electricidad-small-finetuned-amazon-review-classification

This model is a fine-tuned version of mrm8488/electricidad-small-discriminator on the amazon_reviews_multi dataset. It achieves the following results on the evaluation set:

• Loss: 0.9601
• Accuracy: 0.581

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0136	1.0	25000	1.0153	0.5414
0.9416	2.0	50000	0.9942	0.5576

Framework versions

• Transformers 4.17.0
• Pytorch 1.10.0+cu111
• Datasets 2.0.0
• Tokenizers 0.11.6