bert-small-spm

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

• Loss: 2.5919
• Accuracy: 0.5095

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: 0.0001
• train_batch_size: 256
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 3
• total_train_batch_size: 768
• total_eval_batch_size: 24
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.01
• num_epochs: 14
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.3946	1.0	69473	3.2473	0.4299
3.1526	2.0	138946	2.9987	0.4583
3.0496	3.0	208419	2.8875	0.4715
2.9923	4.0	277892	2.8258	0.4788
2.9429	5.0	347365	2.7765	0.4849
2.912	6.0	416838	2.7482	0.4890
2.8813	7.0	486311	2.7103	0.4938
2.8609	8.0	555784	2.6881	0.4963
2.8352	9.0	625257	2.6702	0.4991
2.8163	10.0	694730	2.6510	0.5010
2.8026	11.0	764203	2.6246	0.5046
2.7894	12.0	833676	2.6172	0.5055
2.7728	13.0	903149	2.5994	0.5083
2.761	14.0	972622	2.5919	0.5095

Framework versions

• Transformers 4.19.2
• Pytorch 1.12.0+cu116
• Datasets 2.2.2
• Tokenizers 0.12.1