frame_classification_longformer_earlystopping_2

This model is a fine-tuned version of allenai/longformer-base-4096 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5698
• Accuracy: 0.9317
• F1: 0.9601
• Precision: 0.9346
• Recall: 0.9869

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.8651	1.0	1288	1.1520	0.8323	0.9085	0.8323	1.0
0.7998	2.0	2576	0.6925	0.9130	0.9495	0.9181	0.9832
0.6819	3.0	3864	0.4349	0.9255	0.9560	0.9404	0.9720
0.7033	4.0	5152	0.5836	0.9224	0.9547	0.9278	0.9832
0.757	5.0	6440	0.5795	0.9224	0.9544	0.9339	0.9757
0.6963	6.0	7728	0.5488	0.9317	0.9599	0.9393	0.9813
0.6656	7.0	9016	0.6232	0.9255	0.9564	0.9311	0.9832
0.6957	8.0	10304	0.6441	0.9255	0.9564	0.9311	0.9832
0.6852	9.0	11592	0.6009	0.9224	0.9548	0.9263	0.9851
0.6846	10.0	12880	0.5947	0.9255	0.9564	0.9311	0.9832
0.728	11.0	14168	0.5873	0.9224	0.9550	0.9219	0.9907
0.6456	12.0	15456	0.5781	0.9332	0.9609	0.9363	0.9869
0.662	13.0	16744	0.5128	0.9301	0.9588	0.9408	0.9776
0.6017	14.0	18032	0.6430	0.9177	0.9514	0.9351	0.9683
0.7525	15.0	19320	0.5631	0.9193	0.9522	0.9384	0.9664
0.6802	16.0	20608	0.5949	0.9224	0.9544	0.9339	0.9757
0.6489	17.0	21896	0.6540	0.9022	0.9416	0.9355	0.9478
0.707	18.0	23184	0.5921	0.9239	0.9554	0.9325	0.9795
0.7158	19.0	24472	0.6170	0.9255	0.9564	0.9311	0.9832
0.6264	20.0	25760	0.5303	0.9348	0.9617	0.9395	0.9851
0.6667	21.0	27048	0.6288	0.9255	0.9566	0.9281	0.9869
0.6648	22.0	28336	0.6579	0.9208	0.9540	0.9232	0.9869
0.6204	23.0	29624	0.5716	0.9301	0.9592	0.9330	0.9869
0.6693	24.0	30912	0.6138	0.9270	0.9575	0.9282	0.9888
0.6555	25.0	32200	0.6369	0.9255	0.9566	0.9281	0.9869
0.6446	26.0	33488	0.5609	0.9301	0.9591	0.9345	0.9851
0.6675	27.0	34776	0.5622	0.9301	0.9591	0.9361	0.9832
0.5946	28.0	36064	0.5740	0.9301	0.9591	0.9345	0.9851
0.5707	29.0	37352	0.5661	0.9317	0.9601	0.9346	0.9869
0.6703	30.0	38640	0.5698	0.9317	0.9601	0.9346	0.9869

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2