char-bert-base-uncased

This model is a fine-tuned version of char-bert-base-uncased/checkpoint-1840240 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1760

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss
0.8329	1.0	92012	0.4066
0.4066	2.0	184024	0.3223
0.3422	3.0	276036	0.2803
0.3044	4.0	368048	0.2560
0.2782	5.0	460060	0.2399
0.2593	6.0	552072	0.2265
0.2693	7.0	644084	0.2366
0.2559	8.0	736096	0.2228
0.2431	9.0	828108	0.2112
0.2334	10.0	920120	0.2103
0.2453	11.0	1012132	0.2164
0.2372	12.0	1104144	0.2113
0.2288	13.0	1196156	0.2004
0.2208	14.0	1288168	0.2002
0.2152	15.0	1380180	0.1941
0.2241	16.0	1472192	0.1940
0.2188	17.0	1564204	0.1954
0.2132	18.0	1656216	0.1968
0.2077	19.0	1748228	0.1887
0.2036	20.0	1840240	0.1863
0.2109	21.0	1932252	0.2009
0.2075	22.0	2024264	0.1840
0.2031	23.0	2116276	0.1884
0.1992	24.0	2208288	0.1902
0.196	25.0	2300300	0.1760

Framework versions

• Transformers 4.20.1
• Pytorch 1.9.0+cu111
• Datasets 2.3.2
• Tokenizers 0.12.1