bert-base-turkish-cased-191123-stopword-product-names-classification

This model is a fine-tuned version of dbmdz/bert-base-turkish-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4409
• Accuracy: 0.9347

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	314	2.3636	0.6008
3.1969	2.0	628	1.1851	0.7984
3.1969	3.0	942	0.8059	0.8598
1.0529	4.0	1256	0.6367	0.8733
0.509	5.0	1570	0.5475	0.8813
0.509	6.0	1884	0.4913	0.8988
0.3084	7.0	2198	0.4796	0.9036
0.2041	8.0	2512	0.4450	0.9155
0.2041	9.0	2826	0.4339	0.9131
0.1229	10.0	3140	0.4227	0.9267
0.1229	11.0	3454	0.4097	0.9243
0.0942	12.0	3768	0.4232	0.9211
0.0619	13.0	4082	0.4164	0.9259
0.0619	14.0	4396	0.4191	0.9283
0.0442	15.0	4710	0.4479	0.9227
0.0336	16.0	5024	0.4521	0.9251
0.0336	17.0	5338	0.4484	0.9283
0.0247	18.0	5652	0.4625	0.9259
0.0247	19.0	5966	0.4409	0.9347
0.0177	20.0	6280	0.4560	0.9299
0.0165	21.0	6594	0.4425	0.9307
0.0165	22.0	6908	0.4435	0.9323
0.01	23.0	7222	0.4398	0.9347
0.0097	24.0	7536	0.4531	0.9331
0.0097	25.0	7850	0.4610	0.9291
0.0071	26.0	8164	0.4595	0.9315
0.0071	27.0	8478	0.4567	0.9323
0.0053	28.0	8792	0.4599	0.9339
0.0047	29.0	9106	0.4626	0.9323
0.0047	30.0	9420	0.4632	0.9315

Framework versions

• Transformers 4.33.0
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.13.3