t5-base-finetuned-xsum

This model is a fine-tuned version of t5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.1894
• Rouge1: 34.6512
• Rouge2: 19.303
• Rougel: 32.5996
• Rougelsum: 32.471
• Gen Len: 7.0

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
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	13	2.5904	31.7728	17.9506	31.6663	32.1172	7.2609
No log	2.0	26	2.5238	33.4485	17.2356	33.0828	33.5748	7.8696
No log	3.0	39	2.4739	31.3343	15.996	31.2341	31.646	9.0435
No log	4.0	52	2.4413	31.2481	16.395	31.2318	31.633	8.0435
No log	5.0	65	2.4156	28.9275	13.9479	28.7067	29.3501	8.2609
No log	6.0	78	2.3768	25.1989	12.2726	25.1301	25.331	9.6087
No log	7.0	91	2.3599	25.7646	12.6915	25.745	26.2307	9.0
No log	8.0	104	2.3460	27.3509	12.7398	27.3198	27.7394	8.3043
No log	9.0	117	2.3134	26.4175	12.3464	26.3116	26.6575	8.2609
No log	10.0	130	2.3089	25.4818	12.381	25.4929	25.9985	7.913
No log	11.0	143	2.2976	27.5605	13.3195	27.4523	27.7723	7.0
No log	12.0	156	2.2951	29.0566	13.3195	29.1363	29.4032	7.0
No log	13.0	169	2.2770	27.4586	13.2436	27.3583	27.7153	7.1304
No log	14.0	182	2.2675	27.4586	13.2436	27.3583	27.7153	7.1304
No log	15.0	195	2.2573	27.4586	13.2436	27.3583	27.7153	7.0
No log	16.0	208	2.2523	28.0801	13.2436	28.0208	28.4869	6.6957
No log	17.0	221	2.2410	29.7194	13.9268	29.6547	29.9577	6.7826
No log	18.0	234	2.2355	29.8201	13.9268	29.7716	30.1114	7.0
No log	19.0	247	2.2350	28.4851	13.6025	28.3633	28.6581	7.0435
No log	20.0	260	2.2219	26.4416	12.7674	25.9806	26.2749	6.6522
No log	21.0	273	2.2036	26.6032	13.0124	26.25	26.4627	6.8261
No log	22.0	286	2.1974	27.4656	13.0124	27.2275	27.2769	6.2174
No log	23.0	299	2.1962	26.4416	12.7674	25.9806	26.2749	6.6957
No log	24.0	312	2.2099	26.1686	12.6018	25.7118	25.9094	6.7391
No log	25.0	325	2.1990	27.3084	12.7709	27.0236	27.0354	6.1739
No log	26.0	338	2.1942	29.0825	12.7709	28.6736	29.3337	6.3478
No log	27.0	351	2.2058	29.0825	12.7709	27.5837	28.1575	6.0
No log	28.0	364	2.2012	29.0825	12.7709	27.5837	28.1575	6.0
No log	29.0	377	2.1992	28.1338	12.7709	26.6664	27.0659	6.8696
No log	30.0	390	2.1840	32.5399	17.1325	31.2654	31.4223	6.6087
No log	31.0	403	2.1824	32.5399	17.1325	31.2654	31.4223	6.7826
No log	32.0	416	2.1830	34.6512	19.303	33.1484	33.1392	6.8261
No log	33.0	429	2.1846	33.0599	19.303	31.3736	31.5588	6.6522
No log	34.0	442	2.1868	33.0599	19.303	31.3736	31.5588	6.6522
No log	35.0	455	2.1803	35.5538	19.303	34.135	34.0635	6.087
No log	36.0	468	2.1779	35.5538	19.303	33.5533	33.5085	6.087
No log	37.0	481	2.1770	34.9683	19.303	33.2356	33.1109	6.1739
No log	38.0	494	2.1845	35.5538	19.303	33.5533	33.5085	6.3478
1.8275	39.0	507	2.1867	34.6512	19.303	32.5996	32.471	7.0
1.8275	40.0	520	2.1881	36.4717	19.7895	34.9234	34.7549	6.913
1.8275	41.0	533	2.1877	36.4717	19.7895	34.9234	34.7549	6.913
1.8275	42.0	546	2.1842	36.4717	19.7895	34.9234	34.7549	6.913
1.8275	43.0	559	2.1869	36.4717	19.7895	34.3175	34.1247	6.913
1.8275	44.0	572	2.1914	36.4717	19.7895	34.3175	34.1247	6.913
1.8275	45.0	585	2.1921	36.4717	19.7895	34.3175	34.1247	6.913
1.8275	46.0	598	2.1910	36.4717	19.7895	34.3175	34.1247	6.913
1.8275	47.0	611	2.1903	34.6512	19.303	32.5996	32.471	7.0
1.8275	48.0	624	2.1904	34.6512	19.303	32.5996	32.471	7.0
1.8275	49.0	637	2.1896	34.6512	19.303	32.5996	32.471	7.0
1.8275	50.0	650	2.1894	34.6512	19.303	32.5996	32.471	7.0

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.13.0
• Tokenizers 0.13.3