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
Studio tests: read video.py as UTF-8 in the H3 fused SwiGLU test (#17437)
Daniel Han011129b35583
diff --git a/studio/backend/tests/test_h3_fused_swiglu.py b/studio/backend/tests/test_h3_fused_swiglu.pyindex f8ba9f778e8..dc2fa31b897 100644--- a/studio/backend/tests/test_h3_fused_swiglu.py+++ b/studio/backend/tests/test_h3_fused_swiglu.py@@ -151,7 +151,7 @@ def test_h3_blocks_are_eligible(env):def test_video_installs_it_before_the_first_forward():- src = (_BACKEND / "core" / "inference" / "video.py").read_text()+ src = (_BACKEND / "core" / "inference" / "video.py").read_text(encoding = "utf-8")i = src.index("from .video_minimax_h3_ffn import install as install_h3_ffn")assert src.index("compile_blocks_below_offload_hooks(", src.index("install_int8_gemm(")) < iassert '"h3_fused_swiglu"' in src[i : i + 600]