Wav2vec2-base-checkpoint-6 open-source speech recognition model - Practical, free, and accurate recognition of speech content

Wav2vec2 Base Checkpoint 6

Developed by jiobiala24

A speech recognition model fine-tuned on the Common Voice dataset based on wav2vec2-base-checkpoint-5

Speech Recognition

Transformers

Open Source License:Apache-2.0 #Speech Recognition #Fine-tuned Model #Low Word Error Rate

Downloads 16

Release Time : 3/2/2022

Model Overview

This model is a speech recognition model based on the wav2vec2 architecture, fine-tuned on the Common Voice dataset, primarily used for converting speech to text.

Model Features

Efficient Fine-tuning

Fine-tuned based on wav2vec2-base-checkpoint-5, optimized for performance on the Common Voice dataset

Low Word Error Rate

Achieved a word error rate (WER) of 0.3323 on the evaluation set

Model Capabilities

Speech Recognition

Audio to Text

Use Cases

Speech Transcription

Speech to Text

Convert speech content into editable text format

Word error rate 0.3323

🚀 wav2vec2-base-checkpoint-6

This model is a fine - tuned version of jiobiala24/wav2vec2-base-checkpoint-5 on the common_voice dataset, achieving specific results on the evaluation set.

🚀 Quick Start

This model is a fine - tuned version of jiobiala24/wav2vec2-base-checkpoint-5 on the common_voice dataset. It achieves the following results on the evaluation set:

Loss: 0.9738
Wer: 0.3323

📚 Documentation

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: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3435	1.82	1000	0.5637	0.3419
0.2599	3.65	2000	0.5804	0.3473
0.2043	5.47	3000	0.6481	0.3474
0.1651	7.3	4000	0.6937	0.3452
0.1376	9.12	5000	0.7221	0.3429
0.118	10.95	6000	0.7634	0.3441
0.105	12.77	7000	0.7789	0.3444
0.0925	14.6	8000	0.8209	0.3444
0.0863	16.42	9000	0.8293	0.3440
0.0756	18.25	10000	0.8553	0.3412
0.0718	20.07	11000	0.9006	0.3430
0.0654	21.9	12000	0.9541	0.3458
0.0605	23.72	13000	0.9400	0.3350
0.0552	25.55	14000	0.9547	0.3363
0.0543	27.37	15000	0.9715	0.3348
0.0493	29.2	16000	0.9738	0.3323