From cbde541082f5eb50a38af027b2b5c83d78718d91 Mon Sep 17 00:00:00 2001 From: Yang Zhang Date: Wed, 20 May 2020 10:11:41 -0700 Subject: [PATCH] fix trt benchmark column names 2nd time to fix trt benchmark column names, since it was overwritten by another commit 1st time washttps://github.com/NVIDIA/DeepLearningExamples/pull/267 --- PyTorch/SpeechRecognition/Jasper/trt/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/PyTorch/SpeechRecognition/Jasper/trt/README.md b/PyTorch/SpeechRecognition/Jasper/trt/README.md index d6545dae..c26c1c2b 100644 --- a/PyTorch/SpeechRecognition/Jasper/trt/README.md +++ b/PyTorch/SpeechRecognition/Jasper/trt/README.md @@ -281,7 +281,7 @@ The results are obtained by running inference on LibriSpeech dev-clean dataset o #### Inference performance: NVIDIA T4 -| Sequence Length (in seconds) | Batch size | PyTorch FP16 Throughput (#sequences/second) Percentiles | | | | TensorRT FP16 Throughput (#sequences/second) Percentiles | | | | PyT/TRT Speedup | +| Sequence Length (in seconds) | Batch size | TensorRT FP16 Throughput (#sequences/second) Percentiles | | | | PyTorch FP16 Throughput (#sequences/second) Percentiles | | | | TRT/PyT Speedup | |---------------|------------|---------------------|---------|---------|---------|-----------------|---------|---------|---------|-----------------| | | | 90% | 95% | 99% | Avg | 90% | 95% | 99% | Avg | | |2|1|71.002|70.897|70.535|71.987|42.974|42.932|42.861|43.166|1.668|