	108	0.7460	0.2074	0.7460	0.8637
No log	7.3333	110	0.6841	0.3061	0.6841	0.8271
No log	7.4667	112	0.8200	0.2000	0.8200	0.9055
No log	7.6	114	0.9356	0.1545	0.9356	0.9673
No log	7.7333	116	0.9215	0.1807	0.9215	0.9599
No log	7.8667	118	0.8087	0.2287	0.8087	0.8993
No log	8.0	120	0.7111	0.2830	0.7111	0.8433
No log	8.1333	122	0.6965	0.3263	0.6965	0.8345
No log	8.2667	124	0.6882	0.2990	0.6882	0.8296
No log	8.4	126	0.6860	0.3089	0.6860	0.8283
No log	8.5333	128	0.6846	0.3089	0.6846	0.8274
No log	8.6667	130	0.7019	0.2487	0.7019	0.8378
No log	8.8	132	0.7228	0.2536	0.7228	0.8501
No log	8.9333	134	0.7172	0.2536	0.7172	0.8469
No log	9.0667	136	0.7039	0.2464	0.7039	0.8390
No log	9.2	138	0.6924	0.3684	0.6924	0.8321
No log	9.3333	140	0.6928	0.3684	0.6928	0.8323
No log	9.4667	142	0.6969	0.3641	0.6969	0.8348
No log	9.6	144	0.7050	0.2850	0.7050	0.8397
No log	9.7333	146	0.7058	0.2850	0.7058	0.8401
No log	9.8667	148	0.7082	0.2850	0.7082	0.8415
No log	10.0	150	0.7080	0.2850	0.7080	0.8414

Framework versions

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