flan-t5-small

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0344
• Rouge1: 69.6097
• Rouge2: 38.3871
• Rougel: 69.5382
• Rougelsum: 69.6105
• Gen Len: 7.5996

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• 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	185	0.9477	71.6359	39.5346	71.4741	71.4583	7.7331
No log	2.0	370	0.9463	71.0852	39.5355	71.0038	70.978	7.6489
0.2285	3.0	555	0.9467	71.5319	40.2552	71.4544	71.4181	7.7351
0.2285	4.0	740	0.9539	70.4693	40.232	70.4544	70.4556	7.7823
0.2285	5.0	925	0.9325	70.7711	39.3649	70.6562	70.681	7.6057
0.2253	6.0	1110	0.9316	70.9811	39.6512	70.8691	70.8853	7.6222
0.2253	7.0	1295	0.9532	70.8027	40.185	70.7144	70.7021	7.7556
0.2253	8.0	1480	0.9502	71.0405	39.34	70.9893	71.0025	7.7105
0.2115	9.0	1665	0.9828	70.9613	39.5827	70.9241	70.9235	7.8234
0.2115	10.0	1850	0.9547	70.9949	39.0672	70.9072	70.9983	7.5873
0.2078	11.0	2035	0.9735	70.5543	39.3234	70.4865	70.4698	7.6509
0.2078	12.0	2220	0.9807	70.9333	39.7692	70.8078	70.8372	7.616
0.2078	13.0	2405	0.9802	71.2548	40.1212	71.1681	71.1966	7.6797
0.197	14.0	2590	0.9810	70.9522	39.7054	70.8424	70.8882	7.7269
0.197	15.0	2775	0.9727	71.0681	39.4757	70.9472	70.9828	7.5934
0.197	16.0	2960	0.9610	71.0263	38.9423	71.0483	71.0445	7.4764
0.189	17.0	3145	1.0003	71.0454	40.407	70.929	71.0031	7.7844
0.189	18.0	3330	0.9698	70.178	39.1381	70.0794	70.042	7.5318
0.1828	19.0	3515	0.9792	70.593	39.3776	70.4514	70.4502	7.5339
0.1828	20.0	3700	0.9918	70.8841	39.9677	70.745	70.7878	7.807
0.1828	21.0	3885	1.0140	70.656	39.7687	70.5639	70.5805	7.7823
0.1823	22.0	4070	0.9897	70.9188	39.4871	70.8637	70.828	7.614
0.1823	23.0	4255	0.9735	71.1061	39.3874	70.9184	70.9809	7.5339
0.1823	24.0	4440	0.9961	70.6611	39.253	70.4768	70.5304	7.6407
0.1778	25.0	4625	1.0104	70.4577	39.1874	70.3119	70.3523	7.6591
0.1778	26.0	4810	0.9870	70.5865	38.787	70.42	70.4843	7.4743
0.1778	27.0	4995	1.0201	70.5721	39.5023	70.3954	70.442	7.6838
0.1704	28.0	5180	1.0246	70.6781	39.2569	70.4955	70.5349	7.6735
0.1704	29.0	5365	1.0361	70.1736	38.7315	70.031	70.0288	7.7166
0.1703	30.0	5550	1.0163	70.3424	38.8374	70.2202	70.2418	7.614
0.1703	31.0	5735	1.0410	70.2047	39.0462	70.0922	70.0899	7.7515
0.1703	32.0	5920	1.0262	70.4281	39.1786	70.3153	70.365	7.6078
0.1625	33.0	6105	1.0175	70.2575	38.9449	70.1802	70.2177	7.5195
0.1625	34.0	6290	1.0152	70.2141	38.7015	70.1391	70.1653	7.5914
0.1625	35.0	6475	1.0340	70.0245	38.8692	69.9427	69.9565	7.5893
0.1645	36.0	6660	1.0378	70.0414	38.8613	69.956	70.0099	7.6366
0.1645	37.0	6845	1.0362	69.9449	38.8858	69.9163	69.9452	7.6427
0.1628	38.0	7030	1.0434	69.8905	38.8858	69.8392	69.8622	7.6304
0.1628	39.0	7215	1.0365	70.0064	39.0114	69.957	70.0052	7.616
0.1628	40.0	7400	1.0357	69.9773	38.9875	69.927	69.979	7.6222
0.1554	41.0	7585	1.0391	70.0719	39.0349	70.031	70.0612	7.653
0.1554	42.0	7770	1.0270	69.8766	38.6497	69.8359	69.8445	7.6119
0.1554	43.0	7955	1.0327	69.7936	38.6497	69.7422	69.7614	7.6201
0.1567	44.0	8140	1.0341	69.7648	38.5709	69.7178	69.7306	7.6366
0.1567	45.0	8325	1.0347	69.734	38.4414	69.6589	69.7016	7.616
0.1545	46.0	8510	1.0371	69.5376	38.2615	69.4623	69.5408	7.6283
0.1545	47.0	8695	1.0402	69.5376	38.2615	69.4623	69.5408	7.6283
0.1545	48.0	8880	1.0363	69.5376	38.2615	69.4623	69.5408	7.6283
0.1575	49.0	9065	1.0337	69.6097	38.3871	69.5382	69.6105	7.5996
0.1575	50.0	9250	1.0344	69.6097	38.3871	69.5382	69.6105	7.5996

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3