rubert-tiny2-sentiment-tuned

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

• Loss: 0.2380
• Accuracy: 0.9248
• F1: 0.9274

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: 3e-07
• train_batch_size: 32
• eval_batch_size: 32
• 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: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.4865	0.1926	500	0.3455	0.8853	0.9123
0.3198	0.3852	1000	0.2753	0.9127	0.9218
0.2655	0.5778	1500	0.2513	0.9217	0.9260
0.2506	0.7704	2000	0.2415	0.9240	0.9270
0.2511	0.9630	2500	0.2380	0.9248	0.9274

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.1
• Tokenizers 0.21.0