kingabzpro/wav2vec2-large-xls-r-1b-Indonesian

🤗 Hugging Face sourceautomatic-speech-recognitionapache-2.0963M params3.9 GBsafetensors✓ 3 checksumsupdated today

Submit in one command

Run it next to your model folder. It makes the torrent, checks your files against Hugging Face, and submits it. You just start seeding and paste your key from your account. It only reads your files and never changes them. Read the script first if you like.

curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo kingabzpro/wav2vec2-large-xls-r-1b-Indonesian ./model-folder

Needs a seeder →

wav2vec2-large-xls-r-1b-Indonesian

This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the common_voice dataset. It achieves the following results on the evaluation set:

Loss: 0.9550
Wer: 0.4551
Cer: 0.1643

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0003
train_batch_size: 64
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 400
num_epochs: 50
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
3.663	7.69	200	0.7898	0.6039	0.1848
0.7424	15.38	400	1.0215	0.5615	0.1924
0.4494	23.08	600	1.0901	0.5249	0.1932
0.5075	30.77	800	1.1013	0.5079	0.1935
0.4671	38.46	1000	1.1034	0.4916	0.1827
0.1928	46.15	1200	0.9550	0.4551	0.1643

Framework versions

Transformers 4.17.0.dev0
Pytorch 1.10.2+cu102
Datasets 1.18.2.dev0
Tokenizers 0.11.0