phi-3-mini-LoRA-MEDQA-V3

This model is a fine-tuned version of microsoft/Phi-3-mini-4k-instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6086

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• 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_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.6898	0.3633	100	0.6195
0.6113	0.7266	200	0.6134
0.6095	1.0899	300	0.6110
0.6034	1.4532	400	0.6103
0.6037	1.8165	500	0.6093
0.6043	2.1798	600	0.6089
0.5986	2.5431	700	0.6089
0.5993	2.9064	800	0.6086

Framework versions

• PEFT 0.12.0
• Transformers 4.43.3
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1