gbert-large-finetuned-cust18

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

• Loss: 0.1232

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: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.604	1.0	391	0.3560
0.3497	2.0	782	0.2838
0.2812	3.0	1173	0.2484
0.2452	4.0	1564	0.2232
0.2253	5.0	1955	0.2240
0.2202	6.0	2346	0.1993
0.1922	7.0	2737	0.1747
0.182	8.0	3128	0.1631
0.1609	9.0	3519	0.1555
0.1553	10.0	3910	0.1434
0.147	11.0	4301	0.1399
0.144	12.0	4692	0.1340
0.1307	13.0	5083	0.1319
0.128	14.0	5474	0.1490
0.1304	15.0	5865	0.1338
0.1165	16.0	6256	0.1233
0.1456	17.0	6647	0.1673
0.1419	18.0	7038	0.1591
0.1447	19.0	7429	0.1360
0.1317	20.0	7820	0.1232

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.3