jackmedda/google-bigbird-roberta-base_finetuned_augmented_augmented_deepseek

This model is a fine-tuned version of google/bigbird-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3040
• Accuracy: 0.7647
• F1: 0.8571
• Precision: 0.8
• Recall: 0.9231

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.4568	1.0	46	0.7344	0.7	0.8235	0.7	1.0
0.6053	2.0	92	0.6889	0.7	0.8235	0.7	1.0
0.5004	3.0	138	0.6289	0.7	0.8235	0.7	1.0
0.6076	4.0	184	0.6493	0.7	0.8235	0.7	1.0
0.3798	5.0	230	0.7945	0.7	0.8235	0.7	1.0
0.2401	6.0	276	1.2776	0.7	0.8235	0.7	1.0
0.4076	7.0	322	1.1325	0.7	0.8235	0.7	1.0
0.1193	8.0	368	1.1424	0.7	0.8235	0.7	1.0
0.3332	9.0	414	0.9214	0.8	0.875	0.7778	1.0
0.1445	10.0	460	0.2424	0.9	0.9333	0.875	1.0
0.0029	11.0	506	1.6181	0.7	0.8235	0.7	1.0
0.1174	12.0	552	0.0031	1.0	1.0	1.0	1.0
0.0014	13.0	598	1.2314	0.8	0.875	0.7778	1.0
0.0011	14.0	644	1.1105	0.8	0.875	0.7778	1.0
0.0008	15.0	690	1.2286	0.8	0.875	0.7778	1.0
0.0008	16.0	736	1.2704	0.8	0.875	0.7778	1.0
0.0006	17.0	782	1.3157	0.8	0.875	0.7778	1.0
0.0005	18.0	828	1.3290	0.8	0.875	0.7778	1.0
0.0005	19.0	874	1.3752	0.8	0.875	0.7778	1.0
0.0004	20.0	920	1.3951	0.8	0.875	0.7778	1.0

Framework versions

• Transformers 4.48.3
• Pytorch 2.3.0+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0