vit5-base-skill-extraction-lora-ver2

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: 1.0795

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: 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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.2896	0.5051	300	1.9677
1.8529	1.0101	600	1.6651
1.6366	1.5152	900	1.4975
1.5251	2.0202	1200	1.3960
1.4233	2.5253	1500	1.3370
1.35	3.0303	1800	1.2921
1.2733	3.5354	2100	1.2446
1.2032	4.0404	2400	1.2183
1.1618	4.5455	2700	1.1813
1.1292	5.0505	3000	1.1564
1.0818	5.5556	3300	1.1442
1.021	6.0606	3600	1.1281
0.9924	6.5657	3900	1.1033
0.9653	7.0707	4200	1.0956
0.9468	7.5758	4500	1.0899
0.9299	8.0808	4800	1.0839
0.8824	8.5859	5100	1.0800
0.8931	9.0909	5400	1.0817
0.9111	9.5960	5700	1.0795

Framework versions

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