metadata
library_name: transformers
language:
- spa
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-spa
results: []
speaker-segmentation-fine-tuned-callhome-spa
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: 0.5595
- Der: 0.2894
- False Alarm: 0.2353
- Missed Detection: 0.0536
- Confusion: 0.0005
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: 64
- 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 |
|---|---|---|---|---|---|---|---|
| 0.3614 | 1.0 | 226 | 0.4962 | 0.2910 | 0.2389 | 0.0520 | 0.0001 |
| 0.3465 | 2.0 | 452 | 0.5067 | 0.2860 | 0.2179 | 0.0679 | 0.0002 |
| 0.3325 | 3.0 | 678 | 0.5343 | 0.2941 | 0.2300 | 0.0636 | 0.0005 |
| 0.3189 | 4.0 | 904 | 0.5613 | 0.2906 | 0.2380 | 0.0522 | 0.0004 |
| 0.3238 | 5.0 | 1130 | 0.5595 | 0.2894 | 0.2353 | 0.0536 | 0.0005 |
Framework versions
- Transformers 4.45.1
- Pytorch 2.4.1
- Datasets 3.0.1
- Tokenizers 0.20.0