feat: Draft ollama test (#566)

<!-- .github/pull_request_template.md -->

## Description
<!-- Provide a clear description of the changes in this PR -->

## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

- **Tests**
- Introduced new automated testing workflows for Ollama and Gemini,
triggered by pull requests and manual dispatch.
- The Ollama workflow sets up the service and executes a simple example
test to enhance continuous integration.
- Enhanced dependency update workflow with new triggers for push and
pull request events, and added an optional debug logging parameter.
- Added new capabilities for audio and image transcription within the
Ollama API adapter.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Daniel Molnar <soobrosa@gmail.com>

.github/workflows/test_gemini.yml

@@ -0,0 +1,33 @@
+name: test | gemini
+
+on:
+  workflow_dispatch:
+  pull_request:
+    types: [labeled, synchronize]
+
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  run_simple_example_test:
+      uses: ./.github/workflows/reusable_python_example.yml
+      with:
+        example-location: ./examples/python/simple_example.py
+      secrets:
+        OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
+        GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
+        EMBEDDING_PROVIDER: "gemini"
+        EMBEDDING_API_KEY: ${{ secrets.GEMINI_API_KEY }}
+        EMBEDDING_MODEL: "gemini/text-embedding-004"
+        EMBEDDING_ENDPOINT: "https://generativelanguage.googleapis.com/v1beta/models/text-embedding-004"
+        EMBEDDING_API_VERSION: "v1beta"
+        EMBEDDING_DIMENSIONS: 768
+        EMBEDDING_MAX_TOKENS: 8076
+        LLM_PROVIDER: "gemini"
+        LLM_API_KEY: ${{ secrets.GEMINI_API_KEY }}
+        LLM_MODEL: "gemini/gemini-1.5-flash"
+        LLM_ENDPOINT: "https://generativelanguage.googleapis.com/"
+        LLM_API_VERSION: "v1beta"

.github/workflows/test_ollama.yml

@@ -0,0 +1,116 @@
+name: test | ollama
+
+on:
+  workflow_dispatch:
+  pull_request:
+    types: [ labeled, synchronize ]
+
+jobs:
+
+  run_simple_example_test:
+
+    # needs 16 Gb RAM for phi4
+    runs-on: buildjet-4vcpu-ubuntu-2204
+#    services:
+#      ollama:
+#        image: ollama/ollama
+#        ports:
+#          - 11434:11434
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Setup Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.12.x'
+
+      - name: Install Poetry
+        uses: snok/install-poetry@v1.4.1
+        with:
+          virtualenvs-create: true
+          virtualenvs-in-project: true
+          installer-parallel: true
+
+      - name: Install dependencies
+        run: |
+          poetry install --no-interaction --all-extras
+          poetry add torch
+
+#      - name: Install ollama
+#        run: curl -fsSL https://ollama.com/install.sh | sh
+#      - name: Run ollama
+#        run: |
+#          ollama serve --openai &
+#          ollama pull llama3.2 &
+#          ollama pull avr/sfr-embedding-mistral:latest
+
+      - name: Start Ollama container
+        run: |
+          docker run -d --name ollama -p 11434:11434 ollama/ollama
+          sleep 5
+          docker exec -d ollama bash -c "ollama serve --openai"
+
+      - name: Check Ollama logs
+        run: docker logs ollama
+
+      - name: Wait for Ollama to be ready
+        run: |
+          for i in {1..30}; do
+            if curl -s http://localhost:11434/v1/models > /dev/null; then
+              echo "Ollama is ready"
+              exit 0
+            fi
+            echo "Waiting for Ollama... attempt $i"
+            sleep 2
+          done
+          echo "Ollama failed to start"
+          exit 1
+
+      - name: Pull required Ollama models
+        run: |
+          curl -X POST http://localhost:11434/api/pull -d '{"name": "phi4"}'
+          curl -X POST http://localhost:11434/api/pull -d '{"name": "avr/sfr-embedding-mistral:latest"}'
+
+      - name: Call ollama API
+        run: |
+          curl -X POST http://localhost:11434/v1/chat/completions \
+            -H "Content-Type: application/json" \
+            -d '{
+              "model": "phi4",
+              "stream": false,
+              "messages": [
+                { "role": "system", "content": "You are a helpful assistant." },
+                { "role": "user", "content": "Whatever I say, answer with Yes." }
+              ]
+            }'
+          curl -X POST http://127.0.0.1:11434/v1/embeddings \
+            -H "Content-Type: application/json" \
+             -d '{
+              "model": "avr/sfr-embedding-mistral:latest",
+            "input": "This is a test sentence to generate an embedding."
+            }'
+
+      - name: Dump Docker logs
+        run: |
+          docker ps
+          docker logs $(docker ps --filter "ancestor=ollama/ollama" --format "{{.ID}}")
+
+
+      - name: Run example test
+        env:
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+          GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
+          GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
+          PYTHONFAULTHANDLER: 1
+          LLM_PROVIDER: "ollama"
+          LLM_API_KEY: "ollama"
+          LLM_ENDPOINT: "http://localhost:11434/v1/"
+          LLM_MODEL: "phi4"
+          EMBEDDING_PROVIDER: "ollama"
+          EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
+          EMBEDDING_ENDPOINT: "http://localhost:11434/v1/"
+          EMBEDDING_DIMENSIONS: "4096"
+          HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
+        run: poetry run python ./examples/python/simple_example.py

.github/workflows/upgrade_deps.yml

@@ -2,8 +2,29 @@ name: Update Poetry Dependencies
 
 on:
   schedule:
-    - cron: '0 3 * * 0'
+    - cron: '0 3 * * 0'  # Runs at 3 AM every Sunday
+  push:
+    paths:
+      - 'poetry.lock'
+      - 'pyproject.toml'
+    branches:
+      - main
+      - dev
+  pull_request:
+    paths:
+      - 'poetry.lock'
+      - 'pyproject.toml'
+    types: [opened, synchronize, reopened]
+    branches:
+      - main
+      - dev
   workflow_dispatch:
+    inputs:
+      debug_enabled:
+        type: boolean
+        description: 'Run the update with debug logging'
+        required: false
+        default: false
 
 jobs:
   update-dependencies:

cognee/infrastructure/llm/ollama/adapter.py

@@ -4,6 +4,8 @@ import instructor
 from cognee.infrastructure.llm.llm_interface import LLMInterface
 from cognee.infrastructure.llm.config import get_llm_config
 from openai import OpenAI
+import base64
+import os
 
 
 class OllamaAPIAdapter(LLMInterface):
@@ -42,3 +44,54 @@ class OllamaAPIAdapter(LLMInterface):
         )
 
         return response
+
+    def create_transcript(self, input_file: str) -> str:
+        """Generate an audio transcript from a user query."""
+
+        if not os.path.isfile(input_file):
+            raise FileNotFoundError(f"The file {input_file} does not exist.")
+
+        with open(input_file, "rb") as audio_file:
+            transcription = self.aclient.audio.transcriptions.create(
+                model="whisper-1",  # Ensure the correct model for transcription
+                file=audio_file,
+                language="en",
+            )
+
+        # Ensure the response contains a valid transcript
+        if not hasattr(transcription, "text"):
+            raise ValueError("Transcription failed. No text returned.")
+
+        return transcription.text
+
+    def transcribe_image(self, input_file: str) -> str:
+        """Transcribe content from an image using base64 encoding."""
+
+        if not os.path.isfile(input_file):
+            raise FileNotFoundError(f"The file {input_file} does not exist.")
+
+        with open(input_file, "rb") as image_file:
+            encoded_image = base64.b64encode(image_file.read()).decode("utf-8")
+
+        response = self.aclient.chat.completions.create(
+            model=self.model,
+            messages=[
+                {
+                    "role": "user",
+                    "content": [
+                        {"type": "text", "text": "What’s in this image?"},
+                        {
+                            "type": "image_url",
+                            "image_url": {"url": f"data:image/jpeg;base64,{encoded_image}"},
+                        },
+                    ],
+                }
+            ],
+            max_tokens=300,
+        )
+
+        # Ensure response is valid before accessing .choices[0].message.content
+        if not hasattr(response, "choices") or not response.choices:
+            raise ValueError("Image transcription failed. No response received.")
+
+        return response.choices[0].message.content