feat: Add helm to core

.github/workflows/test_ollama.yml

@@ -9,12 +9,15 @@ jobs:
 
   run_simple_example_test:
 
-    runs-on: ubuntu-latest
-    services:
-      ollama:
-        image: ollama/ollama
-        ports:
-          - 11434:11434
+
+    # needs 16 Gb RAM for phi4
+    runs-on: buildjet-4vcpu-ubuntu-2204
+#    services:
+#      ollama:
+#        image: ollama/ollama
+#        ports:
+#          - 11434:11434
+
 
     steps:
       - name: Checkout repository
@@ -36,21 +39,32 @@ jobs:
         run: |
           poetry install --no-interaction --all-extras
 
-      - name: Install ollama
-        run: curl -fsSL https://ollama.com/install.sh | sh
-      - name: Run ollama
-        run: |
-          ollama serve &
-          ollama pull llama3.2 &
-          ollama pull avr/sfr-embedding-mistral:latest
-      - name: Call ollama API
+          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: |
-          curl -d '{"model": "llama3.2", "stream": false, "prompt":"Whatever I say, asnwer with Yes"}' http://localhost:11434/api/generate
+          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/api/tags > /dev/null; then
+
+            if curl -s http://localhost:11434/v1/models > /dev/null; then
+
               echo "Ollama is ready"
               exit 0
             fi
@@ -60,6 +74,31 @@ jobs:
           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
@@ -72,13 +111,15 @@ jobs:
           GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
           GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
           PYTHONFAULTHANDLER: 1
-          LLM_API_KEY: "ollama"
+
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://127.0.0.1:11434/api/generate"
-          LLM_MODEL: "ollama/llama3.2"
+          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://127.0.0.1:11434/api/embeddings"
+          EMBEDDING_ENDPOINT: "http://localhost:11434/v1/"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
-        run: poetry run python ./examples/python/simple_example.py
+        run: poetry run python ./examples/python/simple_example.py
+

cognee/infrastructure/llm/ollama/adapter.py

@@ -1,14 +1,12 @@
-from sys import api_version
 from typing import Type
 from pydantic import BaseModel
 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
-from pathlib import Path
-import os
 
+import os
 
 class OllamaAPIAdapter(LLMInterface):
     """Adapter for a Ollama API LLM provider using instructor with an OpenAI backend."""
@@ -17,21 +15,13 @@ class OllamaAPIAdapter(LLMInterface):
 
     MAX_RETRIES = 5
 
-    def __init__(
-        self,
-        endpoint: str,
-        api_key: str,
-        model: str,
-        name: str,
-        max_tokens: int,
-        api_version: str = None,
-    ) -> None:
+    def __init__(self, endpoint: str, api_key: str, model: str, name: str, max_tokens: int, api_version: str = None) -> None:
         self.name = name
         self.model = model
         self.api_key = api_key
         self.endpoint = endpoint
         self.max_tokens = max_tokens
-        self.api_version = api_version
+        self.api_version= api_version
 
         self.aclient = instructor.from_openai(
             OpenAI(base_url=self.endpoint, api_key=self.api_key), mode=instructor.Mode.JSON
@@ -60,52 +50,57 @@ async def acreate_structured_output(
 
         return response
 
-    def create_transcript(self, input):
-        """Generate a audio transcript from a user query."""
 
-        if not os.path.isfile(input):
-            raise FileNotFoundError(f"The file {input} does not exist.")
+    def create_transcript(self, input_file: str) -> str:
+        """Generate an audio transcript from a user query."""
 
-        # with open(input, 'rb') as audio_file:
-        #     audio_data = audio_file.read()
+        if not os.path.isfile(input_file):
+            raise FileNotFoundError(f"The file {input_file} does not exist.")
 
-        transcription = self.aclient.transcription(
-            model=self.transcription_model,
-            file=Path(input),
-            api_key=self.api_key,
-            api_base=self.endpoint,
-            api_version=self.api_version,
-            max_retries=self.MAX_RETRIES,
-        )
+        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
+        return transcription.text
 
-    def transcribe_image(self, input) -> BaseModel:
-        with open(input, "rb") as image_file:
+    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")
 
-        return self.aclient.completion(
+        response = self.aclient.chat.completions.create(
+
             model=self.model,
             messages=[
                 {
                     "role": "user",
                     "content": [
-                        {
-                            "type": "text",
-                            "text": "What’s in this image?",
-                        },
+
+                        {"type": "text", "text": "What’s in this image?"},
                         {
                             "type": "image_url",
-                            "image_url": {
-                                "url": f"data:image/jpeg;base64,{encoded_image}",
-                            },
+                            "image_url": {"url": f"data:image/jpeg;base64,{encoded_image}"},
                         },
                     ],
                 }
             ],
-            api_key=self.api_key,
-            api_base=self.endpoint,
-            api_version=self.api_version,
             max_tokens=300,
-            max_retries=self.MAX_RETRIES,
         )
+
+        # 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
+