distilbert-cased-lft

This model is a fine-tuned version of distilbert/distilbert-base-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1017
• Precision: 0.8722
• Recall: 0.8905
• F1: 0.8813
• Accuracy: 0.9764

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.4065	100	0.1106	0.8374	0.7969	0.8167	0.9647
No log	0.8130	200	0.0926	0.8242	0.8474	0.8356	0.9690
No log	1.2195	300	0.0898	0.8325	0.8671	0.8494	0.9704
No log	1.6260	400	0.0873	0.8591	0.8614	0.8602	0.9729
0.0943	2.0325	500	0.0829	0.8563	0.8765	0.8662	0.9740
0.0943	2.4390	600	0.0864	0.8656	0.8747	0.8701	0.9746
0.0943	2.8455	700	0.0842	0.8652	0.8761	0.8706	0.9746
0.0943	3.2520	800	0.0875	0.8627	0.8823	0.8724	0.9746
0.0943	3.6585	900	0.0887	0.8564	0.8829	0.8694	0.9744
0.0444	4.0650	1000	0.0875	0.8801	0.8797	0.8799	0.9763
0.0444	4.4715	1100	0.0944	0.8516	0.8901	0.8704	0.9746
0.0444	4.8780	1200	0.0906	0.8607	0.8891	0.8746	0.9752
0.0444	5.2846	1300	0.0934	0.8706	0.8896	0.8800	0.9765
0.0444	5.6911	1400	0.0914	0.8784	0.8862	0.8823	0.9765
0.0248	6.0976	1500	0.0918	0.8796	0.8896	0.8846	0.9772
0.0248	6.5041	1600	0.0960	0.8711	0.8916	0.8812	0.9765
0.0248	6.9106	1700	0.0970	0.8678	0.8876	0.8776	0.9763
0.0248	7.3171	1800	0.1008	0.8690	0.8887	0.8787	0.9759
0.0248	7.7236	1900	0.1012	0.8650	0.8926	0.8786	0.9759
0.0153	8.1301	2000	0.1002	0.8715	0.8921	0.8817	0.9762
0.0153	8.5366	2100	0.1003	0.8749	0.8889	0.8818	0.9763
0.0153	8.9431	2200	0.1015	0.8680	0.8917	0.8797	0.9760
0.0153	9.3496	2300	0.1015	0.8716	0.8882	0.8798	0.9764
0.0153	9.7561	2400	0.1017	0.8722	0.8905	0.8813	0.9764

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1