Qwen3 EdgeLLM Jetson Artifacts

This private model repository stores generated Qwen3 ASR/TTS artifacts for Jetson Voice deployments.

Code repository: https://github.com/suharvest/qwen3-edgellm-jetson Product repository: jetson-voice

Layout

The canonical required layout is recorded in deploy/artifacts/qwen3_manifest.json.

Expected artifact families:

  • engines/orin-nano/highperf/...
  • engines/orin-nx/highperf/...
  • engines/orin-nano/official/...
  • tts/talker/...
  • tts/tokenizer/...

Large TensorRT engine files are target-device and TensorRT-version sensitive. Upload Nano and NX engine sets separately and keep checksums in the manifest.

Jetson Voice Usage

JETSON_VOICE_PROFILE=multilanguage-qwen-highperf \
QWEN3_HF_REPO_ID=harvestsu/qwen3-edgellm-jetson-artifacts \
docker compose -f deploy/docker-compose.yml up -d
Downloads last month
16
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support