dataset-5400

This model is a fine-tuned version of facebook/mbart-large-50 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.6678
• Gen Len: 16.2956
• Rouge-1: 61.8854
• Rouge-2: 51.9874
• Rouge-l: 61.5994

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: polynomial
• lr_scheduler_warmup_steps: 1000
• num_epochs: 50
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Gen Len	Rouge-1	Rouge-2	Rouge-l
No log	1.0	642	3.1256	18.2467	34.4667	18.0906	33.8972
No log	2.0	1284	3.0535	19.7	35.0866	18.7498	34.6592
No log	3.0	1926	2.9782	18.5556	36.3096	19.6731	35.8024
No log	4.0	2568	2.9259	15.9622	37.2212	19.9257	36.5795
No log	5.0	3210	2.8784	18.0756	39.3406	21.8734	38.7753
No log	6.0	3852	2.8503	17.6089	38.3231	21.1923	37.7442
No log	7.0	4494	2.8441	17.0822	39.6139	23.1286	39.0491
3.1832	8.0	5136	2.7900	17.0111	42.3608	24.9947	41.8427
3.1832	9.0	5778	2.7731	16.1467	41.4778	24.2298	41.044
3.1832	10.0	6420	2.7838	17.5978	42.3125	25.3928	41.6762
3.1832	11.0	7062	2.7627	15.3511	42.4708	25.7843	42.0607
3.1832	12.0	7704	2.7382	16.6333	45.7431	28.9995	45.1587
3.1832	13.0	8346	2.7240	16.8978	44.2626	28.2948	43.6046
3.1832	14.0	8988	2.7129	16.4311	47.5648	32.3034	47.1341
3.1832	15.0	9630	2.6917	16.54	47.0207	31.3636	46.4102
2.4158	16.0	10272	2.7043	16.2956	47.5201	31.9858	47.1196
2.4158	17.0	10914	2.6951	16.1467	48.0773	32.6974	47.7123
2.4158	18.0	11556	2.7118	16.18	49.4704	34.8365	49.2157
2.4158	19.0	12198	2.6950	16.7111	50.4711	36.0529	50.0663
2.4158	20.0	12840	2.6708	16.7133	51.5121	36.6734	51.2274
2.4158	21.0	13482	2.6730	16.1444	50.5072	35.9911	49.9929
2.4158	22.0	14124	2.6710	15.9867	50.9642	36.5418	50.5949
2.4158	23.0	14766	2.6748	15.9156	52.8178	39.0075	52.5719
2.1866	24.0	15408	2.6639	15.5133	52.3247	37.9551	51.9185
2.1866	25.0	16050	2.6949	16.4578	53.5261	40.8743	53.3567
2.1866	26.0	16692	2.6709	16.9267	54.6274	42.0288	54.2419
2.1866	27.0	17334	2.6668	15.5622	53.2566	40.0637	53.136
2.1866	28.0	17976	2.6578	16.2756	57.4156	44.4816	57.045
2.1866	29.0	18618	2.6522	15.5689	54.1314	41.6894	54.0261
2.1866	30.0	19260	2.6645	16.1222	56.879	44.8673	56.6811
2.1866	31.0	19902	2.6625	16.6333	58.1119	46.0585	57.8146
2.0501	32.0	20544	2.6512	15.8844	57.1061	45.1091	56.9354
2.0501	33.0	21186	2.6499	16.2022	58.2516	47.2457	57.8831
2.0501	34.0	21828	2.6727	16.1467	57.596	46.2813	57.3532
2.0501	35.0	22470	2.6673	16.1	58.9716	48.0406	58.6796
2.0501	36.0	23112	2.6556	16.6867	59.6493	48.5384	59.347
2.0501	37.0	23754	2.6523	16.3778	59.1905	48.2495	58.7407
2.0501	38.0	24396	2.6416	16.8333	60.8048	50.8419	60.5507
1.9593	39.0	25038	2.6553	16.2311	59.0907	48.4354	58.9027
1.9593	40.0	25680	2.6510	16.0222	60.5903	49.821	60.3967
1.9593	41.0	26322	2.6703	16.0044	59.9114	49.1521	59.7299
1.9593	42.0	26964	2.6593	16.1356	60.7095	50.1594	60.4683
1.9593	43.0	27606	2.6678	16.2956	61.8854	51.9874	61.5994

Framework versions

• Transformers 4.35.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.15.2