NVIDIA-NeMo/Speech

Speed up RNNT model inference using TRT

オープン

#14,531 opened on 2025/08/20

 (1 件のコメント) (0 件のリアクション) (1 人の担当者)Python (3,558 件のフォーク)github user discovery
ASRcommunity-requesthelp wantedwaiting-on-customer

Repository metrics

Stars
 (18,152 個のスター)
PR merge metrics
 (30d に merged PR はありません)

説明

Hi,

I previously trained an RNNT model and now want to accelerate it by converting it to TensorRT. I’ve exported the model to ONNX and have encoder.onnx and decoder.onnx.

I’m using the TensorRT 25.03 Docker image and trtexec to convert the models. The decoder works fine with --fp16, but when I use --fp16 for the encoder, some outputs return NaN and the results are incorrect.

Has anyone encountered this issue or knows how to fix it?

Are there any methods to accelerate RNNT model inference?

コントリビューターガイド