BeyzaAkyildiz's picture
End of training
a42c6d2 verified
metadata
language:
  - jpn
license: mit
base_model: pyannote/speaker-diarization-3.1
tags:
  - speaker-diarization
  - speaker-segmentation
  - generated_from_trainer
datasets:
  - diarizers-community/callhome
model-index:
  - name: speaker-segmentation-fine-tuned-callhome-jpn
    results: []

speaker-segmentation-fine-tuned-callhome-jpn

This model is a fine-tuned version of pyannote/speaker-diarization-3.1 on the diarizers-community/callhome dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4585
  • Der: 0.1815
  • False Alarm: 0.0615
  • Missed Detection: 0.0694
  • Confusion: 0.0506

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: 0.001
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • num_epochs: 5

Training results

Training Loss Epoch Step Validation Loss Der False Alarm Missed Detection Confusion
0.3855 1.0 362 0.4769 0.1895 0.0554 0.0764 0.0577
0.3977 2.0 724 0.4610 0.1879 0.0668 0.0693 0.0518
0.3778 3.0 1086 0.4577 0.1805 0.0597 0.0703 0.0505
0.3558 4.0 1448 0.4600 0.1812 0.0606 0.0703 0.0503
0.3335 5.0 1810 0.4585 0.1815 0.0615 0.0694 0.0506

Framework versions

  • Transformers 4.42.4
  • Pytorch 2.3.1+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1