w2v2-base-pretrained_lr1e-4_at0.5_da1 – AI Model by MelanieKoe | AlphaNeural AI

You can deploy this model and start earning money today!

MelanieKoe/
w2v2-base-pretrained_lr1e-4_at0.5_da1

like

0

automatic-speech-recognition

generated_from_trainer

facebook/wav2vec2-base

endpoints_compatible

Views

No views yet

w2v2-base-pretrained_lr1e-4_at0.5_da1

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: 5.7414
Wer: 1.0541

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: 60
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
23.8591	4.17	250	22.7534	1.0
3.9846	8.33	500	3.8634	1.0
3.1237	12.5	750	3.8990	1.0
2.0534	16.67	1000	2.5944	1.0469
0.3404	20.83	1250	3.4531	1.0554
0.1803	25.0	1500	3.5649	1.0358
0.0986	29.17	1750	4.4418	1.0473
0.0664	33.33	2000	4.8448	1.0443
0.05	37.5	2250	5.1965	1.0575
0.0382	41.67	2500	5.3657	1.0490
0.0278	45.83	2750	5.4654	1.0563
0.0242	50.0	3000	5.5346	1.0554
0.0196	54.17	3250	5.6232	1.0546
0.0181	58.33	3500	5.7414	1.0541

Framework versions

Transformers 4.35.0
Pytorch 2.0.0
Datasets 2.14.6
Tokenizers 0.14.1