metadata

license: apache-2.0
base_model: facebook/hubert-base-ls960
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - recall
  - precision
model-index:
  - name: hubert-base-ls960-finetuned-common_voice
    results: []

hubert-base-ls960-finetuned-common_voice

This model is a fine-tuned version of facebook/hubert-base-ls960 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.2002
Accuracy: 0.955
F1: 0.9549
Recall: 0.9550
Precision: 0.9551
Mcc: 0.9438
Auc: 0.9942

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: 1e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Recall	Precision	Mcc	Auc
1.5544	1.0	200	1.5193	0.405	0.3628	0.4050	0.5940	0.2904	0.8407
1.1406	2.0	400	0.9811	0.6375	0.5780	0.6375	0.6712	0.5734	0.9464
0.7902	3.0	600	0.6775	0.8125	0.7969	0.8125	0.8181	0.7740	0.9724
0.5346	4.0	800	0.5083	0.8725	0.8683	0.8725	0.8774	0.8438	0.9834
0.5139	5.0	1000	0.3943	0.9025	0.8988	0.9025	0.9074	0.8809	0.9879
0.5136	6.0	1200	0.3314	0.915	0.9145	0.915	0.9174	0.8945	0.9881
0.3726	7.0	1400	0.2894	0.925	0.9241	0.925	0.9258	0.9069	0.9878
0.3072	8.0	1600	0.2267	0.9325	0.9314	0.9325	0.9349	0.9167	0.9914
0.1948	9.0	1800	0.2117	0.945	0.9445	0.945	0.9461	0.9317	0.9931
0.2312	10.0	2000	0.2002	0.955	0.9549	0.9550	0.9551	0.9438	0.9942

Framework versions

Transformers 4.41.0
Pytorch 2.3.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1