ASR Model Unavailable on Jetson Orin AGX

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

NVIDIA Jetson AGX Orin Developer Kit
L4T 36.3.0
Ubuntu 22.04.5 LTS
Kernel Version: 5.15.136-tegra
CUDA 12.2.140
CUDA Architecture: 8.7
OpenCV version: 4.8.0
OpenCV Cuda: NO
CUDNN: 8.9.4.25
TensorRT: 8.6.2.3
Vision Works: NOT_INSTALLED
VPI: 3.1.5
Vulcan: 1.3.204

Architecture: aarch64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 12
On-line CPU(s) list: 0-11
Vendor ID: ARM
Model name: Cortex-A78AE

RIVA version: 2.17.0-l4t-aarch64

I run the following from the RIVA start guide and get a message that the model is unavailable:

root@2a386c161a78:/opt/riva# riva_asr_client --audio_file=/opt/riva/wav/en-US_sample.wav
I1126 22:54:51.504103 211 grpc.h:94] Using Insecure Server Credentials
Loading eval dataset…
filename: /opt/riva/wav/en-US_sample.wav
Done loading 1 files
RPC failed: Error: Unavailable model requested given these parameters: language_code=en-US; type=offline;
Done processing 1 responses
Some requests failed to complete properly, not printing performance stats

(I could not upload a .sh file so config.sh is attached as a .txt file)

riva-speech-logs.txt (130 Bytes)
config.txt (20.4 KB)

2

Hi @angelrodriguez0617 ,
I assume that you need to enable the offline ASR model to get that example working… we provide streaming ASR as prebuilt, but offline needs to be enabled in config.sh.
by uncommenting

### Offline w/ CPU decoder
          #"${riva_ngc_org}/${riva_ngc_team}/rmir_asr_${asr_acoustic_model}${modified_asr_acoustic_model_variant}_${modified_lang_code}_ofl${decoder}:${riva_ngc_model_version}"
      )