a way to run this model on DGX Spark GB10 since its basicly arm64

[saikanov]

Self Checks

• I have thoroughly reviewed the project documentation (installation, training, inference) but couldn't find any relevant information that meets my needs. English 中文 日本語 Portuguese (Brazil)
• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• Please do not modify this template :) and fill in all the required fields.

1. Is this request related to a challenge you're experiencing? Tell us your story.

yes, since my device is not standar amd64 i got problem when using arm64 device like DGX Spark GB10

2. What is your suggested solution?

i tried blindly using this docker command

docker build --platform linux/arm64 -f docker/Dockerfile --build-arg BACKEND=cuda --build-arg CUDA_VER=12.9.0 --build-arg UV_EXTRA=cu129 --target webui -t fish-speech-webui:cuda .

and got this error

=> [app-base 5/10] WORKDIR /app 0.0s
=> [app-base 6/10] COPY --chown=1000:1000 pyproject.toml uv.lock README.md ./ 0.1s
=> ERROR [app-base 7/10] RUN --mount=type=cache,target=/tmp/uv-cache,uid=1000,gid=1000 uv python pin 3.12 && uv sync --extra cu129 --frozen --no-install-project 0.3s

[app-base 7/10] RUN --mount=type=cache,target=/tmp/uv-cache,uid=1000,gid=1000 uv python pin 3.12 && uv sync --extra cu129 --frozen --no-install-project:
0.244 Pinned .python-version to 3.12
0.251 Using CPython 3.12.3 interpreter at: /usr/bin/python3.12
0.251 Creating virtual environment at: .venv
0.281 error: Distribution triton==3.4.0 @ registry+https://pypi.org/simple can't be installed because it doesn't have a source distribution or wheel for the current platform
0.281
0.281 hint: You're on Linux (manylinux_2_39_aarch64), but triton (v3.4.0) only has wheels for the following platforms: manylinux_2_27_x86_64, manylinux_2_28_x86_64; consider adding your platform to tool.uv.required-environments to ensure uv resolves to a version with compatible wheels

---

Dockerfile:211

210 | # Use a generic cache path that works regardless of username
211 | >>> RUN --mount=type=cache,target=/tmp/uv-cache,uid=${USER_UID},gid=${USER_GID}
212 | >>> uv python pin ${PY_VER}
213 | >>> && uv sync --extra ${UV_EXTRA} --frozen --no-install-project
214 |

ERROR: failed to build: failed to solve: process "/bin/sh -c uv python pin ${PY_VER} && uv sync --extra ${UV_EXTRA} --frozen --no-install-project" did not complete successfully: exit code: 2

before i spend time, can i know that this model could run on my device?

3. Additional context or comments

No response

4. Can you help us with this feature?

• I am interested in contributing to this feature.

[nic-nvidia]

Any luck?