wav2vec2-8 – AI Model by chrisvinsen | AlphaNeural AI

You can deploy this model and start earning money today!

chrisvinsen/
wav2vec2-8

like

0

automatic-speech-recognition

generated_from_trainer

endpoints_compatible

Views

No views yet

wav2vec2-8

This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:

Loss: 3.1169
Wer: 1.0

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: 0.0006
train_batch_size: 32
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_steps: 400
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
3.9398	1.56	200	3.1250	1.0
2.8703	3.12	400	3.1608	1.0
2.8632	4.69	600	3.1329	1.0
2.8638	6.25	800	3.0795	1.0
2.8595	7.81	1000	3.1410	1.0
2.8611	9.38	1200	3.0952	1.0
2.861	10.94	1400	3.1391	1.0
2.8603	12.5	1600	3.0639	1.0
2.8568	14.06	1800	3.1180	1.0
2.8563	15.62	2000	3.1170	1.0
2.857	17.19	2200	3.0846	1.0
2.8574	18.75	2400	3.0740	1.0
2.8543	20.31	2600	3.1482	1.0
2.8567	21.88	2800	3.1604	1.0
2.8561	23.44	3000	3.1055	1.0
2.858	25.0	3200	3.0669	1.0
2.8524	26.56	3400	3.0992	1.0
2.8557	28.12	3600	3.1050	1.0
2.8527	29.69	3800	3.1169	1.0

Framework versions

Transformers 4.19.2
Pytorch 1.11.0+cu113
Datasets 2.2.2
Tokenizers 0.12.1