Version3ASAP_FineTuningBERT_AugV14_k3_task1_organization_k3_k3_fold1

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

• Loss: 0.6729
• Qwk: 0.4639
• Mse: 0.6720
• Rmse: 0.8198

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Qwk	Mse	Rmse
No log	1.0	3	9.0966	0.0	9.0940	3.0156
No log	2.0	6	8.4419	0.0	8.4394	2.9051
No log	3.0	9	6.4159	0.0	6.4136	2.5325
No log	4.0	12	5.2018	0.0040	5.1996	2.2803
No log	5.0	15	4.0671	0.0	4.0650	2.0162
No log	6.0	18	3.0890	0.0	3.0872	1.7570
No log	7.0	21	2.4075	0.0130	2.4057	1.5510
No log	8.0	24	1.9348	0.0211	1.9331	1.3904
No log	9.0	27	1.4871	0.0	1.4855	1.2188
No log	10.0	30	1.2310	0.0	1.2295	1.1088
No log	11.0	33	1.1277	0.0	1.1264	1.0613
No log	12.0	36	1.1318	0.0315	1.1306	1.0633
No log	13.0	39	1.3079	0.0921	1.3066	1.1431
No log	14.0	42	0.9406	0.0211	0.9393	0.9692
No log	15.0	45	1.3367	0.1621	1.3355	1.1556
No log	16.0	48	1.2721	0.1918	1.2710	1.1274
No log	17.0	51	0.8128	0.3399	0.8117	0.9010
No log	18.0	54	1.0414	0.2475	1.0404	1.0200
No log	19.0	57	0.7162	0.4805	0.7152	0.8457
No log	20.0	60	0.9405	0.4317	0.9395	0.9693
No log	21.0	63	0.6584	0.4827	0.6574	0.8108
No log	22.0	66	0.8425	0.4801	0.8415	0.9173
No log	23.0	69	0.6581	0.4787	0.6571	0.8106
No log	24.0	72	0.8453	0.4781	0.8443	0.9189
No log	25.0	75	0.6620	0.4740	0.6611	0.8131
No log	26.0	78	0.7372	0.4777	0.7362	0.8580
No log	27.0	81	0.7330	0.5392	0.7320	0.8556
No log	28.0	84	0.9004	0.4905	0.8992	0.9483
No log	29.0	87	0.6992	0.5583	0.6983	0.8356
No log	30.0	90	0.8284	0.4360	0.8273	0.9096
No log	31.0	93	0.6354	0.5317	0.6345	0.7965
No log	32.0	96	0.6318	0.4991	0.6309	0.7943
No log	33.0	99	0.9968	0.3960	0.9957	0.9978
No log	34.0	102	0.6545	0.5274	0.6536	0.8085
No log	35.0	105	0.6306	0.5069	0.6296	0.7935
No log	36.0	108	0.8291	0.4420	0.8279	0.9099
No log	37.0	111	0.6919	0.5162	0.6910	0.8312
No log	38.0	114	0.6214	0.5275	0.6205	0.7877
No log	39.0	117	0.7159	0.4554	0.7149	0.8455
No log	40.0	120	0.6125	0.5356	0.6116	0.7821
No log	41.0	123	0.7335	0.5055	0.7325	0.8559
No log	42.0	126	0.6880	0.5177	0.6870	0.8288
No log	43.0	129	0.6104	0.5356	0.6095	0.7807
No log	44.0	132	0.7677	0.4429	0.7666	0.8755
No log	45.0	135	0.6317	0.5317	0.6308	0.7942
No log	46.0	138	0.6310	0.5221	0.6301	0.7938
No log	47.0	141	0.7758	0.4124	0.7748	0.8802
No log	48.0	144	0.6201	0.5082	0.6192	0.7869
No log	49.0	147	0.6729	0.4639	0.6720	0.8198

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0