final_best

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

• Loss: 0.2002
• Accuracy: 0.9346
• Precision: 0.9344
• Recall: 0.9346
• F1: 0.9344

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: 7e-06
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.1126	0.1255	100	0.9980	0.5913	0.6925	0.5913	0.5476
0.7959	0.2509	200	0.5448	0.8411	0.8541	0.8411	0.8412
0.4298	0.3764	300	0.3201	0.9016	0.9043	0.9016	0.9023
0.3292	0.5019	400	0.2677	0.9141	0.9155	0.9141	0.9145
0.2899	0.6274	500	0.2556	0.9157	0.9177	0.9157	0.9159
0.277	0.7528	600	0.2352	0.9224	0.9235	0.9224	0.9227
0.2546	0.8783	700	0.2221	0.9264	0.9268	0.9264	0.9264
0.2507	1.0038	800	0.2176	0.9283	0.9283	0.9283	0.9283
0.2491	1.1292	900	0.2242	0.9267	0.9277	0.9267	0.9265
0.2378	1.2547	1000	0.2178	0.9267	0.9274	0.9267	0.9266
0.2216	1.3802	1100	0.2217	0.9261	0.9267	0.9261	0.9259
0.2222	1.5056	1200	0.2163	0.93	0.9300	0.93	0.9299
0.2467	1.6311	1300	0.2192	0.929	0.9295	0.929	0.9288
0.2433	1.7566	1400	0.2170	0.9267	0.9278	0.9267	0.9263
0.2413	1.8821	1500	0.2113	0.9302	0.9306	0.9302	0.9300
0.2184	2.0075	1600	0.2146	0.9303	0.9306	0.9303	0.9302
0.2204	2.1330	1700	0.2078	0.9289	0.9292	0.9289	0.9287
0.2321	2.2585	1800	0.2068	0.9309	0.9309	0.9309	0.9306
0.1983	2.3839	1900	0.2105	0.9311	0.9310	0.9311	0.9309
0.2048	2.5094	2000	0.2048	0.9316	0.9314	0.9316	0.9313
0.2104	2.6349	2100	0.2085	0.9321	0.9320	0.9321	0.9319
0.229	2.7604	2200	0.2002	0.9346	0.9344	0.9346	0.9344
0.2051	2.8858	2300	0.2078	0.9329	0.9330	0.9329	0.9327
0.222	3.0113	2400	0.2021	0.9328	0.9330	0.9328	0.9327
0.1985	3.1368	2500	0.2068	0.9341	0.9342	0.9341	0.9340
0.1968	3.2622	2600	0.2145	0.932	0.9323	0.932	0.9318

Framework versions

• Transformers 4.56.1
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.0