phi3.5-mini-adapter_v0

This model is a fine-tuned version of microsoft/Phi-3.5-mini-instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0423

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: 0.0002
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• training_steps: 250

Training results

Training Loss	Epoch	Step	Validation Loss
11.9711	0.1509	10	10.5546
0.0847	0.3019	20	0.1104
0.0989	0.4528	30	0.0840
0.0543	0.6038	40	0.0588
0.0392	0.7547	50	0.0490
0.0447	0.9057	60	0.0457
0.0465	1.0566	70	0.0435
0.0317	1.2075	80	0.0445
0.0443	1.3585	90	0.0423

Framework versions

• PEFT 0.11.1
• Transformers 4.43.1
• Pytorch 2.4.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1