metadata
library_name: transformers
license: mit
base_model: pyannote/segmentation-3.0
tags:
- speaker-diarization
- speaker-segmentation
- generated_from_trainer
datasets:
- diarizers-community/simsamu
model-index:
- name: speaker-segmentation-simsamu-fra
results: []
speaker-segmentation-simsamu-fra
This model is a fine-tuned version of pyannote/segmentation-3.0 on the diarizers-community/simsamu dataset. It achieves the following results on the evaluation set:
- Loss: 0.2243
- Model Preparation Time: 0.0033
- Der: 0.0906
- False Alarm: 0.0239
- Missed Detection: 0.0427
- Confusion: 0.0240
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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- num_epochs: 5.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Der | False Alarm | Missed Detection | Confusion |
|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 56 | 0.2303 | 0.0033 | 0.0985 | 0.0298 | 0.0430 | 0.0257 |
| 0.2116 | 2.0 | 112 | 0.2301 | 0.0033 | 0.0968 | 0.0218 | 0.0524 | 0.0227 |
| 0.2116 | 3.0 | 168 | 0.2247 | 0.0033 | 0.0923 | 0.0230 | 0.0462 | 0.0231 |
| 0.1681 | 4.0 | 224 | 0.2244 | 0.0033 | 0.0909 | 0.0246 | 0.0424 | 0.0240 |
| 0.1681 | 5.0 | 280 | 0.2243 | 0.0033 | 0.0906 | 0.0239 | 0.0427 | 0.0240 |
Framework versions
- Transformers 4.52.3
- Pytorch 2.6.0+cu126
- Datasets 3.6.0
- Tokenizers 0.21.1