no_vague_no_downsample

This model is a fine-tuned version of Mardiyyah/cellate2.0-tapt_base-LR_5e-05 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0743
• Precision: 0.7128
• Recall: 0.7825
• F1: 0.7460
• Accuracy: 0.9815

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: 32
• eval_batch_size: 16
• seed: 3407
• 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: 0.01
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.7697	0.4950	100	0.1502	0.3079	0.2172	0.2547	0.9608
0.1727	0.9901	200	0.1198	0.4065	0.6694	0.5058	0.9620
0.1057	1.4851	300	0.0818	0.7075	0.6856	0.6964	0.9804
0.0753	1.9802	400	0.0765	0.7167	0.7244	0.7205	0.9807
0.0555	2.4752	500	0.1019	0.3659	0.8505	0.5117	0.9471
0.0511	2.9703	600	0.0741	0.7128	0.7825	0.7460	0.9815
0.0381	3.4653	700	0.0898	0.7111	0.7458	0.7280	0.9811
0.0369	3.9604	800	0.0846	0.7078	0.7804	0.7423	0.9818
0.0295	4.4554	900	0.0919	0.6923	0.7723	0.7301	0.9809

Framework versions

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