bert-small-UnidicUnigram

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

• Loss: 1.1279
• Accuracy: 0.7455

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: 14.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.5872	1.0	69473	1.4531	0.6867
1.4695	2.0	138946	1.3340	0.7073
1.4136	3.0	208419	1.2793	0.7173
1.3779	4.0	277892	1.2490	0.7227
1.3546	5.0	347365	1.2227	0.7277
1.3353	6.0	416838	1.2070	0.7307
1.3182	7.0	486311	1.1895	0.7334
1.3058	8.0	555784	1.1777	0.7360
1.2974	9.0	625257	1.1660	0.7378
1.2857	10.0	694730	1.1543	0.7401
1.2755	11.0	764203	1.1514	0.7408
1.2694	12.0	833676	1.1377	0.7431
1.2623	13.0	903149	1.1338	0.7442
1.2587	14.0	972622	1.1279	0.7455

Framework versions

• Transformers 4.19.2
• Pytorch 1.12.0+cu116
• Datasets 2.9.0
• Tokenizers 0.12.1