metadata
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
datasets:
- superb
metrics:
- accuracy
- recall
- f1
model-index:
- name: wav2vec2-base-finetuned-ks
results:
- task:
name: Audio Classification
type: audio-classification
dataset:
name: superb
type: superb
config: ks
split: validation
args: ks
metrics:
- name: Accuracy
type: accuracy
value: 0.9832303618711385
- name: Recall
type: recall
value: 0.9664413018718482
- name: F1
type: f1
value: 0.9719648106690262
wav2vec2-base-finetuned-ks
This model is a fine-tuned version of facebook/wav2vec2-base on the superb dataset. It achieves the following results on the evaluation set:
- Loss: 0.0711
- Accuracy: 0.9832
- Recall: 0.9664
- F1: 0.9720
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: 3e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Recall | F1 |
|---|---|---|---|---|---|---|
| 0.4397 | 1.0 | 798 | 0.2810 | 0.9651 | 0.9289 | 0.9361 |
| 0.2067 | 2.0 | 1597 | 0.1142 | 0.9769 | 0.9536 | 0.9593 |
| 0.1881 | 3.0 | 2395 | 0.0829 | 0.9821 | 0.9644 | 0.9693 |
| 0.1167 | 4.0 | 3194 | 0.0752 | 0.9831 | 0.9644 | 0.9726 |
| 0.13 | 5.0 | 3990 | 0.0711 | 0.9832 | 0.9664 | 0.9720 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu121
- Datasets 2.14.6
- Tokenizers 0.14.1