flan-t5-base-samsum

This model is a fine-tuned version of google/flan-t5-base on the samsum dataset. It achieves the following results on the evaluation set:

• Loss: 1.3716
• Rouge1: 47.3683
• Rouge2: 24.0343
• Rougel: 39.9874
• Rougelsum: 43.6453
• Gen Len: 17.3004

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
• 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: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.4531	1.0	1842	1.3836	46.4391	23.0513	39.1448	42.8774	17.1868
1.3433	2.0	3684	1.3729	47.0465	23.4504	39.8361	43.3316	17.2613
1.2767	3.0	5526	1.3716	47.3683	24.0343	39.9874	43.6453	17.3004
1.2117	4.0	7368	1.3739	47.6321	24.1445	40.3378	43.9123	17.0989
1.1622	5.0	9210	1.3826	47.6786	23.9568	40.2743	43.7625	17.0879
1.1387	6.0	11052	1.3920	47.6434	24.0265	40.2093	43.9179	17.3712
1.1011	7.0	12894	1.3947	47.6658	24.0395	40.3477	43.8425	17.2186
1.0755	8.0	14736	1.4059	47.5613	24.0555	40.181	43.7645	17.1490
1.0514	9.0	16578	1.4053	47.9552	24.2395	40.3731	44.0694	17.3602
1.0311	10.0	18420	1.4114	48.0582	24.3022	40.4713	44.1136	17.3175

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.3
• Tokenizers 0.13.3