banglat5-finetuned-new-method

This model is a fine-tuned version of csebuetnlp/banglat5 on the None dataset. It achieves the following results on the evaluation set:

• Rouge1 Precision: 13.6496
• Rouge1 Recall: 27.8778
• Rouge1 Fmeasure: 17.8803
• Rouge2 Precision: 4.9428
• Rouge2 Recall: 10.9976
• Rouge2 Fmeasure: 6.6131
• Rougel Precision: 12.916
• Rougel Recall: 26.5923
• Rougel Fmeasure: 16.9585
• Gen Len: 18.6998
• Loss: 4.9511

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Rouge1 Precision	Rouge1 Recall	Rouge1 Fmeasure	Rouge2 Precision	Rouge2 Recall	Rouge2 Fmeasure	Rougel Precision	Rougel Recall	Rougel Fmeasure	Gen Len	Validation Loss
9.5004	1.0	202	9.8059	20.2097	12.8916	2.9133	6.3753	3.8695	9.092	18.8644	11.9714	19.0	6.5711
7.6455	2.0	404	11.168	22.9627	14.6726	3.8094	8.4291	5.0876	10.3227	21.5565	13.6235	19.0	5.7313
6.8425	3.0	606	12.4439	25.6372	16.361	4.3303	9.7531	5.8212	11.7523	24.4085	15.4884	18.9851	5.2068
6.6708	4.0	808	13.2142	27.0973	17.3176	4.8216	10.7206	6.442	12.4235	25.6883	16.3239	18.7965	5.0076
6.6577	5.0	1010	13.6496	27.8778	17.8803	4.9428	10.9976	6.6131	12.916	26.5923	16.9585	18.6998	4.9511

Framework versions

• Transformers 4.40.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1