ArabicNewSplits4_WithDuplicationsForScore5_FineTuningAraBERT_run3_AugV5_k5_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: 0.6251
• Qwk: 0.4694
• Mse: 0.6251
• Rmse: 0.7907

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	0.0833	2	3.4150	-0.0160	3.4150	1.8480
No log	0.1667	4	1.7435	-0.0070	1.7435	1.3204
No log	0.25	6	1.0272	0.0335	1.0272	1.0135
No log	0.3333	8	1.0208	0.1405	1.0208	1.0104
No log	0.4167	10	0.9405	0.1736	0.9405	0.9698
No log	0.5	12	0.5890	0.0534	0.5890	0.7674
No log	0.5833	14	0.8211	0.1908	0.8211	0.9062
No log	0.6667	16	0.6020	0.1176	0.6020	0.7759
No log	0.75	18	0.6791	0.0739	0.6791	0.8241
No log	0.8333	20	0.5753	0.2516	0.5753	0.7585
No log	0.9167	22	0.5785	0.1756	0.5785	0.7606
No log	1.0	24	0.7771	0.1895	0.7771	0.8816
No log	1.0833	26	0.7128	0.1888	0.7128	0.8443
No log	1.1667	28	0.6123	0.0725	0.6123	0.7825
No log	1.25	30	0.6922	0.1732	0.6922	0.8320
No log	1.3333	32	0.7012	0.1605	0.7012	0.8374
No log	1.4167	34	0.7217	0.1707	0.7217	0.8495
No log	1.5	36	0.8159	0.1287	0.8159	0.9033
No log	1.5833	38	0.8331	0.1174	0.8331	0.9128
No log	1.6667	40	0.7404	0.0769	0.7404	0.8605
No log	1.75	42	1.2267	-0.0185	1.2267	1.1076
No log	1.8333	44	1.3862	-0.0196	1.3862	1.1774
No log	1.9167	46	0.8902	0.1675	0.8902	0.9435
No log	2.0	48	0.8758	0.1416	0.8758	0.9359
No log	2.0833	50	1.2632	0.1467	1.2632	1.1239
No log	2.1667	52	0.9453	0.1515	0.9453	0.9722
No log	2.25	54	0.7840	0.1759	0.7840	0.8854
No log	2.3333	56	1.0163	0.0968	1.0163	1.0081
No log	2.4167	58	0.8018	0.1759	0.8018	0.8954
No log	2.5	60	0.7951	0.1045	0.7951	0.8917
No log	2.5833	62	0.7563	0.1515	0.7563	0.8697
No log	2.6667	64	0.7493	0.1739	0.7493	0.8656
No log	2.75	66	0.8497	0.24	0.8497	0.9218
No log	2.8333	68	0.9383	0.1289	0.9383	0.9687
No log	2.9167	70	0.8001	0.2161	0.8001	0.8945
No log	3.0	72	0.8016	0.1636	0.8016	0.8953
No log	3.0833	74	0.8118	0.3214	0.8118	0.9010
No log	3.1667	76	1.0396	0.1373	1.0396	1.0196
No log	3.25	78	0.9867	0.1040	0.9867	0.9933
No log	3.3333	80	1.1141	0.0606	1.1141	1.0555
No log	3.4167	82	0.8770	0.2356	0.8770	0.9365
No log	3.5	84	0.9628	0.1304	0.9628	0.9812
No log	3.5833	86	0.8962	0.1273	0.8962	0.9467
No log	3.6667	88	0.8968	0.2579	0.8968	0.9470
No log	3.75	90	0.8814	0.1712	0.8814	0.9388
No log	3.8333	92	0.9067	0.1928	0.9067	0.9522
No log	3.9167	94	0.9232	0.2829	0.9232	0.9608
No log	4.0	96	0.9173	0.2520	0.9173	0.9578
No log	4.0833	98	0.8399	0.2340	0.8399	0.9164
No log	4.1667	100	0.8449	0.3147	0.8449	0.9192
No log	4.25	102	1.2335	0.1065	1.2335	1.1106
No log	4.3333	104	1.4293	0.1611	1.4293	1.1955
No log	4.4167	106	1.0125	0.1127	1.0125	1.0062
No log	4.5	108	0.8379	0.1852	0.8379	0.9154
No log	4.5833	110	0.9678	0.1417	0.9678	0.9837
No log	4.6667	112	0.8150	0.2212	0.8150	0.9028
No log	4.75	114	0.7526	0.3778	0.7526	0.8675
No log	4.8333	116	0.7260	0.3607	0.7260	0.8521
No log	4.9167	118	0.6948	0.3514	0.6948	0.8336
No log	5.0	120	0.6716	0.3052	0.6716	0.8195
No log	5.0833	122	0.6620	0.3077	0.6620	0.8136
No log	5.1667	124	0.6784	0.3571	0.6784	0.8237
No log	5.25	126	0.6316	0.3077	0.6316	0.7948
No log	5.3333	128	0.6568	0.2830	0.6568	0.8104
No log	5.4167	130	0.6543	0.3462	0.6543	0.8089
No log	5.5	132	0.8699	0.2829	0.8699	0.9327
No log	5.5833	134	0.9898	0.1746	0.9898	0.9949
No log	5.6667	136	0.7723	0.2900	0.7723	0.8788
No log	5.75	138	0.6683	0.2390	0.6683	0.8175
No log	5.8333	140	0.7710	0.1925	0.7710	0.8781
No log	5.9167	142	0.7059	0.2075	0.7059	0.8402
No log	6.0	144	0.6669	0.2709	0.6669	0.8166
No log	6.0833	146	0.9252	0.2000	0.9252	0.9619
No log	6.1667	148	0.9479	0.2000	0.9479	0.9736
No log	6.25	150	0.7282	0.2692	0.7282	0.8533
No log	6.3333	152	0.6482	0.2990	0.6482	0.8051
No log	6.4167	154	0.7996	0.1930	0.7996	0.8942
No log	6.5	156	0.7762	0.1930	0.7762	0.8810
No log	6.5833	158	0.6515	0.3684	0.6515	0.8072
No log	6.6667	160	0.6909	0.2161	0.6909	0.8312
No log	6.75	162	0.7002	0.2161	0.7002	0.8368
No log	6.8333	164	0.6704	0.3077	0.6704	0.8188
No log	6.9167	166	0.6531	0.3171	0.6531	0.8082
No log	7.0	168	0.6585	0.28	0.6585	0.8115
No log	7.0833	170	0.6480	0.3469	0.6480	0.8050
No log	7.1667	172	0.6499	0.3469	0.6499	0.8062
No log	7.25	174	0.6663	0.2709	0.6663	0.8163
No log	7.3333	176	0.6829	0.2233	0.6829	0.8264
No log	7.4167	178	0.6780	0.3143	0.6780	0.8234
No log	7.5	180	0.6772	0.2676	0.6772	0.8229
No log	7.5833	182	0.6698	0.3831	0.6698	0.8184
No log	7.6667	184	0.6716	0.2871	0.6716	0.8195
No log	7.75	186	0.6617	0.4	0.6617	0.8135
No log	7.8333	188	0.6627	0.3860	0.6627	0.8141
No log	7.9167	190	0.6842	0.3267	0.6842	0.8272
No log	8.0	192	0.6981	0.2850	0.6981	0.8355
No log	8.0833	194	0.6823	0.2780	0.6823	0.8260
No log	8.1667	196	0.6637	0.3860	0.6637	0.8147
No log	8.25	198	0.6601	0.3860	0.6601	0.8124
No log	8.3333	200	0.6573	0.4341	0.6573	0.8107
No log	8.4167	202	0.6532	0.3548	0.6532	0.8082
No log	8.5	204	0.6511	0.3786	0.6511	0.8069
No log	8.5833	206	0.6460	0.4229	0.6460	0.8038
No log	8.6667	208	0.6399	0.4229	0.6399	0.7999
No log	8.75	210	0.6361	0.4694	0.6361	0.7975
No log	8.8333	212	0.6462	0.3333	0.6462	0.8039
No log	8.9167	214	0.6817	0.2727	0.6817	0.8257
No log	9.0	216	0.7541	0.2632	0.7541	0.8684
No log	9.0833	218	0.8071	0.2618	0.8071	0.8984
No log	9.1667	220	0.8152	0.2618	0.8152	0.9029
No log	9.25	222	0.8004	0.2618	0.8004	0.8947
No log	9.3333	224	0.7517	0.2632	0.7517	0.8670
No log	9.4167	226	0.7089	0.3303	0.7089	0.8420
No log	9.5	228	0.6721	0.2332	0.6721	0.8198
No log	9.5833	230	0.6432	0.3333	0.6432	0.8020
No log	9.6667	232	0.6287	0.4694	0.6287	0.7929
No log	9.75	234	0.6238	0.4694	0.6238	0.7898
No log	9.8333	236	0.6237	0.4694	0.6237	0.7897
No log	9.9167	238	0.6247	0.4694	0.6247	0.7904
No log	10.0	240	0.6251	0.4694	0.6251	0.7907

Framework versions

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