testc8-2

This model is a fine-tuned version of shafin/chemical-bert-uncased-finetuned-cust-c2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2346

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.6173	1.0	16	0.3874
0.5383	2.0	32	0.3227
0.4756	3.0	48	0.3142
0.4399	4.0	64	0.3404
0.4462	5.0	80	0.3112
0.4187	6.0	96	0.3185
0.4023	7.0	112	0.2628
0.3712	8.0	128	0.2807
0.3922	9.0	144	0.2516
0.3483	10.0	160	0.1995
0.3417	11.0	176	0.2452
0.3585	12.0	192	0.2236
0.3413	13.0	208	0.2031
0.3452	14.0	224	0.2238
0.317	15.0	240	0.2229
0.3161	16.0	256	0.2591
0.3338	17.0	272	0.2599
0.2949	18.0	288	0.2618
0.3035	19.0	304	0.2436
0.3108	20.0	320	0.2015
0.289	21.0	336	0.2329
0.3144	22.0	352	0.1940
0.2606	23.0	368	0.2334
0.2842	24.0	384	0.1996
0.2892	25.0	400	0.2330
0.2612	26.0	416	0.2163
0.2669	27.0	432	0.2053
0.3147	28.0	448	0.1555
0.286	29.0	464	0.1983
0.2857	30.0	480	0.2346

Framework versions

• Transformers 4.24.0
• Pytorch 1.12.1+cu113
• Datasets 2.7.1
• Tokenizers 0.13.2