Modelpublic

HouraMor/wh-loraft-lr5e6-dtstf5-adm-ga1ba16-st15k-v2-evalstp50-pat20-trainvalch

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face

0likes11downloads

Model Card

wh-loraft-lr5e6-dtstf5-adm-ga1ba16-st15k-v2-evalstp50-pat20-trainvalch

This model is a fine-tuned version of HouraMor/wh-ft-lr5e6-dtstf5-adm-ga1ba16-st15k-v2-evalstp500-pat5 on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.5748

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-06
—trainbatchsize: 16
—evalbatchsize: 8
—seed: 42
—optimizer: Use adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_steps: 250
—training_steps: 5000
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.3046	0.1004	50	0.5668
0.3436	0.2008	100	0.5666
0.3602	0.3012	150	0.5670
0.2937	0.4016	200	0.5672
0.3096	0.5020	250	0.5671
0.2671	0.6024	300	0.5678
0.3461	0.7028	350	0.5690
0.2273	0.8032	400	0.5701
0.2773	0.9036	450	0.5710
0.4484	1.0040	500	0.5714
0.2035	1.1044	550	0.5712
0.1835	1.2048	600	0.5720
0.278	1.3052	650	0.5729
0.2929	1.4056	700	0.5741
0.3285	1.5060	750	0.5742
0.3162	1.6064	800	0.5747
0.2685	1.7068	850	0.5748

Framework versions

—PEFT 0.15.2
—Transformers 4.52.3
—Pytorch 2.7.0+cu118
—Datasets 3.6.0
—Tokenizers 0.21.1