speaker-segmentation-fine-tuned-callhome-eng
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: 0.4607
- Der: 0.1815
- False Alarm: 0.0596
- Missed Detection: 0.0708
- Confusion: 0.0511
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 |
---|---|---|---|---|---|---|---|
0.426 | 1.0 | 362 | 0.4667 | 0.1875 | 0.0549 | 0.0784 | 0.0542 |
0.392 | 2.0 | 724 | 0.4678 | 0.1852 | 0.0594 | 0.0721 | 0.0536 |
0.3722 | 3.0 | 1086 | 0.4561 | 0.1801 | 0.0578 | 0.0714 | 0.0509 |
0.351 | 4.0 | 1448 | 0.4565 | 0.1810 | 0.0597 | 0.0699 | 0.0515 |
0.3493 | 5.0 | 1810 | 0.4607 | 0.1815 | 0.0596 | 0.0708 | 0.0511 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1
- Downloads last month
- 3
Model tree for KMayanja/speaker-segmentation-fine-tuned-callhome-eng
Base model
pyannote/segmentation-3.0
Finetuned
this model