This model is a fine-tuned version of pyannote/speaker-diarization-3.1 on the speaker-segmentation dataset. It achieves the following results on the evaluation set:

Loss: 0.6644
Model Preparation Time: 0.0038
Der: 0.2459
False Alarm: 0.0831
Missed Detection: 0.0574
Confusion: 0.1054

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: 43
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

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Der	False Alarm	Missed Detection	Confusion
0.7188	1.0	244	0.7130	0.0038	0.2572	0.0828	0.0607	0.1138
0.677	2.0	488	0.6822	0.0038	0.2561	0.0828	0.0611	0.1122
0.6583	3.0	732	0.6737	0.0038	0.2499	0.0812	0.0598	0.1088
0.6407	4.0	976	0.6639	0.0038	0.2470	0.0829	0.0578	0.1063
0.6099	5.0	1220	0.6644	0.0038	0.2459	0.0831	0.0574	0.1054

Framework versions

Transformers 4.50.3
Pytorch 2.6.0+cu124
Datasets 3.5.0
Tokenizers 0.21.1

whitneyten/pydiarize-synthetic-data-Indonesia-200

speaker-segmentation-fine-tuned-id-dataset-id-HuggingFace

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

whitneyten/
pydiarize-synthetic-data-Indonesia-200