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-eng-2
results: []
speaker-segmentation-fine-tuned-callhome-eng-2
This model is a fine-tuned version of pyannote/segmentation-3.0 on the diarizers-community/callhome eng dataset. It achieves the following results on the evaluation set:
- Loss: 0.4666
- Der: 0.1814
- False Alarm: 0.0552
- Missed Detection: 0.0739
- Confusion: 0.0523
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: 64
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Der | False Alarm | Missed Detection | Confusion |
|---|---|---|---|---|---|---|---|
| 0.4548 | 1.0 | 181 | 0.4943 | 0.1966 | 0.0564 | 0.0811 | 0.0590 |
| 0.4171 | 2.0 | 362 | 0.4845 | 0.1951 | 0.0644 | 0.0754 | 0.0552 |
| 0.396 | 3.0 | 543 | 0.4633 | 0.1856 | 0.0502 | 0.0825 | 0.0529 |
| 0.3856 | 4.0 | 724 | 0.4609 | 0.1843 | 0.0571 | 0.0739 | 0.0534 |
| 0.3693 | 5.0 | 905 | 0.4639 | 0.1821 | 0.0531 | 0.0761 | 0.0528 |
| 0.3634 | 6.0 | 1086 | 0.4610 | 0.1821 | 0.0588 | 0.0716 | 0.0517 |
| 0.3655 | 7.0 | 1267 | 0.4638 | 0.1827 | 0.0566 | 0.0740 | 0.0521 |
| 0.3608 | 8.0 | 1448 | 0.4603 | 0.1814 | 0.0567 | 0.0732 | 0.0515 |
| 0.3545 | 9.0 | 1629 | 0.4645 | 0.1805 | 0.0530 | 0.0761 | 0.0514 |
| 0.3508 | 10.0 | 1810 | 0.4666 | 0.1814 | 0.0552 | 0.0739 | 0.0523 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.2.0+cu121
- Datasets 2.17.0
- Tokenizers 0.19.1