imdb_sentiment_distilbert

This model is a fine-tuned version of distilbert-base-uncased-finetuned-sst-2-english on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1952
• Accuracy: 0.9272
• Precision: 0.9329
• Recall: 0.9242
• F1: 0.9285

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: 1e-06
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• 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	Accuracy	Precision	Recall	F1
0.2944	1.0	625	0.2280	0.9144	0.9157	0.9172	0.9164
0.235	2.0	1250	0.2165	0.9176	0.9235	0.9148	0.9191
0.2188	3.0	1875	0.2076	0.9208	0.9184	0.9277	0.9230
0.2044	4.0	2500	0.2030	0.9222	0.9275	0.9199	0.9237
0.1958	5.0	3125	0.2003	0.9228	0.9236	0.9258	0.9247
0.1905	6.0	3750	0.1984	0.9234	0.9237	0.9269	0.9253
0.1833	7.0	4375	0.1968	0.9248	0.9333	0.9187	0.9260
0.1708	8.0	5000	0.1958	0.9264	0.9268	0.9297	0.9282
0.1726	9.0	5625	0.1959	0.9268	0.9232	0.9347	0.9289
0.1677	10.0	6250	0.1940	0.9252	0.9299	0.9234	0.9267
0.1677	11.0	6875	0.1943	0.9266	0.9238	0.9336	0.9287
0.1585	12.0	7500	0.1952	0.927	0.9339	0.9226	0.9282
0.1585	13.0	8125	0.1952	0.9272	0.9329	0.9242	0.9285

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1