cellate-tapt_freeze_llrd_ww_mask-LR_2e-05

This model is a fine-tuned version of microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.3251
• Accuracy: 0.7222

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: 16
• eval_batch_size: 16
• seed: 3407
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3722	1.0	6	1.3843	0.7176
1.3817	2.0	12	1.3727	0.7164
1.3559	3.0	18	1.4378	0.7126
1.3708	4.0	24	1.3784	0.7197
1.3981	5.0	30	1.4502	0.7055
1.3353	6.0	36	1.4600	0.7063
1.3235	7.0	42	1.4827	0.7000
1.323	8.0	48	1.3655	0.7285
1.3173	9.0	54	1.3318	0.7340
1.3163	10.0	60	1.4321	0.7118
1.3113	11.0	66	1.3115	0.7256
1.3202	12.0	72	1.4071	0.7101
1.3057	13.0	78	1.2777	0.7319
1.3789	14.0	84	1.3215	0.7256
1.3064	15.0	90	1.3583	0.7059
1.3159	16.0	96	1.3715	0.7264
1.412	16.7273	100	1.3251	0.7222

Framework versions

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