Riva-build speech_recognition cannot be started due to the Error

AI & Data Science Deep Learning (Training & Inference) Riva
,
1

Please provide the following information when requesting support.

Hardware - GPU T4
Hardware - CPU x86_64 Intel(R) Xeon(R) Platinum
Operating System - Linux
Riva Version - NVIDIA Release devel (build 22382700)
TLT Version (if relevant)
How to reproduce the issue ?

The error appears after launching the following command in Jupyter notebook cell:

!docker run --rm --gpus 1 \
    -v $EXPORT_MODEL_LOC:/tao \
    -v $RMIR_LOC:/riva \
    $RIVA_SM_CONTAINER -- \
    riva-build speech_recognition --decoder_type=greedy \
    -f /riva/$RMIR_MODEL_NAME:$KEY /tao/$EXPORT_MODEL_NAME:$KEY \
    2>&1|tee my_assessment/step5.txt # DO NOT REMOVE THIS LINE

Where environment variables were assigned the following values

# Set the Riva paths for the project
WORKSPACE = "/dli/task"

##### Riva Paths
# ServiceMaker Docker
RIVA_SM_CONTAINER = "nvcr.io/nvidia/riva/riva-speech:1.4.0-beta-servicemaker"

# Model output directories
RMIR_LOC = WORKSPACE + "/riva/riva_quickstart/models_repo_assessment/rmir"
RIVA_MODEL_LOC = WORKSPACE + '/riva/riva_quickstart/models_repo_assessment'

# Model Names
EXPORT_MODEL_NAME = "tc-model.riva"  
RMIR_MODEL_NAME = "tc-model.rmir"

# Riva Quick Start 
RIVA_QS = WORKSPACE + "/riva/riva_quickstart"
2

Hi @german.sergei

Thanks for your interest in riva

Can you share the Jupyter notebook you are trying, if it is a NGC based notebook please share the link with us

Please share the error logs with us

Please share the above requested files and logs

Also the version you are using seems to be pretty old
we recommend using the latest one

# ServiceMaker Docker
RIVA_SM_CONTAINER = "nvcr.io/nvidia/riva/riva-speech:2.8.1-servicemaker"

Thanks