Version5ASAP_FineTuningBERT_AugV12_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: 1.5932
• Qwk: 0.4596
• Mse: 1.5927
• Rmse: 1.2620

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	1	12.0362	0.0	12.0334	3.4689
No log	2.0	2	10.6754	0.0077	10.6727	3.2669
No log	3.0	3	9.4378	0.0	9.4352	3.0717
No log	4.0	4	8.1517	0.0	8.1491	2.8547
No log	5.0	5	7.3227	0.0	7.3202	2.7056
No log	6.0	6	6.4618	0.0	6.4594	2.5415
No log	7.0	7	5.5808	0.0541	5.5787	2.3619
No log	8.0	8	5.0813	0.0131	5.0792	2.2537
No log	9.0	9	4.8201	-0.0038	4.8180	2.1950
No log	10.0	10	4.2233	0.0	4.2214	2.0546
No log	11.0	11	3.7775	0.0	3.7756	1.9431
No log	12.0	12	3.3785	0.0	3.3766	1.8375
No log	13.0	13	3.1868	0.0	3.1848	1.7846
No log	14.0	14	2.9126	0.0	2.9107	1.7061
No log	15.0	15	2.5109	0.0	2.5091	1.5840
No log	16.0	16	2.3052	0.1209	2.3035	1.5177
No log	17.0	17	2.0089	0.1068	2.0072	1.4168
No log	18.0	18	1.9206	0.0359	1.9189	1.3852
No log	19.0	19	1.9374	-0.0189	1.9357	1.3913
No log	20.0	20	1.6355	0.0	1.6339	1.2782
No log	21.0	21	1.3152	0.0	1.3137	1.1462
No log	22.0	22	1.2093	0.0	1.2078	1.0990
No log	23.0	23	1.1210	0.0	1.1197	1.0581
No log	24.0	24	1.1343	0.0	1.1329	1.0644
No log	25.0	25	1.0416	0.0	1.0402	1.0199
No log	26.0	26	0.9135	0.1348	0.9123	0.9551
No log	27.0	27	0.8540	0.3340	0.8528	0.9235
No log	28.0	28	0.8547	0.2314	0.8535	0.9239
No log	29.0	29	0.8636	0.1865	0.8625	0.9287
No log	30.0	30	0.8326	0.1951	0.8315	0.9119
No log	31.0	31	0.8926	0.1608	0.8915	0.9442
No log	32.0	32	0.9123	0.1487	0.9113	0.9546
No log	33.0	33	0.8308	0.1782	0.8298	0.9109
No log	34.0	34	0.8417	0.1661	0.8408	0.9169
No log	35.0	35	0.7955	0.1834	0.7945	0.8914
No log	36.0	36	0.8381	0.1834	0.8372	0.9150
No log	37.0	37	0.8884	0.1834	0.8875	0.9421
No log	38.0	38	0.7824	0.1970	0.7816	0.8841
No log	39.0	39	0.7569	0.2394	0.7560	0.8695
No log	40.0	40	0.9385	0.2110	0.9378	0.9684
No log	41.0	41	0.8765	0.2992	0.8758	0.9358
No log	42.0	42	0.6547	0.5357	0.6540	0.8087
No log	43.0	43	0.6471	0.5642	0.6464	0.8040
No log	44.0	44	0.8809	0.4351	0.8802	0.9382
No log	45.0	45	0.8588	0.4811	0.8581	0.9263
No log	46.0	46	0.6181	0.6149	0.6175	0.7858
No log	47.0	47	0.6101	0.6193	0.6095	0.7807
No log	48.0	48	0.7119	0.5900	0.7113	0.8434
No log	49.0	49	0.6437	0.6326	0.6431	0.8019
No log	50.0	50	0.5538	0.6668	0.5532	0.7438
No log	51.0	51	0.6654	0.6357	0.6648	0.8153
No log	52.0	52	0.6373	0.6499	0.6367	0.7979
No log	53.0	53	0.8067	0.5789	0.8060	0.8978
No log	54.0	54	0.7786	0.6019	0.7779	0.8820
No log	55.0	55	0.5907	0.6701	0.5900	0.7681
No log	56.0	56	0.6136	0.6621	0.6130	0.7829
No log	57.0	57	0.8337	0.5967	0.8331	0.9127
No log	58.0	58	1.0219	0.5567	1.0212	1.0106
No log	59.0	59	0.9518	0.5694	0.9512	0.9753
No log	60.0	60	0.9126	0.5837	0.9119	0.9549
No log	61.0	61	0.7433	0.6427	0.7426	0.8617
No log	62.0	62	0.8063	0.6430	0.8056	0.8976
No log	63.0	63	1.0811	0.5743	1.0806	1.0395
No log	64.0	64	1.1731	0.5555	1.1727	1.0829
No log	65.0	65	1.0231	0.5793	1.0225	1.0112
No log	66.0	66	0.9954	0.5884	0.9948	0.9974
No log	67.0	67	1.1531	0.5610	1.1526	1.0736
No log	68.0	68	1.2649	0.5192	1.2645	1.1245
No log	69.0	69	1.2054	0.5372	1.2050	1.0977
No log	70.0	70	1.2789	0.5043	1.2784	1.1307
No log	71.0	71	1.1868	0.5378	1.1863	1.0892
No log	72.0	72	1.2911	0.5014	1.2906	1.1361
No log	73.0	73	1.5075	0.4661	1.5070	1.2276
No log	74.0	74	1.6494	0.4370	1.6488	1.2841
No log	75.0	75	1.5932	0.4596	1.5927	1.2620

Framework versions

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