Hugging Face's logo

juanquivilla
/
phi-3-mini-NER-PII-Vast3

Token Classification
Transformers
Safetensors
phi3
Generated from Trainer
custom_code
text-generation-inference
Inference Endpoints
Model card Files Community
phi-3-mini-NER-PII-Vast3 / README.md
juanquivilla's picture
juanquivilla
Model save
d0e46ec verified
preview code
|
raw
history blame
1.82 kB
metadata
license: mit
base_model: microsoft/Phi-3-mini-4k-instruct
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: phi-3-mini-NER-PII-Vast3
    results: []

phi-3-mini-NER-PII-Vast3

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: 0.1122
  • Precision: 0.6826
  • Recall: 0.8382
  • F1: 0.7524
  • Accuracy: 0.9697

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: 2e-05
  • 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
  • num_epochs: 3

Training results

Training Loss Epoch Step Accuracy F1 Validation Loss Precision Recall
0.1017 1.0 11105 0.9693 0.7506 0.1123 0.6807 0.8364
0.0782 2.0 22210 0.1119 0.6819 0.8382 0.7520 0.9697
0.0944 3.0 33315 0.1122 0.6826 0.8382 0.7524 0.9697

Framework versions

  • Transformers 4.41.1
  • Pytorch 2.3.0
  • Datasets 2.19.1
  • Tokenizers 0.19.1