Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
AMD iGPU routing tests: clear the visibility masks for every test (#11752)
Daniel Hanba18ca78a7ec
diff --git a/tests/studio/install/test_amd_hybrid_igpu_routing_11143.py b/tests/studio/install/test_amd_hybrid_igpu_routing_11143.pyindex 89e63b3c9ef..bdd481b1444 100644--- a/tests/studio/install/test_amd_hybrid_igpu_routing_11143.py+++ b/tests/studio/install/test_amd_hybrid_igpu_routing_11143.py@@ -41,7 +41,10 @@ def _rocminfo(*arches: str) -> str:)+# Autouse: _pick_rocm_gfx_target reads the visibility masks, so a test that forgot to ask for+# this passed on a bare CI runner and failed wherever CUDA_VISIBLE_DEVICES is set, a GPU box+# included. Tests that want a mask set it after this runs.[email protected](autouse = True)def unmasked(monkeypatch):for name in _VISIBILITY_ENV:monkeypatch.delenv(name, raising = False)