phi-3-mini-LoRA

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: 0.6528

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• 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
0.7804	0.2203	100	0.6979
0.6811	0.4405	200	0.6706
0.6681	0.6608	300	0.6644
0.6622	0.8811	400	0.6613
0.6602	1.1013	500	0.6592
0.6581	1.3216	600	0.6576
0.6564	1.5419	700	0.6563
0.6557	1.7621	800	0.6553
0.6541	1.9824	900	0.6545
0.6531	2.2026	1000	0.6540
0.6506	2.4229	1100	0.6534
0.651	2.6432	1200	0.6530
0.6512	2.8634	1300	0.6528

Framework versions

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