Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
wav2vec2-demo – AI Model by M-Quan | AlphaNeural AI
You can deploy this model and start earning money today!
M-Quan
/
wav2vec2-demo
like
0
transformers
pytorch
tensorboard
wav2vec2
automatic-speech-recognition
generated_from_trainer
apache-2.0
endpoints_compatible
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
wav2vec2-demo
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: 0.4239
Wer: 0.3508
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: 0.0001
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: 1000
num_epochs: 20
mixed_precision_training: Native AMP
Training results
Training Loss
Epoch
Step
Validation Loss
Wer
3.4093
4.0
500
1.2405
0.8685
0.5597
8.0
1000
0.4538
0.4437
0.2113
12.0
1500
0.4106
0.3749
0.1188
16.0
2000
0.4609
0.3775
0.0776
20.0
2500
0.4239
0.3508
Framework versions
Transformers 4.11.3
Pytorch 1.10.0+cu111
Datasets 1.18.4
Tokenizers 0.10.3
