File size: 634 Bytes
4f961c8 1b26b69 4f961c8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 |
---
metrics:
- accuracy
pipeline_tag: audio-classification
tags:
- music
---
This model is finetuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) for music detection.
# This model was finetuned to distinguish between music and non-music audio.
# Training hyperparameters
The following hyperparameters were used during training:
- metric_for_best_model = accuracy
- learning_rate = 5e-5
- seed = 42
- per_device_train_batch_size = 4
- per_device_eval_batch_size = 4
- gradient_accumulation_steps = 1
- warmup_ratio = 0.1
- fp16 = True
- adam_epsilon = 1e-08
- adam_beta1 = 0.9
- adam_beta2 = 0.999 |