en_bn_summarize_v7

This model is a fine-tuned version of csebuetnlp/mT5_m2m_crossSum on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.8058
• Rouge-1: 18.1261
• Rouge-2: 6.4386
• Rouge-l: 15.755
• Gen Len: 43.3354

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge-1	Rouge-2	Rouge-l	Gen Len
1.8582	1.0	154	1.8089	17.2361	6.3031	15.1651	42.4348
1.6492	2.0	308	1.7993	16.9045	6.083	14.6343	41.472
1.6278	3.0	462	1.8006	16.909	6.1661	14.6043	43.4969
1.5656	4.0	616	1.8016	17.1664	6.3668	15.0702	42.1925
1.5456	5.0	770	1.7983	16.8696	5.9485	14.729	42.2298
1.5146	6.0	924	1.8060	17.2806	5.98	14.7861	43.3602
1.4575	7.0	1078	1.8024	17.6126	6.1446	15.1649	43.3665
1.4988	8.0	1232	1.8046	17.619	6.1422	15.1738	43.3913
1.4637	9.0	1386	1.8059	17.6713	6.2475	15.3152	44.0621
1.4593	10.0	1540	1.8058	18.1261	6.4386	15.755	43.3354

Framework versions

• Transformers 4.33.1
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3