metadata

license: apache-2.0
base_model: distilbert-base-uncased
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: emotions_distilbert_im
    results: []

emotions_distilbert_im

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

Loss: 0.5190
F1 Micro: 0.6929
F1 Macro: 0.5934
Accuracy: 0.2362

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: 0.0001
train_batch_size: 128
eval_batch_size: 128
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Micro	F1 Macro	Accuracy
0.7627	0.41	20	0.6619	0.5987	0.3531	0.1599
0.6258	0.82	40	0.5710	0.6609	0.4929	0.1929
0.5609	1.22	60	0.5496	0.6672	0.5097	0.2233
0.504	1.63	80	0.5260	0.6765	0.5840	0.1948
0.4856	2.04	100	0.5152	0.6864	0.5912	0.1981
0.4246	2.45	120	0.5190	0.6929	0.5934	0.2362
0.4157	2.86	140	0.5321	0.6757	0.5763	0.2058
0.4027	3.27	160	0.5186	0.6800	0.5914	0.2149
0.346	3.67	180	0.5274	0.6725	0.5888	0.1981
0.3562	4.08	200	0.5346	0.6811	0.5884	0.2272
0.3097	4.49	220	0.5413	0.6767	0.5863	0.2136
0.2989	4.9	240	0.5637	0.6832	0.5886	0.2259
0.2701	5.31	260	0.5745	0.6803	0.5911	0.2272
0.2505	5.71	280	0.5946	0.6783	0.5807	0.2155
0.2508	6.12	300	0.6194	0.6822	0.5764	0.2272
0.2171	6.53	320	0.6293	0.6800	0.5790	0.2181
0.2164	6.94	340	0.6322	0.6805	0.5806	0.2097
0.1949	7.35	360	0.6663	0.6775	0.5709	0.2155
0.1852	7.76	380	0.6763	0.6768	0.5749	0.2129
0.1821	8.16	400	0.6757	0.6791	0.5743	0.2227
0.1653	8.57	420	0.6862	0.6757	0.5728	0.2155
0.166	8.98	440	0.6989	0.6786	0.5749	0.2233
0.1593	9.39	460	0.7019	0.6784	0.5765	0.2220
0.1512	9.8	480	0.7010	0.6781	0.5744	0.2233

Framework versions

Transformers 4.39.3
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2