Modelpublic

me1adi/Kvasir-VQA-x1-lora_260823-1829

sourceHugging Faceupdated 1mo agoView on Hugging Face

0likes12downloads

Model Card

Kvasir-VQA-x1-lora_260823-1829

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.4089
—Token Acc: 0.8573

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: 0.0002
—trainbatchsize: 4
—evalbatchsize: 4
—seed: 42
—gradientaccumulationsteps: 4
—totaltrainbatch_size: 16
—optimizer: Use adamwtorchfused with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_steps: 0.03
—num_epochs: 2.0
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Token Acc
0.6074	0.1114	1000	0.6064	0.8035
0.5380	0.2228	2000	0.5472	0.8191
0.5199	0.3343	3000	0.5312	0.8215
0.5344	0.4457	4000	0.5069	0.8283
0.4680	0.5571	5000	0.4901	0.8312
0.4942	0.6685	6000	0.4727	0.8360
0.4518	0.7800	7000	0.4644	0.8396
0.4598	0.8914	8000	0.4593	0.8425
0.4164	1.0028	9000	0.4526	0.8438
0.4103	1.1142	10000	0.4401	0.8479
0.3861	1.2256	11000	0.4390	0.8485
0.3911	1.3371	12000	0.4342	0.8513
0.4091	1.4485	13000	0.4259	0.8492
0.3802	1.5599	14000	0.4209	0.8530
0.3847	1.6713	15000	0.4190	0.8545
0.3460	1.7828	16000	0.4133	0.8554
0.3598	1.8942	17000	0.4109	0.8563
0.3611	2.0	17950	0.4089	0.8573

Framework versions

—PEFT 0.19.1
—Transformers 5.15.1
—Pytorch 2.13.0+cu130
—Datasets 4.8.4
—Tokenizers 0.22.2