Lolimorimorf/ranlp_2023_russian_propaganda_opposition_model

ru_propaganda_opposition_model_without_foreign_agent_mask_large_6

RANLP_2023_ru_propaganda_opposition_model

This model is a fine-tuned version of DeepPavlov/bert-base-bg-cs-pl-ru-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 0.0003
• Validation Loss: 0.2103
• Train Accuracy: 0.9606
• Epoch: 14

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:

• optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 6150, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Validation Loss	Train Accuracy	Epoch
0.2650	0.1440	0.9452	0
0.0881	0.1171	0.9529	1
0.0377	0.1506	0.9573	2
0.0202	0.1724	0.9562	3
0.0116	0.1534	0.9617	4
0.0064	0.2133	0.9540	5
0.0093	0.1844	0.9628	6
0.0074	0.1427	0.9606	7
0.0018	0.1757	0.9639	8
0.0017	0.2006	0.9606	9
0.0006	0.1996	0.9595	10
0.0007	0.2004	0.9628	11
0.0004	0.2052	0.9628	12
0.0003	0.2090	0.9606	13
0.0003	0.2103	0.9606	14

Framework versions

• Transformers 4.28.1
• TensorFlow 2.12.0
• Datasets 2.11.0
• Tokenizers 0.13.3