BERT_V8_sp10_lw40_ex100_lo50_k7_k7_fold4

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: 0.5177
• Qwk: 0.5782
• Mse: 0.5177
• Rmse: 0.7195

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 OptimizerNames.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	4	7.1697	0.0	7.1697	2.6776
No log	2.0	8	5.3751	0.0090	5.3751	2.3184
No log	3.0	12	3.7582	0.0040	3.7582	1.9386
No log	4.0	16	2.2137	0.1420	2.2137	1.4879
No log	5.0	20	1.3653	0.0107	1.3653	1.1685
No log	6.0	24	1.0163	0.0107	1.0163	1.0081
No log	7.0	28	0.8223	0.2236	0.8223	0.9068
No log	8.0	32	0.6480	0.3986	0.6480	0.8050
No log	9.0	36	0.7958	0.4648	0.7958	0.8921
No log	10.0	40	0.5216	0.5131	0.5216	0.7222
No log	11.0	44	1.0154	0.4034	1.0154	1.0077
No log	12.0	48	0.5674	0.5171	0.5674	0.7532
No log	13.0	52	1.1029	0.4372	1.1029	1.0502
No log	14.0	56	0.6530	0.5470	0.6530	0.8081
No log	15.0	60	0.5819	0.5441	0.5819	0.7628
No log	16.0	64	0.7779	0.5024	0.7779	0.8820
No log	17.0	68	0.5471	0.5878	0.5471	0.7397
No log	18.0	72	0.5176	0.6038	0.5176	0.7195
No log	19.0	76	0.6601	0.5165	0.6601	0.8125
No log	20.0	80	0.5787	0.6124	0.5787	0.7607
No log	21.0	84	0.5214	0.6173	0.5214	0.7221
No log	22.0	88	0.5008	0.6244	0.5008	0.7076
No log	23.0	92	0.5051	0.6270	0.5051	0.7107
No log	24.0	96	0.5532	0.5571	0.5532	0.7438
No log	25.0	100	0.6777	0.5217	0.6777	0.8232
No log	26.0	104	0.5681	0.6268	0.5681	0.7537
No log	27.0	108	0.5558	0.5669	0.5558	0.7455
No log	28.0	112	0.5692	0.5637	0.5692	0.7544
No log	29.0	116	0.5192	0.5930	0.5192	0.7206
No log	30.0	120	0.5157	0.6063	0.5157	0.7181
No log	31.0	124	0.5222	0.5851	0.5222	0.7226
No log	32.0	128	0.5304	0.5964	0.5304	0.7283
No log	33.0	132	0.5591	0.5907	0.5591	0.7477
No log	34.0	136	0.5311	0.5845	0.5311	0.7287
No log	35.0	140	0.5083	0.5820	0.5083	0.7129
No log	36.0	144	0.5273	0.5753	0.5273	0.7261
No log	37.0	148	0.5238	0.6247	0.5238	0.7237
No log	38.0	152	0.5506	0.5770	0.5506	0.7421
No log	39.0	156	0.5044	0.5981	0.5044	0.7102
No log	40.0	160	0.5823	0.5518	0.5823	0.7631
No log	41.0	164	0.5361	0.6167	0.5361	0.7322
No log	42.0	168	0.5310	0.6142	0.5310	0.7287
No log	43.0	172	0.5645	0.5635	0.5645	0.7513
No log	44.0	176	0.5060	0.6033	0.5060	0.7113
No log	45.0	180	0.5616	0.5474	0.5616	0.7494
No log	46.0	184	0.5153	0.5990	0.5153	0.7178
No log	47.0	188	0.5607	0.5689	0.5607	0.7488
No log	48.0	192	0.5215	0.5977	0.5215	0.7221
No log	49.0	196	0.5649	0.5657	0.5649	0.7516
No log	50.0	200	0.5058	0.5880	0.5058	0.7112
No log	51.0	204	0.5078	0.5969	0.5078	0.7126
No log	52.0	208	0.5540	0.5582	0.5540	0.7443
No log	53.0	212	0.5177	0.5782	0.5177	0.7195

Framework versions

• Transformers 4.51.1
• Pytorch 2.5.1+cu124
• Datasets 3.5.0
• Tokenizers 0.21.0