feat: alora adapters for ollama

docs/docs/advanced/intrinsics.md

@@ -14,7 +14,7 @@ Adapter functions are adapter-accelerated operations for RAG quality checks. The
 LoRA/aLoRA adapters loaded directly into the Hugging Face backend — faster and more
 reliable than prompting a general-purpose model for these specialized micro-tasks.
 
-> **Backend note:** Adapter functions work with two backends:
+> **Backend note:** Adapter functions work with three backends:
 >
 > - **LocalHFBackend** — loads LoRA/aLoRA adapters from the catalog at runtime.
 >   A local Granite Switch checkpoint can instead use
@@ -25,8 +25,13 @@ reliable than prompting a general-purpose model for these specialized micro-task
 >   `load_embedded_adapters=True`. Only adapter functions embedded in the model are
 >   available — check the model's `adapter_index.json` for the list.
 >   See `docs/docs/examples/granite-switch/README.md`
+> - **OllamaModelBackend**: uses an Ollama model that bundles the adapter,
+>   for example `gabegoodhart/granite4.1-uncertainty:3b`, which is `granite4.1:3b`
+>   plus the uncertainty aLoRA. Ollama bundles one adapter per model, so pass
+>   `adapter_models={"uncertainty": "<tag>", ...}` to route each adapter function
+>   to its model. Install `mellea[switch]` to download the adapter's `io.yaml`.
 >
-> Adapter functions do not work with Ollama or other remote backends.
+> Adapter functions do not work with other remote backends.
 
 Set up the backend once and reuse it across adapter function calls:
 

docs/docs/advanced/lora-and-alora-adapters.md

@@ -16,7 +16,9 @@ Apple Silicon Mac with sufficient VRAM for the chosen base model. Uploading requ
 Hugging Face account.
 
 > **Backend note:** Custom-trained adapters can only be loaded into `LocalHFBackend`.
-> They do not work with Ollama, OpenAI, or other remote backends.
+> They do not work with OpenAI or other remote backends. To use one with Ollama,
+> convert it to GGUF and bundle it into an Ollama model with a Modelfile `ADAPTER`
+> line, then pass that model tag via `OllamaModelBackend(adapter_models=...)`.
 >
 > Granite Switch models ship with pre-trained adapter functions embedded in the
 > model weights. Use them through `OpenAIBackend` with a served checkpoint, or

docs/examples/intrinsics/uncertainty_ollama.py

@@ -0,0 +1,35 @@
+# pytest: e2e, ollama, qualitative
+
+"""Example usage of the uncertainty/certainty intrinsic with Ollama.
+
+Evaluates how certain the model is about its response to a user question.
+The context should contain a user question followed by an assistant answer.
+
+Ollama bundles one adapter per model, so the uncertainty adapter is served by
+its own model tag (`granite4.1:3b` plus the uncertainty aLoRA). Pass that tag
+via `adapter_models`; normal chat still uses the base model.
+
+Requires `mellea[switch]` to download the adapter's `io.yaml`.
+
+To run this script from the root of the Mellea source tree, use the command:
+```
+uv run python docs/examples/intrinsics/uncertainty_ollama.py
+```
+"""
+
+from mellea import model_ids, start_backend
+from mellea.stdlib import functional as mfuncs
+from mellea.stdlib.components.intrinsic import core
+
+ctx, backend = start_backend(
+    "ollama",
+    model_id=model_ids.IBM_GRANITE_4_1_3B,
+    context_type="chat",
+    adapter_models={"uncertainty": "gabegoodhart/granite4.1-uncertainty:3b"},
+)
+
+response, ctx = mfuncs.chat("What is 2 + 2?", ctx, backend)  # type: ignore
+print(f"Response: {response.content}")
+
+result = core.check_certainty(ctx, backend)  # type: ignore
+print(f"Certainty score: {result}")