base-asr-with-tpbs-cv2

krirk/base-asr-with-tpbs-cv2

base-asr-with-tpbs-cv2 is a automatic speech recognition model from krirk. Use it when you need speech turned into text. It is set up for transformers.

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

Downloads · 30 days

1% of all-time downloads

All-time downloads

355

Public

Repo size

6.3 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin1.3 GB · 100%

At a glance

Task: Automatic Speech Recognition
Library: transformers
Model type: wav2vec2
Access: Public
Created: Apr 5, 2023
Updated: Apr 5, 2023
SHA: 3cd5c157

Model card

From the Hugging Face model README

base-asr-with-tpbs-cv2

This model is a fine-tuned version of Botnoi/base-asr-with-lm on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.5224
Wer: 0.2509
Cer: 0.0699
Clean Cer: 0.0628

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
gradient_accumulation_steps: 2
total_train_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 100
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer	Clean Cer
0.1319	26.67	400	0.4728	0.2699	0.0731	0.0656
0.0452	53.33	800	0.5010	0.2561	0.0703	0.0637
0.0293	80.0	1200	0.5224	0.2509	0.0699	0.0628

Framework versions

Transformers 4.27.4
Pytorch 1.13.1+cu117
Datasets 2.11.0
Tokenizers 0.13.2

Task: Automatic Speech Recognition
Library: transformers
Type: wav2vec2
Created: Apr 5, 2023
Updated: Apr 5, 2023