feat: Ollama embeddings api + Ollama tests (#1433)

Co-authored-by: Krishna Mandal <krishna@mandal.us>

.github/workflows/test_ollama.yml

@@ -0,0 +1,38 @@
+name: Endpoint (Ollama)
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+    steps:
+    - name: Checkout
+      uses: actions/checkout@v4
+    
+    - name: Start Ollama Server
+      run: |
+        curl -fsSL https://ollama.com/install.sh | sh
+        ollama serve &
+        sleep 10 # wait for server
+        ollama pull dolphin2.2-mistral:7b-q6_K 
+        ollama pull mxbai-embed-large
+
+    - name: "Setup Python, Poetry and Dependencies"
+      uses: packetcoders/action-setup-cache-python-poetry@main
+      with:
+        python-version: "3.12"
+        poetry-version: "1.8.2"
+        install-args: "-E dev -E ollama"
+
+    - name: Test LLM endpoint
+      run: |
+        poetry run pytest -s -vv tests/test_endpoints.py::test_llm_endpoint_ollama
+
+    - name: Test embedding endpoint
+      run: |
+        poetry run pytest -s -vv tests/test_endpoints.py::test_embedding_endpoint_ollama

.github/workflows/tests.yml

@@ -28,8 +28,7 @@ jobs:
         install-args: "-E dev -E postgres -E milvus"
 
     - name: Initialize credentials
-      run: |
-        poetry run memgpt quickstart --backend memgpt
+      run: poetry run memgpt quickstart --backend memgpt
 
     #- name: Run docker compose server
     #  env: