Inference results are different on different computers

tranmanhdat commented 3 years ago

Bug Description

when i inference model converted (streaming convnet) and use the same converted model, options and audio files; the results are different on different computer.

Reproduction Steps

i follow tutorial inference tutorial

Platform and Hardware

Computer 1: Intel(R) Core(TM) i5-8300H CPU @ 2.30GHz, nvidia geforce 1050 4gb, ubuntu 18.04, run on docker-nvidia-2
Computer 2: Intel(R) Xeon(R) CPU E5-2690 v4@ 2.60GHz, Tesla V100 16gb, ubuntu 18.04, run on docker-nvidia-2
Additional Context

vineelpratap commented 3 years ago

Can you check if this PR is included in your commit https://github.com/fairinternal/wav2letter/commit/e4f6d1d236e653257c0377794f251c6810b4b2e6

tlikhomanenko commented 3 years ago

cc @vineelpratap @avidov

tranmanhdat commented 3 years ago

Can you check if this PR is included in your commit fairinternal/wav2letter@e4f6d1d

i wonder do you check this or forgoten?

tlikhomanenko commented 3 years ago

@vineelpratap can you point the correct link to the commit you mentioned which probably fixes the problem?

flashlight / wav2letter