metadata
license: mit
base_model: pyannote/segmentation-3.0
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/segmentation-3.0 on the diarizers-community/callhome jpn dataset. It achieves the following results on the evaluation set:
- Loss: 0.4828
- Der: 0.1446
- False Alarm: 0.0404
- Missed Detection: 0.0606
- Confusion: 0.0435
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: 5.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Der | False Alarm | Missed Detection | Confusion |
|---|---|---|---|---|---|---|---|
| 0.5955 | 1.0 | 394 | 0.5366 | 0.1609 | 0.0435 | 0.0706 | 0.0468 |
| 0.5648 | 2.0 | 788 | 0.4979 | 0.1509 | 0.0400 | 0.0646 | 0.0462 |
| 0.5392 | 3.0 | 1182 | 0.4852 | 0.1489 | 0.0447 | 0.0588 | 0.0453 |
| 0.5283 | 4.0 | 1576 | 0.4756 | 0.1442 | 0.0412 | 0.0607 | 0.0422 |
| 0.5109 | 5.0 | 1970 | 0.4828 | 0.1446 | 0.0404 | 0.0606 | 0.0435 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.2.2+cu118
- Datasets 2.18.0
- Tokenizers 0.19.1