Modelpublic

aellxx/preprocessed_disaster_tweets

sourceHugging Faceupdated 3y agoView on Hugging Face

0likes6downloads

Model Card

preprocesseddisastertweets

This model is a fine-tuned version of cardiffnlp/twitter-roberta-base-sentiment on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.4125

Training hyperparameters

The following hyperparameters were used during training:

—learning_rate: 5e-05
—trainbatchsize: 64
—evalbatchsize: 64
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
0.6693	0.12	12	0.6626
0.6712	0.25	24	0.6614
0.6604	0.38	36	0.6595
0.6667	0.5	48	0.6571
0.6574	0.62	60	0.6540
0.6595	0.75	72	0.6500
0.6545	0.88	84	0.6456
0.6502	1.0	96	0.6405
0.6437	1.12	108	0.6347
0.6393	1.25	120	0.6280
0.6496	1.38	132	0.6218
0.6196	1.5	144	0.6146
0.6299	1.62	156	0.6068
0.6068	1.75	168	0.5976
0.6038	1.88	180	0.5878
0.5786	2.0	192	0.5770
0.5691	2.12	204	0.5663
0.5619	2.25	216	0.5552
0.5758	2.38	228	0.5442
0.5249	2.5	240	0.5326
0.5264	2.62	252	0.5207
0.5167	2.75	264	0.5090
0.5259	2.88	276	0.4991
0.4738	3.0	288	0.4917
0.4681	3.12	300	0.4795
0.4482	3.25	312	0.4698
0.4354	3.38	324	0.4616
0.465	3.5	336	0.4531
0.4593	3.62	348	0.4469
0.4294	3.75	360	0.4467
0.4284	3.88	372	0.4374
0.4337	4.0	384	0.4347
0.3805	4.12	396	0.4318
0.4113	4.25	408	0.4269
0.4155	4.38	420	0.4201
0.4112	4.5	432	0.4205
0.3921	4.62	444	0.4228
0.4018	4.75	456	0.4151
0.3995	4.88	468	0.4102
0.3891	5.0	480	0.4125

Framework versions

—Transformers 4.28.1
—Pytorch 1.13.0
—Datasets 2.1.0
—Tokenizers 0.13.2