distilbert-sentiment-imdb-cft

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

• Accuracy: 0.9856
• Pos Similarity: 0.9538
• Neg Similarity: 0.4913
• F1: 0.9927
• Precision: 1.0
• Recall: 0.9856
• Loss: 3.5397

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Accuracy	Pos Similarity	Neg Similarity	F1	Precision	Recall	Validation Loss
3.8563	1.1364	200	0.9728	0.9662	0.7048	0.9862	1.0	0.9728	3.5778
3.5857	2.2727	400	0.9848	0.9666	0.5691	0.9923	1.0	0.9848	3.5278
3.5032	3.4091	600	0.9856	0.9538	0.4913	0.9927	1.0	0.9856	3.5397

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1
• Datasets 3.1.0
• Tokenizers 0.20.3