Configuration for an offline Qwen3-ASR model.
More...
#include <c-api.h>
Definition at line 999 of file c-api.h.
◆ conv_frontend
| const char* SherpaOnnxOfflineQwen3ASRModelConfig::conv_frontend |
Path to the conv-frontend ONNX model.
Definition at line 1001 of file c-api.h.
◆ decoder
| const char* SherpaOnnxOfflineQwen3ASRModelConfig::decoder |
Path to the decoder ONNX model (with KV cache).
Definition at line 1005 of file c-api.h.
◆ encoder
| const char* SherpaOnnxOfflineQwen3ASRModelConfig::encoder |
Path to the encoder ONNX model.
Definition at line 1003 of file c-api.h.
◆ hotwords
| const char* SherpaOnnxOfflineQwen3ASRModelConfig::hotwords |
Optional comma-separated hotwords (UTF-8, ASCII ','), e.g. "foo,bar,baz".
Definition at line 1020 of file c-api.h.
◆ max_new_tokens
| int32_t SherpaOnnxOfflineQwen3ASRModelConfig::max_new_tokens |
Maximum number of new tokens to generate.
Definition at line 1011 of file c-api.h.
◆ max_total_len
| int32_t SherpaOnnxOfflineQwen3ASRModelConfig::max_total_len |
Maximum total sequence length supported by the model.
Definition at line 1009 of file c-api.h.
◆ seed
| int32_t SherpaOnnxOfflineQwen3ASRModelConfig::seed |
Random seed for reproducible sampling.
Definition at line 1017 of file c-api.h.
◆ temperature
| float SherpaOnnxOfflineQwen3ASRModelConfig::temperature |
Sampling temperature.
Definition at line 1013 of file c-api.h.
◆ tokenizer
| const char* SherpaOnnxOfflineQwen3ASRModelConfig::tokenizer |
Path to the tokenizer directory (e.g. containing vocab.json).
Definition at line 1007 of file c-api.h.
◆ top_p
| float SherpaOnnxOfflineQwen3ASRModelConfig::top_p |
Top-p (nucleus) sampling threshold.
Definition at line 1015 of file c-api.h.
The documentation for this struct was generated from the following file: