tapt_helpfulness_seq_bn_pretraining_model_full_train

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.4917

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: 21
• eval_batch_size: 21
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 42
• optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-06
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
2.746	1.0	1068	1.8309
1.8701	2.0	2137	1.6877
1.7711	3.0	3205	1.6275
1.7178	4.0	4274	1.5909
1.6876	5.0	5342	1.5788
1.6638	6.0	6411	1.5636
1.6526	7.0	7479	1.5344
1.6357	8.0	8548	1.5402
1.626	9.0	9616	1.5097
1.6144	10.0	10685	1.5111
1.611	11.0	11753	1.5248
1.603	12.0	12822	1.4989
1.6003	13.0	13890	1.5071
1.5915	14.0	14959	1.4807
1.5893	15.0	16027	1.4892
1.5857	16.0	17096	1.4794
1.5839	17.0	18164	1.4893
1.5806	18.0	19233	1.4787
1.5808	19.0	20301	1.4872
1.5781	19.99	21360	1.4917

Framework versions

• Transformers 4.36.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.15.2