CeLLaTe-ner-2class-pubmedbert-tapt-combData-tokenizer-adapted-spanmask-gazetters-lr_2.085

This model is a fine-tuned version of Mardiyyah/CeLLaTe-tapt-pubmedbert-tokenizer-adapted-spanmask-combinedData on the OTAR3088/CeLLaTe-ner-2class-iob_final dataset. It achieves the following results on the evaluation set:

• Loss: 0.1298
• Precision: 0.7765
• Recall: 0.7565
• Micro F1: 0.7664
• Weighted F1: 0.7666
• Macro F1: 0.7751
• Accuracy: 0.9834

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: 2.0854538798e-05
• train_batch_size: 32
• eval_batch_size: 16
• seed: 3407
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• 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_ratio: 0.03
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	Micro F1	Weighted F1	Macro F1	Accuracy
0.2292	1.0	526	0.0584	0.7113	0.7246	0.7179	0.7179	0.7260	0.9820
0.0214	2.0	1052	0.0707	0.7297	0.7288	0.7292	0.7303	0.7454	0.9810
0.0098	3.0	1578	0.0807	0.7614	0.7408	0.7510	0.7513	0.7623	0.9829
0.0059	4.0	2104	0.0893	0.7757	0.7300	0.7522	0.7523	0.7564	0.9837
0.0034	5.0	2630	0.0959	0.7383	0.7462	0.7422	0.7427	0.7530	0.9821
0.0026	6.0	3156	0.0918	0.7337	0.7655	0.7493	0.7500	0.7612	0.9828
0.0017	7.0	3682	0.1079	0.7884	0.7282	0.7571	0.7571	0.7673	0.9830
0.0015	8.0	4208	0.1114	0.7600	0.7523	0.7561	0.7564	0.7643	0.9824
0.0013	9.0	4734	0.1166	0.7723	0.7300	0.7505	0.7504	0.7604	0.9825
0.0012	10.0	5260	0.1188	0.7682	0.7474	0.7577	0.7581	0.7684	0.9829
0.0009	11.0	5786	0.1239	0.7750	0.7354	0.7547	0.7551	0.7664	0.9830
0.0007	12.0	6312	0.1243	0.7693	0.7438	0.7563	0.7566	0.7644	0.9828
0.0007	13.0	6838	0.1241	0.7630	0.7571	0.7600	0.7603	0.7675	0.9832
0.0007	14.0	7364	0.1186	0.7677	0.7492	0.7584	0.7585	0.7667	0.9834
0.0006	15.0	7890	0.1225	0.7829	0.7438	0.7629	0.7632	0.7714	0.9840
0.0005	16.0	8416	0.1230	0.7648	0.7547	0.7597	0.7601	0.7693	0.9834
0.0004	17.0	8942	0.1321	0.7660	0.7577	0.7618	0.7621	0.7701	0.9835
0.0005	18.0	9468	0.1301	0.7668	0.7535	0.7601	0.7603	0.7675	0.9834
0.0004	19.0	9994	0.1292	0.7765	0.7565	0.7664	0.7666	0.7751	0.9834
0.0004	20.0	10520	0.1295	0.7682	0.7571	0.7626	0.7628	0.7714	0.9834

Framework versions

• Transformers 4.48.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.2
• Tokenizers 0.21.0