flan-t5-rouge-durga-q5-clean-4

This model is a fine-tuned version of google/flan-t5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0070
• Rouge1: 0.7173
• Rouge2: 0.6752
• Rougel: 0.7164
• Rougelsum: 0.7174

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: 0.0003
• train_batch_size: 12
• eval_batch_size: 12
• 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
1.9349	1.0	17	1.4016	0.2948	0.1024	0.2865	0.2864
1.7388	2.0	34	0.9216	0.3438	0.1391	0.3309	0.3314
0.9571	3.0	51	0.6340	0.3703	0.1739	0.3615	0.3621
0.8848	4.0	68	0.3948	0.4248	0.2516	0.4184	0.4196
0.6464	5.0	85	0.2382	0.4324	0.2759	0.4225	0.4232
0.3926	6.0	102	0.1711	0.4578	0.3070	0.4530	0.4534
0.3694	7.0	119	0.1300	0.4510	0.3250	0.4473	0.4485
0.2783	8.0	136	0.0952	0.4941	0.3828	0.4924	0.4926
0.2033	9.0	153	0.0825	0.5179	0.4106	0.5156	0.5147
0.1751	10.0	170	0.0788	0.4996	0.3951	0.4973	0.4977
0.1538	11.0	187	0.0589	0.5613	0.4773	0.5582	0.5592
0.2292	12.0	204	0.0544	0.5735	0.4866	0.5708	0.5720
0.1612	13.0	221	0.0437	0.5849	0.5058	0.5844	0.5851
0.0878	14.0	238	0.0388	0.6113	0.5346	0.6117	0.6120
0.0826	15.0	255	0.0403	0.6233	0.5515	0.6220	0.6219
0.0801	16.0	272	0.0290	0.6391	0.5769	0.6402	0.6390
0.1168	17.0	289	0.0242	0.6365	0.5697	0.6357	0.6373
0.0749	18.0	306	0.0242	0.6385	0.5713	0.6374	0.6388
0.0542	19.0	323	0.0175	0.6632	0.6103	0.6623	0.6636
0.0724	20.0	340	0.0154	0.6913	0.6393	0.6906	0.6922
0.0796	21.0	357	0.0177	0.6779	0.6248	0.6768	0.6777
0.0595	22.0	374	0.0116	0.7008	0.6612	0.7002	0.6995
0.0347	23.0	391	0.0135	0.6904	0.6442	0.6895	0.6895
0.0497	24.0	408	0.0110	0.6984	0.6488	0.6979	0.6982
0.0276	25.0	425	0.0103	0.7038	0.6581	0.7029	0.7038
0.0386	26.0	442	0.0087	0.7164	0.6773	0.7169	0.7161
0.0155	27.0	459	0.0084	0.7170	0.6787	0.7176	0.7174
0.0483	28.0	476	0.0077	0.7109	0.6657	0.7106	0.7114
0.0309	29.0	493	0.0071	0.7144	0.6706	0.7138	0.7152
0.0289	30.0	510	0.0070	0.7173	0.6752	0.7164	0.7174

Framework versions

• Transformers 4.46.0
• Pytorch 2.5.0+cu121
• Datasets 3.0.2
• Tokenizers 0.20.1