metadata
library_name: transformers
language:
- hi
license: mit
base_model: pyannote/speaker-diarization-3.1
tags:
- speaker-diarization
- speaker-segmentation
- generated_from_trainer
datasets:
- Samyak29/synthetic-speaker-diarization-dataset-hindi-large
model-index:
- name: speaker-segmentation-fine-tuned-hindi
results: []
speaker-segmentation-fine-tuned-hindi
This model is a fine-tuned version of pyannote/speaker-diarization-3.1 on the Samyak29/synthetic-speaker-diarization-dataset-hindi-large dataset. It achieves the following results on the evaluation set:
- Loss: 0.3905
- Model Preparation Time: 0.0039
- Der: 0.1286
- False Alarm: 0.0227
- Missed Detection: 0.0270
- Confusion: 0.0790
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 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 |
|---|---|---|---|---|---|---|---|---|
| 0.4526 | 1.0 | 219 | 0.4401 | 0.0039 | 0.1423 | 0.0261 | 0.0297 | 0.0864 |
| 0.4004 | 2.0 | 438 | 0.4090 | 0.0039 | 0.1334 | 0.0228 | 0.0297 | 0.0810 |
| 0.3571 | 3.0 | 657 | 0.3891 | 0.0039 | 0.1249 | 0.0224 | 0.0273 | 0.0752 |
| 0.3497 | 4.0 | 876 | 0.3877 | 0.0039 | 0.1269 | 0.0238 | 0.0264 | 0.0767 |
| 0.3609 | 5.0 | 1095 | 0.3905 | 0.0039 | 0.1286 | 0.0227 | 0.0270 | 0.0790 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0