Modelpublic

jayanta/distilbert-base-uncased-sentiment-finetuned-memes-20epoch

sourceHugging Faceapache-2.0updated 4y agoView on Hugging Face

0likes25downloads

Model Card

distilbert-base-uncased-sentiment-finetuned-memes-20epoch

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

—Loss: 2.2252
—Accuracy: 0.8160
—Precision: 0.8165
—Recall: 0.8160
—F1: 0.8162

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
—trainbatchsize: 4
—evalbatchsize: 4
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.5319	1.0	4293	0.5560	0.7699	0.7777	0.7699	0.7704
0.4627	2.0	8586	0.6588	0.7856	0.7868	0.7856	0.7860
0.3735	3.0	12879	0.8583	0.7835	0.7854	0.7835	0.7813
0.3549	4.0	17172	1.0078	0.7872	0.7869	0.7872	0.7866
0.2995	5.0	21465	1.0007	0.8024	0.8030	0.8024	0.8012
0.2579	6.0	25758	1.1501	0.8150	0.8152	0.8150	0.8151
0.2078	7.0	30051	1.2604	0.8097	0.8137	0.8097	0.8102
0.1635	8.0	34344	1.3755	0.8092	0.8092	0.8092	0.8085
0.1453	9.0	38637	1.4639	0.8097	0.8137	0.8097	0.8102
0.1431	10.0	42930	1.5612	0.8050	0.8048	0.8050	0.8044
0.085	11.0	47223	1.8216	0.8097	0.8121	0.8097	0.8101
0.0693	12.0	51516	1.7761	0.8087	0.8100	0.8087	0.8090
0.041	13.0	55809	1.8538	0.8082	0.8083	0.8082	0.8082
0.0391	14.0	60102	2.0022	0.8160	0.8158	0.8160	0.8158
0.0299	15.0	64395	2.0101	0.8124	0.8121	0.8124	0.8121
0.0226	16.0	68688	2.0396	0.8150	0.8152	0.8150	0.8151
0.0229	17.0	72981	2.1071	0.8171	0.8170	0.8171	0.8171
0.0133	18.0	77274	2.1047	0.8181	0.8182	0.8181	0.8182
0.0268	19.0	81567	2.2037	0.8208	0.8208	0.8208	0.8208
0.0068	20.0	85860	2.2252	0.8160	0.8165	0.8160	0.8162

Framework versions

—Transformers 4.22.1
—Pytorch 1.12.1+cu113
—Datasets 2.4.0
—Tokenizers 0.12.1