results_pegasus6_hiba_wiki

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

• Loss: 0.0699
• Rouge1 Fmeasure: 0.3654
• Rouge2 Fmeasure: 0.2931
• Rougel Fmeasure: 0.3466
• Meteor: 0.2623
• Bleu: 0.0000
• Bertscore P: 0.9267
• Bertscore R: 0.9204
• Bertscore F1: 0.9235

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
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 250
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1 Fmeasure	Rouge2 Fmeasure	Rougel Fmeasure	Meteor	Bleu	Bertscore P	Bertscore R	Bertscore F1
0.2433	0.5222	500	0.2175	0.0259	0.0060	0.0259	0.1659	0.0000	0.9282	0.8784	0.9026
0.2165	1.0444	1000	0.1818	0.0532	0.0247	0.0501	0.2343	0.0000	0.9405	0.9278	0.9341
0.1815	1.5666	1500	0.1535	0.1065	0.0529	0.0833	0.2365	0.0000	0.9369	0.9215	0.9291
0.1558	2.0888	2000	0.1352	0.1544	0.0836	0.1246	0.2442	0.0000	0.9342	0.9247	0.9294
0.1449	2.6110	2500	0.1250	0.1934	0.0940	0.1573	0.2439	0.0000	0.9338	0.9242	0.9289
0.1356	3.1332	3000	0.1107	0.2681	0.1641	0.2233	0.2477	0.0000	0.9259	0.9199	0.9229
0.1256	3.6554	3500	0.1063	0.2722	0.1793	0.2380	0.2497	0.0000	0.9263	0.9204	0.9233
0.1187	4.1775	4000	0.0975	0.2946	0.2058	0.2663	0.2534	0.0000	0.9272	0.9207	0.9239
0.1157	4.6997	4500	0.0936	0.2898	0.2032	0.2546	0.2456	0.0000	0.9289	0.9205	0.9247
0.1078	5.2219	5000	0.0876	0.3094	0.2179	0.2811	0.2521	0.0000	0.9272	0.9207	0.9239
0.1029	5.7441	5500	0.0849	0.3324	0.2481	0.3016	0.2571	0.0000	0.9267	0.9204	0.9235
0.1029	6.2663	6000	0.0827	0.3357	0.2546	0.3146	0.2575	0.0000	0.9284	0.9199	0.9241
0.0992	6.7885	6500	0.0783	0.3510	0.2765	0.3238	0.2606	0.0000	0.9268	0.9202	0.9234
0.0954	7.3107	7000	0.0754	0.3327	0.2611	0.3139	0.2588	0.0000	0.9267	0.9204	0.9235
0.0938	7.8329	7500	0.0746	0.3592	0.2863	0.3404	0.2614	0.0000	0.9300	0.9216	0.9258
0.0938	8.3551	8000	0.0728	0.3718	0.3022	0.3492	0.2613	0.0000	0.9268	0.9204	0.9236
0.09	8.8773	8500	0.0707	0.3600	0.2882	0.3442	0.2633	0.0000	0.9267	0.9204	0.9235
0.0911	9.3995	9000	0.0705	0.3677	0.2933	0.3464	0.2646	0.0000	0.9267	0.9204	0.9235
0.0881	9.9217	9500	0.0699	0.3654	0.2931	0.3466	0.2623	0.0000	0.9267	0.9204	0.9235

Framework versions

• Transformers 4.42.0.dev0
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1