From 8321fcb9224c241e95cc014829e42f90c48cccc8 Mon Sep 17 00:00:00 2001 From: Guillaume Klein Date: Fri, 17 Feb 2023 14:42:09 +0100 Subject: [PATCH] Recompute the performance numbers on GPU --- README.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index 5a6f224..c668153 100644 --- a/README.md +++ b/README.md @@ -17,7 +17,8 @@ For reference, here's the time and memory usage that are required to transcribe | Implementation | Precision | Beam size | Time | Max. GPU memory | Max. CPU memory | | --- | --- | --- | --- | --- | --- | | openai/whisper | fp16 | 5 | 2m56s | 6285MB | 5598MB | -| faster-whisper | fp16 | 5 | 0m36s | 2803MB | 1722MB | +| faster-whisper | fp16 | 5 | 0m42s | 3283MB | 1722MB | +| faster-whisper | int8 | 5 | 0m41s | 1939MB | 2159MB | *Executed with CUDA 11.7.1 on a NVIDIA Tesla V100S.*