Modelpublic

yoeel/bart-cnn-summarizer

sourceHugging Faceapache-2.0updated 21d agoView on Hugging Face

0likes500downloads

Model Card

bart-cnn-summarizer

This model is a fine-tuned version of facebook/bart-base on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 3.2134

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-05
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—gradientaccumulationsteps: 4
—totaltrainbatch_size: 32
—optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_steps: 100
—num_epochs: 1
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
19.4901	0.0178	100	3.7653
17.5543	0.0356	200	3.6700
17.3140	0.0533	300	3.5888
16.9665	0.0711	400	3.5865
16.7950	0.0889	500	3.5545
16.4116	0.1067	600	3.5126
16.6196	0.1244	700	3.4994
16.4861	0.1422	800	3.4613
16.3399	0.16	900	3.4798
16.1388	0.1778	1000	3.5040
16.0895	0.1956	1100	3.4076
16.0361	0.2133	1200	3.4577
15.9889	0.2311	1300	3.4254
16.0983	0.2489	1400	3.3988
15.9938	0.2667	1500	3.4130
15.7351	0.2844	1600	3.4220
15.9561	0.3022	1700	3.3938
15.9253	0.32	1800	3.3993
15.8845	0.3378	1900	3.3602
15.6629	0.3556	2000	3.3580
15.4323	0.3733	2100	3.3202
15.7600	0.3911	2200	3.3595
15.6233	0.4089	2300	3.3359
15.5126	0.4267	2400	3.3498
15.4304	0.4444	2500	3.3059
15.4607	0.4622	2600	3.3204
15.4800	0.48	2700	3.3184
15.3404	0.4978	2800	3.3151
15.3150	0.5156	2900	3.3021
15.4926	0.5333	3000	3.2872
15.3738	0.5511	3100	3.3185
15.5217	0.5689	3200	3.2836
15.3160	0.5867	3300	3.2977
15.3971	0.6044	3400	3.2818
15.2491	0.6222	3500	3.2802
15.2616	0.64	3600	3.2686
15.3559	0.6578	3700	3.2731
15.1660	0.6756	3800	3.2616
15.1404	0.6933	3900	3.2743
15.1153	0.7111	4000	3.2549
15.2221	0.7289	4100	3.2640
15.0089	0.7467	4200	3.2657
15.2302	0.7644	4300	3.2426
15.0945	0.7822	4400	3.2425
15.0907	0.8	4500	3.2428
15.0047	0.8178	4600	3.2483
15.1461	0.8356	4700	3.2180
15.0522	0.8533	4800	3.2334
14.9980	0.8711	4900	3.2294
15.0223	0.8889	5000	3.2200
14.8098	0.9067	5100	3.2266
14.9831	0.9244	5200	3.2287
14.9949	0.9422	5300	3.2185
14.8842	0.96	5400	3.2211
14.9345	0.9778	5500	3.2116
15.0163	0.9956	5600	3.2134

Framework versions

—Transformers 5.0.0
—Pytorch 2.10.0+cu128
—Datasets 5.0.0
—Tokenizers 0.22.2