ruBertTiny_attr_name_addv3

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: 0.2154
• Accuracy: 0.9188

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5515	0.2739	10000	0.4502	0.7949
0.49	0.5478	20000	0.3874	0.8205
0.4632	0.8217	30000	0.3556	0.8376
0.4367	1.0956	40000	0.3381	0.8419
0.4122	1.3695	50000	0.3138	0.8590
0.3989	1.6434	60000	0.2787	0.8932
0.389	1.9173	70000	0.2741	0.8846
0.3669	2.1912	80000	0.2523	0.9017
0.3566	2.4651	90000	0.2459	0.8932
0.3502	2.7391	100000	0.2343	0.9017
0.3458	3.0130	110000	0.2248	0.9145
0.3281	3.2869	120000	0.2203	0.9145
0.3255	3.5608	130000	0.2162	0.9145
0.3234	3.8347	140000	0.2176	0.9274
0.3174	4.1086	150000	0.2147	0.9188
0.3126	4.3825	160000	0.2138	0.9145
0.3127	4.6564	170000	0.2155	0.9188
0.3126	4.9303	180000	0.2154	0.9188

Framework versions

• Transformers 4.44.1
• Pytorch 2.0.1+cu118
• Datasets 2.21.0
• Tokenizers 0.19.1