distilbert-classn-LinearAlg-finetuned-span-width

This model is a fine-tuned version of dslim/distilbert-NER on the None dataset. It achieves the following results on the evaluation set:

Model description

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 1e-05
train_batch_size: 2
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 8
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
lr_scheduler_warmup_steps: 500
num_epochs: 25
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
9.5553	1.3562	50	2.4410	0.1111	0.0988	0.0991	0.1111
9.5082	2.7123	100	2.4145	0.1032	0.0946	0.1005	0.1032
9.2844	4.0548	150	2.3758	0.1429	0.1286	0.1283	0.1429
9.1664	5.4110	200	2.3287	0.1587	0.1491	0.1781	0.1587
8.8251	6.7671	250	2.2563	0.2460	0.2491	0.3626	0.2460
8.3414	8.1096	300	2.1525	0.3095	0.3109	0.3506	0.3095
7.6498	9.4658	350	2.0206	0.4048	0.4014	0.4357	0.4048
6.7787	10.8219	400	1.8072	0.5159	0.4967	0.5453	0.5159
5.5972	12.1644	450	1.5651	0.5952	0.5807	0.6248	0.5952
4.3206	13.5205	500	1.3148	0.6905	0.6886	0.7173	0.6905
3.1237	14.8767	550	1.1469	0.7063	0.7126	0.7461	0.7063
2.1003	16.2192	600	0.9970	0.7381	0.7431	0.7783	0.7381
1.5158	17.5753	650	0.9129	0.7698	0.7743	0.7961	0.7698
1.0663	18.9315	700	0.8501	0.7778	0.7833	0.8084	0.7778
0.7797	20.2740	750	0.7928	0.7698	0.7726	0.7927	0.7698
0.58	21.6301	800	0.7950	0.7619	0.7649	0.7765	0.7619
0.5137	22.9863	850	0.8031	0.7778	0.7796	0.7905	0.7778
0.4405	24.3288	900	0.8048	0.7619	0.7638	0.7732	0.7619