| { | |
| "architectures": [ | |
| "SenseVoiceSmall" | |
| ], | |
| "blank_id": 0, | |
| "dtype": "float32", | |
| "encoder_conf": { | |
| "attention_dropout_rate": 0.1, | |
| "attention_heads": 4, | |
| "dropout_rate": 0.1, | |
| "input_layer": "pe", | |
| "kernel_size": 11, | |
| "linear_units": 2048, | |
| "normalize_before": true, | |
| "num_blocks": 50, | |
| "output_size": 512, | |
| "positional_dropout_rate": 0.1, | |
| "sanm_shfit": 0, | |
| "selfattention_layer_type": "sanm", | |
| "tp_blocks": 20 | |
| }, | |
| "input_size": 560, | |
| "model_type": "sensevoice_small", | |
| "query_size": 16, | |
| "transformers_version": "4.57.6", | |
| "vocab_size": 25055 | |
| } | |