Merge pull request #725 from FunAudioLLM/dev/lyuxiang.lx

add trt bash script

README.md

@@ -143,7 +143,6 @@ for i, j in enumerate(cosyvoice.inference_zero_shot('收到好友从远方寄来
     torchaudio.save('zero_shot_{}.wav'.format(i), j['tts_speech'], cosyvoice.sample_rate)
 
 # fine grained control, for supported control, check cosyvoice/tokenizer/tokenizer.py#L248
-prompt_speech_16k = load_wav('zero_shot_prompt.wav', 16000)
 for i, j in enumerate(cosyvoice.inference_cross_lingual('在他讲述那个荒诞故事的过程中，他突然[laughter]停下来，因为他自己也被逗笑了[laughter]。', prompt_speech_16k, stream=False)):
     torchaudio.save('fine_grained_control_{}.wav'.format(i), j['tts_speech'], cosyvoice.sample_rate)
 

cosyvoice/bin/export_trt.sh

@@ -0,0 +1,8 @@
+#!/bin/bash
+# download tensorrt from https://developer.nvidia.com/tensorrt/download/10x, check your system and cuda for compatibability
+# for example for linux + cuda12.4, you can download https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/10.0.1/tars/TensorRT-10.0.1.6.Linux.x86_64-gnu.cuda-12.4.tar.gz
+TRT_DIR=<YOUR_TRT_DIR>
+MODEL_DIR=<COSYVOICE2_MODEL_DIR>
+
+export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$TRT_DIR/lib:/usr/local/cuda/lib64
+$TRT_DIR/bin/trtexec --onnx=$MODEL_DIR/flow.decoder.estimator.fp32.onnx --saveEngine=$MODEL_DIR/flow.decoder.estimator.fp16.mygpu.plan --fp16 --minShapes=x:2x80x4,mask:2x1x4,mu:2x80x4,cond:2x80x4 --optShapes=x:2x80x193,mask:2x1x193,mu:2x80x193,cond:2x80x193 --maxShapes=x:2x80x6800,mask:2x1x6800,mu:2x80x6800,cond:2x80x6800 --inputIOFormats=fp16:chw,fp16:chw,fp16:chw,fp16:chw,fp16:chw,fp16:chw --outputIOFormats=fp16:chw