From b0c1ca4f7549d7acdfa52a7dcffc934bc63a43ce Mon Sep 17 00:00:00 2001 From: Daniel Hiltgen Date: Thu, 1 Oct 2026 18:05:01 -0700 Subject: [PATCH 1/3] ci: fix missing build context (#18742) --- .github/workflows/test.yaml | 1 + Dockerfile | 7 +++++++ 2 files changed, 8 insertions(+) diff --git a/.github/workflows/test.yaml b/.github/workflows/test.yaml index f7ea8bf56..156916bbf 100644 --- a/.github/workflows/test.yaml +++ b/.github/workflows/test.yaml @@ -46,6 +46,7 @@ jobs: 'cmake/**/*' \ 'llama/server/**/*' \ 'llama/compat/**/*' \ + 'llama/clef/**/*' \ 'LLAMA_CPP_VERSION' \ 'MLX_VERSION' \ 'MLX_C_VERSION' \ diff --git a/Dockerfile b/Dockerfile index 6c629077e..7099e1c2a 100644 --- a/Dockerfile +++ b/Dockerfile @@ -84,6 +84,7 @@ FROM cpu-deps AS llama-server-cpu COPY LLAMA_CPP_VERSION . COPY llama/server llama/server COPY llama/compat llama/compat +COPY llama/clef llama/clef COPY cmake cmake RUN --mount=type=cache,target=/root/.ccache \ cmake -S llama/server --preset cpu \ @@ -104,6 +105,7 @@ FROM cuda-12-deps AS llama-server-cuda_v12 COPY LLAMA_CPP_VERSION . COPY llama/server llama/server COPY llama/compat llama/compat +COPY llama/clef llama/clef COPY cmake cmake RUN --mount=type=cache,target=/root/.ccache \ cmake -S llama/server --preset llama_cuda_v12_linux \ @@ -117,6 +119,7 @@ FROM cuda-13-deps AS llama-server-cuda_v13 COPY LLAMA_CPP_VERSION . COPY llama/server llama/server COPY llama/compat llama/compat +COPY llama/clef llama/clef COPY cmake cmake RUN --mount=type=cache,target=/root/.ccache \ cmake -S llama/server --preset llama_cuda_v13_linux \ @@ -131,6 +134,7 @@ ENV CC=clang CXX=clang++ CXXFLAGS=--gcc-toolchain=/opt/rh/gcc-toolset-13/root/us COPY LLAMA_CPP_VERSION . COPY llama/server llama/server COPY llama/compat llama/compat +COPY llama/clef llama/clef COPY cmake cmake RUN --mount=type=cache,target=/root/.ccache \ cmake -S llama/server --preset rocm_v7_2_linux \ @@ -145,6 +149,7 @@ FROM vulkan-deps AS llama-server-vulkan COPY LLAMA_CPP_VERSION . COPY llama/server llama/server COPY llama/compat llama/compat +COPY llama/clef llama/clef COPY cmake cmake RUN --mount=type=cache,target=/root/.ccache \ cmake -S llama/server --preset vulkan \ @@ -170,6 +175,7 @@ ENV CMAKE_GENERATOR=Ninja COPY LLAMA_CPP_VERSION . COPY llama/server llama/server COPY llama/compat llama/compat +COPY llama/clef llama/clef COPY cmake cmake RUN --mount=type=cache,target=/root/.ccache \ cmake -S llama/server --preset llama_cuda_jetpack5 \ @@ -191,6 +197,7 @@ ENV CMAKE_GENERATOR=Ninja COPY LLAMA_CPP_VERSION . COPY llama/server llama/server COPY llama/compat llama/compat +COPY llama/clef llama/clef COPY cmake cmake RUN --mount=type=cache,target=/root/.ccache \ cmake -S llama/server --preset llama_cuda_jetpack6 \ From e896a52a4f55f70d2986ae0440c0957911f46294 Mon Sep 17 00:00:00 2001 From: Parth Sareen Date: Fri, 2 Oct 2026 11:51:55 -0700 Subject: [PATCH 2/3] docs: link to available decision models (#18758) --- docs/capabilities/decision.mdx | 2 ++ 1 file changed, 2 insertions(+) diff --git a/docs/capabilities/decision.mdx b/docs/capabilities/decision.mdx index 5ba05daf2..b803a7cf3 100644 --- a/docs/capabilities/decision.mdx +++ b/docs/capabilities/decision.mdx @@ -9,6 +9,8 @@ Currently available locally in Ollama. Use System One to classify text, answer yes/no questions, or score text against a rubric. +See the [full list of decision models](https://ollama.com/search?c=decision). + ## Setup Start [Ollama](/quickstart) v0.35.0 or later, then download [Nimble](https://ollama.com/library/nimble): From 42e911bc3d05798cad729cb474bf62f378cb2e26 Mon Sep 17 00:00:00 2001 From: Parth Sareen Date: Fri, 2 Oct 2026 11:52:27 -0700 Subject: [PATCH 3/3] docs: document image input for decision models (#18757) --- docs/capabilities/decision.mdx | 50 ++++++++++++++++++++++++++++++++++ 1 file changed, 50 insertions(+) diff --git a/docs/capabilities/decision.mdx b/docs/capabilities/decision.mdx index b803a7cf3..264092e9f 100644 --- a/docs/capabilities/decision.mdx +++ b/docs/capabilities/decision.mdx @@ -19,6 +19,8 @@ Start [Ollama](/quickstart) v0.35.0 or later, then download [Nimble](https://oll ollama pull nimble ``` +Nimble, Tev1, [Clef](https://ollama.com/library/clef), and [Clef Flash](https://ollama.com/library/clef-flash) are decision models. The Clef models can also judge images and require v0.35.1 or later. + Local requests do not need an API key. ## Choose a label @@ -62,6 +64,54 @@ Example response: } ``` +## Judge an image + +[Clef](https://ollama.com/library/clef) and [Clef Flash](https://ollama.com/library/clef-flash) read images alongside the state. Send base64-encoded PNG, JPEG, or WebP files in `images`; they are shared by all questions, in request order. URLs and data URLs are not accepted. `state` is still required — it holds the context for the decision. + +```shell image.sh +curl http://localhost:11434/v1/systemone \ + -H 'Content-Type: application/json' \ + -d '{ + "model": "clef-flash", + "state": "A user took this screenshot and wants to know what it shows.", + "images": [""], + "questions": { + "has_ollama": { + "type": "noul", + "instructions": "Does this image contain Ollama?" + }, + "app": { + "type": "choice", + "instructions": "Which application is shown in this screenshot?", + "criteria": { + "vscode": "Visual Studio Code", + "terminal": "A terminal emulator", + "browser": "A web browser", + "other": "A different application" + } + } + } + }' +``` + +Example response: + +```json +{ + "model": "clef-flash", + "answers": { + "has_ollama": {"type": "noul", "noul": 0.959}, + "app": { + "type": "choice", + "choice": "vscode", + "probabilities": {"vscode": 0.966, "other": 0.017, "browser": 0.010, "terminal": 0.007}, + "confidence": 0.868 + } + }, + "usage": {"input_tokens": 677, "output_tokens": 0} +} +``` + ## Question types | Type | Criteria | Result |