| { |
| "model": "parakeet-tdt_ctc-0.6b-ja", |
| "language": "ja (Japanese)", |
| "source": "nvidia/parakeet-tdt_ctc-0.6b-ja", |
| "sample_rate": 16000, |
| "max_audio_seconds": 15.0, |
| "max_samples": 240000, |
| "vocab_size": 3072, |
| "blank_id": 3072, |
| "mel_features": 80, |
| "mel_frames": 1501, |
| "encoder_dim": 1024, |
| "time_steps": 188, |
| "components": { |
| "preprocessor": { |
| "input": { |
| "audio_signal": [ |
| 1, |
| 240000 |
| ], |
| "audio_length": [ |
| 1 |
| ] |
| }, |
| "output": { |
| "mel_features": [ |
| 1, |
| 80, |
| 1501 |
| ], |
| "mel_length": [ |
| 1 |
| ] |
| } |
| }, |
| "encoder": { |
| "input": { |
| "mel_features": [ |
| 1, |
| 80, |
| 1501 |
| ], |
| "mel_length": [ |
| 1 |
| ] |
| }, |
| "output": { |
| "encoder": [ |
| 1, |
| 1024, |
| 188 |
| ], |
| "encoder_length": [ |
| 1 |
| ] |
| }, |
| "quantization": "int8-per-channel" |
| }, |
| "ctc_decoder": { |
| "input": { |
| "encoder": [ |
| 1, |
| 1024, |
| 188 |
| ] |
| }, |
| "output": { |
| "ctc_logits": [ |
| 1, |
| 188, |
| 3073 |
| ] |
| } |
| } |
| } |
| } |
