distilroberta-base-finegrain

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

• Loss: 0.3713
• F1: 0.9129

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1
0.3944	1.0	844	0.3998	0.9163
0.3886	2.0	1688	0.4020	0.9163
0.3899	3.0	2532	0.3423	0.9163
0.4026	4.0	3376	0.3837	0.9163
0.3911	5.0	4220	0.3492	0.9163
0.3856	6.0	5064	0.3504	0.9163
0.4058	7.0	5908	0.3682	0.9163
0.4057	8.0	6752	0.3767	0.9163
0.3807	9.0	7596	0.3519	0.9163
0.394	10.0	8440	0.3603	0.9163
0.39	11.0	9284	0.3630	0.9163
0.3945	12.0	10128	0.3846	0.9163
0.3948	13.0	10972	0.3711	0.9163
0.3981	14.0	11816	0.3516	0.9163
0.4144	15.0	12660	0.3526	0.9163
0.3999	16.0	13504	0.3560	0.9163
0.376	17.0	14348	0.3671	0.9163
0.412	18.0	15192	0.3630	0.9163
0.389	19.0	16036	0.3669	0.9136
0.374	20.0	16880	0.3713	0.9129

Framework versions

• Transformers 4.28.1
• Pytorch 1.13.1+cu117
• Datasets 2.12.0
• Tokenizers 0.13.3