Modelpublic

abdulrahman-nuzha/intfloat-multilingual-e5-large-arabic-fp16-allagree

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes7downloads

Model Card

intfloat-multilingual-e5-large-arabic-fp16-allagree

This model is a fine-tuned version of intfloat/multilingual-e5-large on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.1942
—Accuracy: 0.9347
—Precision: 0.9349
—Recall: 0.9347
—F1: 0.9346

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: 64
—evalbatchsize: 64
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 128
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_ratio: 0.3
—num_epochs: 10
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.9731	0.7463	50	0.6211	0.7603	0.8128	0.7603	0.6907
0.375	1.4925	100	0.1942	0.9347	0.9349	0.9347	0.9346
0.1713	2.2388	150	0.1802	0.9412	0.9414	0.9412	0.9412
0.1466	2.9851	200	0.2002	0.9291	0.9303	0.9291	0.9287
0.0936	3.7313	250	0.2041	0.9356	0.9356	0.9356	0.9356
0.0577	4.4776	300	0.2556	0.9431	0.9436	0.9431	0.9432

Framework versions

—Transformers 4.51.1
—Pytorch 2.6.0+cu124
—Datasets 3.5.0
—Tokenizers 0.21.1