BERT_Emotions_tuned

This model is a fine-tuned version of bert-base-uncased on the emotion dataset. It achieves the following results on the evaluation set:

• Loss: 0.2033
• Accuracy: 0.9295

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.1	100	0.8098	0.7195
No log	0.2	200	0.4054	0.882
No log	0.3	300	0.4686	0.877
No log	0.4	400	0.2850	0.909
0.5652	0.5	500	0.2673	0.92
0.5652	0.6	600	0.2474	0.9255
0.5652	0.7	700	0.1943	0.933
0.5652	0.8	800	0.1779	0.9315
0.5652	0.9	900	0.1720	0.939
0.2212	1.0	1000	0.1747	0.9375
0.2212	1.1	1100	0.1902	0.933
0.2212	1.2	1200	0.1540	0.941
0.2212	1.3	1300	0.1599	0.937
0.2212	1.4	1400	0.1533	0.944
0.1315	1.5	1500	0.1421	0.937
0.1315	1.6	1600	0.1549	0.941
0.1315	1.7	1700	0.1284	0.9435
0.1315	1.8	1800	0.1376	0.934
0.1315	1.9	1900	0.1197	0.943
0.1204	2.0	2000	0.1319	0.9385
0.1204	2.1	2100	0.1535	0.935
0.1204	2.2	2200	0.1488	0.943
0.1204	2.3	2300	0.1583	0.94
0.1204	2.4	2400	0.1426	0.9425
0.0913	2.5	2500	0.1554	0.9395
0.0913	2.6	2600	0.1458	0.944
0.0913	2.7	2700	0.1504	0.943
0.0913	2.8	2800	0.1621	0.9465
0.0913	2.9	2900	0.1521	0.944
0.0842	3.0	3000	0.1533	0.944

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2