From fa5aadca6f8af99276cd084c6fe4f385628a890b Mon Sep 17 00:00:00 2001 From: tzhouam Date: Wed, 23 Sep 2026 23:02:10 +0800 Subject: [PATCH] Record MiniCPM-o GGUF Omni desktop qualification depth --- docs/design/omni_edge_e2e_roadmap.md | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/docs/design/omni_edge_e2e_roadmap.md b/docs/design/omni_edge_e2e_roadmap.md index 68573f70b..31c39e1ea 100644 --- a/docs/design/omni_edge_e2e_roadmap.md +++ b/docs/design/omni_edge_e2e_roadmap.md @@ -55,8 +55,8 @@ process, with 13.10 GB maximum sampled RSS. The initial native attempt [exited 0 with incomplete speech](../../benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/minicpmo_cpp_windows_cpu_attempt1/README.md), and is retained as a failure. Both complete C++ responses were generic rather than input descriptions; the fixture supplied no explicit description task, -so quality remains unqualified. These runs do not close the Omni stage, -admission, cancellation, streaming or M4 video/real-observation gates. +so quality remains unqualified. These earlier standalone runs did not establish +an Omni stage, admission, cancellation, streaming or M4 video/real-observation gate. The same GGUF set now also completes [native Windows Radeon 890M+CPU combined input](../../benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/minicpmo_cpp_radeon890m/README.md). @@ -69,8 +69,21 @@ The [same prompted CPU run](../../benchmarks/edge_harness/results/e2e_expansion_ returned identical text and eight language token IDs in 24.66 s, though speech PCM differed. This advances the M4 desktop feasibility evidence for a CPU+iGPU hybrid; neither the one-case quality check nor the standalone C++ -control path closes Omni integration, admission, streaming, real-input quality, -mobile artifacts or sustained profiling. +control path establishes streaming, real-input quality, mobile artifacts or +sustained profiling. + +The [Omni MiniCPM-o GGUF follow-up](../../benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/minicpmo_omni_cpp/README.md) +now binds that complete audio+image → text+speech request as one graph stage on +native Windows CPU and Radeon 890M+CPU. The same pinned weights and spoken +red-square fixture produced the correct answer and a complete WAV on both; +one cold complete request took 24.34 s on CPU and 18.27 s on Radeon+CPU. +Artifact hashes, actual CPU/Vulkan placement, 24 GiB shared-RAM reservation, +terminal event/acknowledgement and in-flight cancellation with zero remaining +ledger reservation were checked. An 8 GiB demand was refused. This closes the +scoped desktop GGUF Omni stage-binding task, not M4 acceptance: the C++ worker +does not stream playable chunks, post-cancel restart is untested, and the +single synthetic case does not establish broad multimodal quality, loading +peak, concurrency, video, mobile artifacts or sustained power behavior. Native Windows HX370 CPU now also completes two Qwen3-TTS CustomVoice text-to-WAV requests through Qwen's standalone PyTorch wrapper using the pinned