README.md

metadata

tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: flan-t5-base-lamp-7t-finetuned-1
    results: []

flan-t5-base-lamp-7t-finetuned-1

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2121
• Rouge1: 0.5299
• Rouge2: 0.2730
• Rougel: 0.4776
• Rougelsum: 0.4930

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
7.0613	1.0	840	1.2605	0.5053	0.2462	0.4513	0.4657
1.3665	2.0	1680	1.2271	0.5124	0.2562	0.4600	0.4747
1.3009	3.0	2520	1.2114	0.5179	0.2613	0.4652	0.4802
1.2594	4.0	3360	1.2039	0.5229	0.2656	0.4677	0.4839
1.227	5.0	4200	1.1988	0.5247	0.2669	0.4698	0.4860
1.1918	6.0	5040	1.1966	0.5238	0.2698	0.4701	0.4860
1.1696	7.0	5880	1.1934	0.5261	0.2702	0.4723	0.4880
1.1285	8.0	6720	1.1933	0.5237	0.2693	0.4701	0.4849
1.1153	9.0	7560	1.1961	0.5263	0.2708	0.4724	0.4880
1.0927	10.0	8400	1.1961	0.5254	0.2691	0.4720	0.4874
1.0661	11.0	9240	1.2010	0.5234	0.2684	0.4698	0.4854
1.0634	12.0	10080	1.2003	0.5259	0.2723	0.4729	0.4885
1.046	13.0	10920	1.2019	0.5277	0.2726	0.4747	0.4907
1.0273	14.0	11760	1.2045	0.5309	0.2749	0.4776	0.4940
1.0218	15.0	12600	1.2077	0.5295	0.2728	0.4771	0.4925
1.0208	16.0	13440	1.2095	0.5303	0.2728	0.4775	0.4928
1.003	17.0	14280	1.2110	0.5301	0.2726	0.4772	0.4929
1.003	18.0	15120	1.2099	0.5314	0.2729	0.4781	0.4941
0.9893	19.0	15960	1.2115	0.5302	0.2737	0.4779	0.4933
0.9918	20.0	16800	1.2121	0.5299	0.2730	0.4776	0.4930

Framework versions

• Transformers 4.35.2
• Pytorch 2.0.1+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0