amazon_helpfulness_classification_on_base_from_DAPT_5M_pretraining

This model is a fine-tuned version of BigTMiami/amazon_pretraining_5M_model_corrected on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7517
• Accuracy: 0.8699
• F1 Macro: 0.6736

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-06
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro
0.3151	1.0	7204	0.3461	0.8766	0.6664
0.2861	2.0	14408	0.3429	0.8736	0.6544
0.2788	3.0	21612	0.3600	0.8722	0.6466
0.2585	4.0	28816	0.3805	0.8682	0.6789
0.1873	5.0	36020	0.5306	0.871	0.6660
0.1333	6.0	43224	0.6493	0.8674	0.6675
0.1369	7.0	50428	0.7657	0.869	0.6799
0.0936	8.0	57632	0.8041	0.8674	0.6779
0.1062	9.0	64836	0.9458	0.867	0.6633
0.0463	10.0	72040	1.0079	0.8682	0.6684

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2