relevance-classification-v1

This model is a fine-tuned version of allenai/longformer-base-4096 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 5.2284
• Accuracy: 0.6552

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	338	0.6800	0.5448
0.7174	2.0	676	1.4472	0.6276
0.865	3.0	1014	1.2742	0.6621
0.865	4.0	1352	1.4262	0.6621
0.5753	5.0	1690	2.1018	0.6414
0.335	6.0	2028	2.4029	0.6345
0.335	7.0	2366	1.9533	0.6483
0.2503	8.0	2704	2.4815	0.6138
0.1785	9.0	3042	2.5177	0.6897
0.1785	10.0	3380	2.5533	0.6552
0.1067	11.0	3718	2.9023	0.6552
0.0957	12.0	4056	3.2890	0.6345
0.0957	13.0	4394	3.5851	0.6138
0.0166	14.0	4732	3.6766	0.5931
0.1395	15.0	5070	3.6210	0.6069
0.1395	16.0	5408	3.2261	0.6414
0.1005	17.0	5746	3.2913	0.6414
0.0793	18.0	6084	3.6091	0.6207
0.0793	19.0	6422	2.4907	0.6897
0.13	20.0	6760	3.0017	0.6552
0.0467	21.0	7098	3.1797	0.6759
0.0467	22.0	7436	3.4537	0.6414
0.0875	23.0	7774	3.1266	0.6414
0.0677	24.0	8112	3.4799	0.6759
0.0677	25.0	8450	3.3836	0.6690
0.0892	26.0	8788	3.1044	0.6483
0.1089	27.0	9126	3.5136	0.6552
0.1089	28.0	9464	3.3848	0.6483
0.0586	29.0	9802	3.5435	0.6621
0.043	30.0	10140	3.6754	0.6414
0.043	31.0	10478	3.8983	0.6483
0.0026	32.0	10816	3.8528	0.6414
0.0195	33.0	11154	3.9876	0.6483
0.0195	34.0	11492	2.9999	0.6414
0.0781	35.0	11830	3.7963	0.6207
0.0552	36.0	12168	4.2694	0.6138
0.0	37.0	12506	4.3729	0.6138
0.0	38.0	12844	4.4702	0.6138
0.0	39.0	13182	4.5190	0.6138
0.0125	40.0	13520	4.2951	0.6483
0.0125	41.0	13858	3.9059	0.6276
0.0709	42.0	14196	3.4919	0.6621
0.0362	43.0	14534	4.0863	0.6276
0.0362	44.0	14872	3.9934	0.6276
0.0311	45.0	15210	4.3174	0.6207
0.0163	46.0	15548	4.3117	0.6138
0.0163	47.0	15886	4.2067	0.6414
0.0235	48.0	16224	3.2403	0.6483
0.0512	49.0	16562	3.6099	0.6621
0.0512	50.0	16900	3.9438	0.6345
0.0002	51.0	17238	4.0551	0.6345
0.0	52.0	17576	4.1505	0.6345
0.0	53.0	17914	4.2107	0.6345
0.0	54.0	18252	4.1841	0.5931
0.0493	55.0	18590	4.4524	0.6207
0.0493	56.0	18928	4.3673	0.6276
0.0172	57.0	19266	4.4991	0.6345
0.0002	58.0	19604	4.7284	0.6138
0.0002	59.0	19942	4.7207	0.6276
0.0004	60.0	20280	4.8372	0.6276
0.0132	61.0	20618	5.0463	0.6138
0.0132	62.0	20956	4.0695	0.6483
0.0294	63.0	21294	4.4791	0.6276
0.0234	64.0	21632	4.0409	0.6759
0.0234	65.0	21970	4.3323	0.6276
0.0311	66.0	22308	4.5133	0.6345
0.0069	67.0	22646	4.1708	0.6690
0.0069	68.0	22984	4.7436	0.6276
0.0001	69.0	23322	4.8199	0.6276
0.0011	70.0	23660	5.2157	0.5862
0.0011	71.0	23998	5.0111	0.6069
0.0279	72.0	24336	4.7120	0.6621
0.0	73.0	24674	4.8631	0.6207
0.0117	74.0	25012	4.9149	0.6276
0.0117	75.0	25350	4.9518	0.6276
0.0	76.0	25688	4.9781	0.6276
0.0	77.0	26026	5.0057	0.6345
0.0	78.0	26364	5.0409	0.6345
0.0	79.0	26702	5.0909	0.6345
0.0119	80.0	27040	4.4556	0.6552
0.0119	81.0	27378	4.5697	0.6621
0.0	82.0	27716	4.8371	0.6483
0.0	83.0	28054	4.8793	0.6483
0.0	84.0	28392	4.9278	0.6414
0.0	85.0	28730	4.9605	0.6414
0.0	86.0	29068	5.2864	0.6207
0.0	87.0	29406	5.3216	0.6207
0.0	88.0	29744	5.3452	0.6207
0.0	89.0	30082	5.5673	0.6069
0.0	90.0	30420	5.3842	0.6276
0.0	91.0	30758	5.3997	0.6276
0.0	92.0	31096	5.4139	0.6276
0.0	93.0	31434	5.4287	0.6276
0.0	94.0	31772	5.4433	0.6345
0.0	95.0	32110	5.1979	0.6552
0.0	96.0	32448	5.2034	0.6552
0.0001	97.0	32786	5.2129	0.6552
0.0	98.0	33124	5.2220	0.6552
0.0	99.0	33462	5.2267	0.6552
0.0	100.0	33800	5.2284	0.6552

Framework versions

• Transformers 4.36.2
• Pytorch 2.2.2+cu121
• Datasets 2.16.0
• Tokenizers 0.15.2