Model
wav2vec 2.0 base PyTorch checkpoint pre-trained on LibriSpeech 960 h, fine-tuned on LibriSpeech 960 h
Use the NGC CLI to download:
Copied!
1 Version
22.09.0_amp-bf16Selected01/16/2023 11:45 AM UTC915.49 MBAccuracy: 00 EpochsBatch Size: 0GPU: A100 Copied!
22.09.0_amp-bf16Selected
01/16/2023 11:45 AM UTC915.49 MBAccuracy: 00 EpochsBatch Size: 0GPU: A100
Copied!
architecture
| Key | Value |
|---|---|
| task | supervised fine-tuning |
| hourglass_transformer_shortening_rate | 4 |
| type | base |
performance
| Key | Value |
|---|---|
| test_other_wer | 8.44 |
training
| Key | Value |
|---|---|
| max_update | 320000 |
| max_tokens | 3200000 |
| lr | 0.0001 |
| gpus | 8 |
| training_precision | AMP-BF16 |
| dataset | LibriSpeech 960 h |