phi-3-vi-sft-1

This model is a fine-tuned version of microsoft/Phi-3-mini-4k-instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.0711

Model description

Fine-tuning Phi-3-4b with Vietnamese Domain, VI_LIMA and VI_Alpaca Dataset

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: 0.0002
• train_batch_size: 8
• eval_batch_size: 4
• seed: 3407
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 5
• num_epochs: 1
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.4031	0.17	40	1.2004
1.1508	0.34	80	1.1312
1.1055	0.51	120	1.1002
1.0814	0.67	160	1.0820
1.0735	0.84	200	1.0711

Framework versions

• PEFT 0.10.0
• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.16.0
• Tokenizers 0.15.2