Does not work with dgx spark
#13
by
sotaaa
- opened
Tried to follow https://build.nvidia.com/spark/sglang/instructions to run a sglang docker on dgx spark, then ran the model using the provided command in model card. Got this error
launch_server.py: error: argument --reasoning-parser: invalid choice: 'nano_v3' (choose from 'deepseek-r1', 'deepseek-v3', 'glm45', 'gpt-oss', 'kimi', 'qwen3', 'qwen3-thinking', 'minimax', 'minimax-append-think', 'step3')
yes would be awesome to have it work on dgx spark
Same issue here on DGX Spark.
Environment:
- DGX Spark, GB10 GPU, CUDA 13.0.1
- Model downloaded: /home/data/models/nemotron-3-nano-bf16 (all 13 safetensors verified)
Tried:
lmsysorg/sglang:spark→nano_v3parser not available (same error as OP)nvcr.io/nvidia/vllm:25.11-py3→AttributeError: 'NemotronHConfig' object has no attribute 'rms_norm_eps'
Question:
Which exact docker image + tag supports Nemotron-3-Nano on DGX Spark today?