Version2ASAP_FineTuningBERT_AugV5_k3_task1_organization_k3_k3_fold2

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: 0.7558
• Qwk: 0.5156
• Mse: 0.7557
• Rmse: 0.8693

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	8.1704	0.0	8.1703	2.8584
No log	2.0	6	5.6337	0.0151	5.6339	2.3736
No log	3.0	9	3.9708	0.0	3.9711	1.9928
No log	4.0	12	2.8912	0.0	2.8917	1.7005
No log	5.0	15	2.0305	0.0813	2.0309	1.4251
No log	6.0	18	1.5637	0.0241	1.5641	1.2506
No log	7.0	21	1.1896	0.0107	1.1900	1.0909
No log	8.0	24	1.0673	0.0174	1.0677	1.0333
No log	9.0	27	1.3990	0.0952	1.3994	1.1830
No log	10.0	30	1.4724	0.1278	1.4729	1.2136
No log	11.0	33	0.8072	0.2965	0.8075	0.8986
No log	12.0	36	1.0081	0.1113	1.0086	1.0043
No log	13.0	39	1.0223	0.3043	1.0228	1.0113
No log	14.0	42	0.8368	0.4208	0.8372	0.9150
No log	15.0	45	0.7773	0.4574	0.7777	0.8819
No log	16.0	48	0.6592	0.4912	0.6596	0.8121
No log	17.0	51	0.6685	0.4925	0.6690	0.8179
No log	18.0	54	0.6702	0.3937	0.6705	0.8188
No log	19.0	57	0.6111	0.4502	0.6115	0.7820
No log	20.0	60	0.6401	0.3683	0.6405	0.8003
No log	21.0	63	0.6395	0.4647	0.6401	0.8001
No log	22.0	66	0.7136	0.4988	0.7139	0.8449
No log	23.0	69	0.7137	0.5064	0.7142	0.8451
No log	24.0	72	0.7566	0.5069	0.7567	0.8699
No log	25.0	75	0.7273	0.5093	0.7278	0.8531
No log	26.0	78	0.8735	0.4603	0.8733	0.9345
No log	27.0	81	0.7327	0.4814	0.7331	0.8562
No log	28.0	84	0.8657	0.4584	0.8658	0.9305
No log	29.0	87	0.7677	0.5013	0.7680	0.8764
No log	30.0	90	0.7631	0.4900	0.7634	0.8737
No log	31.0	93	0.6823	0.5166	0.6826	0.8262
No log	32.0	96	0.6862	0.5197	0.6864	0.8285
No log	33.0	99	0.7238	0.5182	0.7240	0.8509
No log	34.0	102	0.7817	0.4983	0.7817	0.8842
No log	35.0	105	0.7330	0.4953	0.7331	0.8562
No log	36.0	108	0.7896	0.4811	0.7895	0.8886
No log	37.0	111	0.6777	0.5208	0.6780	0.8234
No log	38.0	114	0.7551	0.5032	0.7550	0.8689
No log	39.0	117	0.7879	0.4853	0.7878	0.8876
No log	40.0	120	0.7302	0.5275	0.7302	0.8545
No log	41.0	123	0.6907	0.5153	0.6907	0.8311
No log	42.0	126	0.6750	0.5430	0.6749	0.8215
No log	43.0	129	0.7290	0.5121	0.7289	0.8537
No log	44.0	132	0.7835	0.5055	0.7833	0.8851
No log	45.0	135	0.9404	0.4895	0.9400	0.9696
No log	46.0	138	0.8185	0.5060	0.8182	0.9046
No log	47.0	141	0.9768	0.4525	0.9764	0.9881
No log	48.0	144	0.7059	0.5123	0.7058	0.8401
No log	49.0	147	0.9336	0.4707	0.9331	0.9660
No log	50.0	150	0.7547	0.5421	0.7545	0.8686
No log	51.0	153	0.6534	0.5480	0.6535	0.8084
No log	52.0	156	0.9622	0.4757	0.9616	0.9806
No log	53.0	159	1.0000	0.4671	0.9995	0.9998
No log	54.0	162	0.7267	0.5395	0.7266	0.8524
No log	55.0	165	0.9276	0.4855	0.9273	0.9629
No log	56.0	168	0.7869	0.5098	0.7867	0.8870
No log	57.0	171	0.6600	0.5366	0.6600	0.8124
No log	58.0	174	0.7200	0.5350	0.7198	0.8484
No log	59.0	177	0.6609	0.5441	0.6609	0.8129
No log	60.0	180	0.9045	0.4971	0.9041	0.9508
No log	61.0	183	0.8774	0.5042	0.8770	0.9365
No log	62.0	186	0.7287	0.5357	0.7286	0.8536
No log	63.0	189	0.8217	0.5161	0.8213	0.9063
No log	64.0	192	0.8331	0.5067	0.8328	0.9126
No log	65.0	195	0.6604	0.5471	0.6604	0.8127
No log	66.0	198	0.7764	0.5161	0.7762	0.8810
No log	67.0	201	0.8033	0.4866	0.8031	0.8962
No log	68.0	204	0.7744	0.5110	0.7743	0.8799
No log	69.0	207	0.8497	0.4793	0.8495	0.9217
No log	70.0	210	0.7067	0.5315	0.7066	0.8406
No log	71.0	213	0.7558	0.5156	0.7557	0.8693

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.1
• Tokenizers 0.21.0