Hugging Face's logo

enriquesaou
/
phi-3-mrqa

PEFT
Safetensors
Generated from Trainer
Model card Files Community
phi-3-mrqa / README.md
enriquesaou's picture
enriquesaou
Update README.md
c589401 verified
preview code
|
raw
history blame
1.78 kB
metadata
license: mit
library_name: peft
tags:
  - generated_from_trainer
base_model: microsoft/Phi-3-mini-4k-instruct
model-index:
  - name: phi-3-mrqa
    results: []
datasets:
  - enriquesaou/mrqa-squadded-sample

Visualize in Weights & Biases

phi-3-mrqa

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

  • Loss: 1.7651

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: 3e-05
  • train_batch_size: 4
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.03
  • training_steps: 300
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
1.7895 0.0267 100 1.8179
1.7531 0.0533 200 1.7730
1.7567 0.08 300 1.7651

Framework versions

  • PEFT 0.11.2.dev0
  • Transformers 4.42.0.dev0
  • Pytorch 2.3.0+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1