flan-t5-base-samsum_model_3

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

• Loss: 1.4529
• Rouge1: 45.6393
• Rouge2: 28.4817
• Rougel: 42.2911
• Rougelsum: 43.1196
• Gen Len: 18.1892

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	33	2.0162	37.0638	19.7336	33.7411	34.3994	17.3604
No log	2.0	66	1.8876	39.4107	21.2853	36.2051	36.9608	17.7658
No log	3.0	99	1.7884	40.6159	22.7651	37.4066	38.0497	18.1171
No log	4.0	132	1.7105	42.3741	25.2422	39.2729	40.1794	18.1982
No log	5.0	165	1.6360	42.1734	25.0239	39.1515	40.0452	18.3874
No log	6.0	198	1.5971	43.8499	27.134	40.6895	41.5456	18.3243
No log	7.0	231	1.5776	44.414	27.7411	41.4884	42.2697	18.1081
No log	8.0	264	1.5445	44.6134	27.3292	41.4632	42.163	18.1892
No log	9.0	297	1.5247	45.5293	28.4399	42.2776	42.9934	18.0360
No log	10.0	330	1.4952	45.9019	29.8113	43.073	43.6686	18.0811
No log	11.0	363	1.4871	45.5634	28.1388	41.8117	42.447	18.3063
No log	12.0	396	1.4604	45.6571	28.4288	42.3541	42.8206	18.3333
No log	13.0	429	1.4595	46.4867	29.3814	43.219	43.8416	18.2523
No log	14.0	462	1.4702	45.4699	28.0184	42.1961	42.9143	18.4324
No log	15.0	495	1.4572	45.7874	28.7692	42.5542	43.2001	18.3604
1.2802	16.0	528	1.4535	46.2596	29.1805	42.8517	43.5507	18.2342
1.2802	17.0	561	1.4529	45.6393	28.4817	42.2911	43.1196	18.1892
1.2802	18.0	594	1.4574	45.5262	28.4542	42.3652	43.1119	18.2162
1.2802	19.0	627	1.4586	46.3006	29.1285	43.1952	43.8849	18.0811
1.2802	20.0	660	1.4712	45.9096	28.4353	42.7121	43.427	18.0631

Framework versions

• Transformers 4.41.0
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1