turkish-sentiment4

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

• Loss: 0.1083
• Accuracy: 0.9615
• F1 Macro: 0.9351
• F1 Weighted: 0.9612
• Precision Macro: 0.9415
• Recall Macro: 0.9291

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 400
• training_steps: 2000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	F1 Weighted	Precision Macro	Recall Macro
0.1987	0.0073	200	0.1654	0.9400	0.9074	0.9413	0.8944	0.9236
0.1978	0.0145	400	0.1653	0.9407	0.9092	0.9427	0.8938	0.9303
0.1444	0.0218	600	0.1572	0.9520	0.9157	0.9502	0.9425	0.8957
0.1446	0.0290	800	0.1252	0.9560	0.9239	0.9549	0.9407	0.9101
0.1067	0.0363	1000	0.1389	0.9574	0.9294	0.9574	0.9313	0.9275
0.1333	0.0436	1200	0.1244	0.9587	0.9279	0.9576	0.9464	0.9128
0.0935	0.0508	1400	0.1297	0.9596	0.9336	0.9598	0.9309	0.9365
0.1072	0.0581	1600	0.1196	0.9604	0.9309	0.9593	0.9497	0.9156
0.1235	0.0654	1800	0.1121	0.9613	0.9338	0.9607	0.9450	0.9240
0.1097	0.0726	2000	0.1083	0.9615	0.9351	0.9612	0.9415	0.9291

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2