---
library_name: transformers
language:
- fr
license: mit
base_model: pyannote/segmentation-3.0
tags:
- speaker-diarization
- speaker-segmentation
- generated_from_trainer
datasets:
- CAENNAIS
model-index:
- name: pyannote/segmentation-3.0
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# pyannote/segmentation-3.0

This model is a fine-tuned version of [pyannote/segmentation-3.0](https://huggingface.co/pyannote/segmentation-3.0) on the CAENNAIS dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8139
- Model Preparation Time: 0.0035
- Der: 0.5111
- False Alarm: 0.1728
- Missed Detection: 0.2406
- Confusion: 0.0978

## 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: 10

### Training results

| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Der    | False Alarm | Missed Detection | Confusion |
|:-------------:|:-----:|:----:|:---------------:|:----------------------:|:------:|:-----------:|:----------------:|:---------:|
| 0.8517        | 1.0   | 300  | 0.8676          | 0.0035                 | 0.5466 | 0.1920      | 0.2425           | 0.1121    |
| 0.7998        | 2.0   | 600  | 0.8499          | 0.0035                 | 0.5307 | 0.1640      | 0.2628           | 0.1039    |
| 0.7867        | 3.0   | 900  | 0.8529          | 0.0035                 | 0.5366 | 0.1602      | 0.2767           | 0.0997    |
| 0.7777        | 4.0   | 1200 | 0.8351          | 0.0035                 | 0.5296 | 0.1912      | 0.2333           | 0.1050    |
| 0.7596        | 5.0   | 1500 | 0.8185          | 0.0035                 | 0.5118 | 0.1817      | 0.2239           | 0.1062    |
| 0.7591        | 6.0   | 1800 | 0.8083          | 0.0035                 | 0.5101 | 0.1655      | 0.2540           | 0.0906    |
| 0.7555        | 7.0   | 2100 | 0.8141          | 0.0035                 | 0.5109 | 0.1711      | 0.2396           | 0.1001    |
| 0.7394        | 8.0   | 2400 | 0.8145          | 0.0035                 | 0.5119 | 0.1726      | 0.2405           | 0.0988    |
| 0.7458        | 9.0   | 2700 | 0.8138          | 0.0035                 | 0.5107 | 0.1721      | 0.2403           | 0.0983    |
| 0.705         | 10.0  | 3000 | 0.8139          | 0.0035                 | 0.5111 | 0.1728      | 0.2406           | 0.0978    |


### Framework versions

- Transformers 4.51.0
- Pytorch 2.7.0+cu126
- Datasets 3.5.0
- Tokenizers 0.21.1