w2v2-libri-10min

GeorgeZaf/w2v2-libri-10min

w2v2-libri-10min is a automatic speech recognition model from GeorgeZaf. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

15% of all-time downloads

All-time downloads

Public

Repo size

3.4 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin378 MB · 100%

At a glance

Task: Automatic Speech Recognition
Library: transformers
License: apache-2.0
Model type: wav2vec2
Access: Public
Created: May 20, 2023
Updated: May 20, 2023
SHA: 7903d78a

Model card

From the Hugging Face model README

w2v2-libri-10min

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: 2.0577
Wer: 0.6432

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: 0.0003
train_batch_size: 16
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: 500
training_steps: 2000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
5.5554	62.5	250	2.9713	1.0
2.5095	125.0	500	1.3802	0.7344
0.1558	187.5	750	1.8509	0.6874
0.0603	250.0	1000	1.8335	0.6584
0.0255	312.5	1250	2.2245	0.6943
0.012	375.0	1500	1.9622	0.6473
0.0063	437.5	1750	2.0621	0.6418
0.0045	500.0	2000	2.0577	0.6432

Framework versions

Transformers 4.29.2
Pytorch 2.0.1+cu118
Datasets 1.18.3
Tokenizers 0.13.3

Task: Automatic Speech Recognition
Library: transformers
Type: wav2vec2
License: apache-2.0
Created: May 20, 2023
Updated: May 20, 2023