vit5-base-v2-masked-paraphase_data

This model is a fine-tuned version of VietAI/vit5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.0781

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.3523	0.4250	3500	1.2567
1.2491	0.8501	7000	1.1676
1.2107	1.2751	10500	1.1382
1.1861	1.7001	14000	1.1190
1.1516	2.1251	17500	1.1055
1.1482	2.5501	21000	1.0943
1.1702	2.9752	24500	1.0878
1.1442	3.4001	28000	1.0838
1.134	3.8252	31500	1.0803
1.1373	4.2502	35000	1.0787
1.1207	4.6752	38500	1.0781

Framework versions

• PEFT 0.10.0
• Transformers 4.49.0
• Pytorch 2.4.1+cu118
• Datasets 4.1.1
• Tokenizers 0.21.0