Audio-to-Audio
Transformers
Safetensors
dashengtokenizer
feature-extraction
audio-classification
signal-processing
custom_code
Instructions to use mispeech/dashengtokenizer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use mispeech/dashengtokenizer with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("mispeech/dashengtokenizer", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "architectures": [ | |
| "DashengTokenizerModel" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_dasheng_tokenizer.DashengTokenizerConfig", | |
| "AutoModel": "modeling_dasheng_tokenizer.DashengTokenizerModel" | |
| }, | |
| "decoder_depth": 12, | |
| "decoder_embed_dim": 1280, | |
| "decoder_intermediate_size": 5120, | |
| "depth": 32, | |
| "dtype": "float32", | |
| "embed_dim": 1280, | |
| "hop_length": 160, | |
| "istft_hop": 320, | |
| "istft_n_fft": 1280, | |
| "model_type": "dashengtokenizer", | |
| "n_mels_patch": 128, | |
| "num_heads": 16, | |
| "transformers_version": "5.1.0", | |
| "upsample_tokens": 2 | |
| } |