Products_NER2

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

• Loss: 0.3482
• Precision: 0.9155
• Recall: 0.9208
• F1: 0.9182
• Accuracy: 0.9400

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: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.2267	1.0	2470	0.1614	0.8379	0.8791	0.8580	0.9212
0.1363	2.0	4940	0.1230	0.8602	0.8968	0.8781	0.9332
0.1047	3.0	7410	0.1183	0.8808	0.9063	0.8934	0.9360
0.0931	4.0	9880	0.1139	0.8909	0.9119	0.9013	0.9387
0.085	5.0	12350	0.1153	0.8889	0.9110	0.8998	0.9390
0.0835	6.0	14820	0.1257	0.9043	0.9165	0.9104	0.9398
0.0728	7.0	17290	0.1218	0.8987	0.9149	0.9067	0.9393
0.069	8.0	19760	0.1457	0.9040	0.9154	0.9097	0.9389
0.0616	9.0	22230	0.1606	0.9090	0.9166	0.9128	0.9386
0.0559	10.0	24700	0.1726	0.9122	0.9189	0.9156	0.9397
0.0504	11.0	27170	0.1998	0.9131	0.9192	0.9161	0.9396
0.043	12.0	29640	0.2015	0.9126	0.9194	0.9160	0.9402
0.0389	13.0	32110	0.2388	0.9129	0.9195	0.9162	0.9394
0.035	14.0	34580	0.2569	0.9135	0.9202	0.9169	0.9397
0.0311	15.0	37050	0.2718	0.9156	0.9207	0.9181	0.9400
0.028	16.0	39520	0.2886	0.9158	0.9208	0.9183	0.9403
0.0246	17.0	41990	0.3054	0.9145	0.9201	0.9173	0.9392
0.0212	18.0	44460	0.3252	0.9155	0.9206	0.9180	0.9398
0.0192	19.0	46930	0.3333	0.9157	0.9210	0.9183	0.9402
0.017	20.0	49400	0.3482	0.9155	0.9208	0.9182	0.9400

Framework versions

• Transformers 4.33.2
• Pytorch 1.13.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3