Modelpublic

hosseinmhmdkhani/mt5-small-finetuned-persian-news-summarization-with-additional-hyperparameters

sourceHugging Faceapache-2.0updated 8d agoView on Hugging Face

0likes97downloads

Model Card

mt5-small-finetuned-persian-news-summarization-with-additional-hyperparameters

This model is a fine-tuned version of google/mt5-small on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 1.2368
—Rouge1 Precision: 61.1637
—Rouge1 Recall: 50.8564
—Rouge1 F1: 53.4468
—Rouge2 Precision: 43.7489
—Rouge2 Recall: 36.4625
—Rouge2 F1: 38.2205
—Rougel Precision: 53.1324
—Rougel Recall: 44.3405
—Rougel F1: 46.5079

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: 4e-05
—trainbatchsize: 4
—evalbatchsize: 4
—seed: 42
—gradientaccumulationsteps: 8
—totaltrainbatch_size: 32
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: cosine
—lrschedulerwarmup_ratio: 0.1
—num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1 Precision	Rouge1 Recall	Rouge1 F1	Rouge2 Precision	Rouge2 Recall	Rouge2 F1	Rougel Precision	Rougel Recall	Rougel F1
2.4339	1.0	302	1.3726	52.4552	30.2138	35.7525	34.2706	19.5416	23.154	45.1553	25.9579	30.6864
1.7536	2.0	604	1.2656	61.595	48.0181	51.6818	44.1499	34.4709	37.0359	53.7054	41.9632	45.111
1.644	3.0	906	1.2390	61.3387	50.3584	53.2007	43.9459	36.1575	38.0967	53.3489	43.9572	46.3486
1.6269	4.0	1208	1.2368	61.1637	50.8564	53.4468	43.7489	36.4625	38.2205	53.1324	44.3405	46.5079

Framework versions

—Transformers 4.57.6
—Pytorch 2.11.0+cu128
—Datasets 4.8.5
—Tokenizers 0.22.2