Why does this model only generate audio content lasting more than 20 seconds when processing a 2-minute WAV audio file?
#25
by xwang99 - opened
qwen-asr-demo --asr-checkpoint Qwen3-ASR-0.6B --aligner-checkpoint Qwen3-ForcedAligner-0.6B --backend transformers --cuda-visible-devices 0 --backend-kwargs '{"device_map":"cuda:0","dtype":"bfloat16","max_inference_batch_size":2,"max_new_tokens":51200}' --aligner-kwargs '{"device_map":"cuda:1","dtype":"bfloat16"}' --ip 0.0.0.0 --port 8000
Perform the conversion through the web port 8000 of the Qwen3 ASR Demo