results

This model is a fine-tuned version of cointegrated/rubert-tiny2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1492
• Model Preparation Time: 0.002
• Precision: 0.7282
• Recall: 0.8213
• F1: 0.7719
• Accuracy: 0.9547

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: 4
• eval_batch_size: 4
• 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
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Precision	Recall	F1	Accuracy
No log	1.0	200	0.4280	0.002	0.4559	0.5042	0.4788	0.8802
No log	2.0	400	0.2747	0.002	0.6052	0.7073	0.6523	0.9269
0.5507	3.0	600	0.2190	0.002	0.6511	0.7656	0.7037	0.9385
0.5507	4.0	800	0.1885	0.002	0.6735	0.7855	0.7252	0.9455
0.2228	5.0	1000	0.1712	0.002	0.6906	0.8009	0.7416	0.9491
0.2228	6.0	1200	0.1656	0.002	0.7214	0.8158	0.7657	0.9518
0.2228	7.0	1400	0.1584	0.002	0.7236	0.8183	0.7680	0.9530
0.1613	8.0	1600	0.1511	0.002	0.7367	0.8218	0.7769	0.9550
0.1613	9.0	1800	0.1497	0.002	0.7313	0.8223	0.7741	0.9547
0.1379	10.0	2000	0.1492	0.002	0.7282	0.8213	0.7719	0.9547

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0
• Datasets 3.5.0
• Tokenizers 0.21.1