From c572e2714580ecfb046313c68d5c539a48b39e4c Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Wed, 19 Feb 2025 15:15:59 -0800
Subject: [PATCH 01/48] Draft ollama test

---
 .github/workflows/test_ollama.yml | 59 +++++++++++++++++++++++++++++++
 1 file changed, 59 insertions(+)
 create mode 100644 .github/workflows/test_ollama.yml

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
new file mode 100644
index 0000000000..8af4bbb963
--- /dev/null
+++ b/.github/workflows/test_ollama.yml
@@ -0,0 +1,59 @@
+name: test | ollama
+
+on:
+  schedule:
+    - cron: "0 0 * * *" # daily schedule
+  workflow_dispatch:
+
+jobs:
+  update-insights:
+    runs-on: ubuntu-latest
+    services:
+      ollama:
+        image: ollama/ollama
+        ports:
+          - 11434:11434
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          token: ${{ secrets.GH_PAT }} # Use PAT instead of default token
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: "3.x"
+
+      - name: Cache pip packages
+        uses: actions/cache@v4
+        with:
+          path: ~/.cache/pip
+          key: ${{ runner.os }}-pip-${{ hashFiles('**/requirements.txt') }}
+          restore-keys: |
+            ${{ runner.os }}-pip-
+
+      - name: Install dependencies
+        run: pip install -r requirements.txt
+
+      - name: Wait for Ollama to be ready
+        run: |
+          for i in {1..30}; do
+            curl -s http://localhost:11434/api/tags && break
+            sleep 2
+          done
+
+      - name: Pull Model
+        run: |
+          curl -X POST http://localhost:11434/api/pull -d '{"name": "mistral"}'
+          
+
+  run_simple_example_test:
+      uses: ./.github/workflows/reusable_python_example.yml
+      with:
+        example-location: ./examples/python/simple_example.py
+      secrets:
+        LLM_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
+        GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
\ No newline at end of file

From 2de364b93731f9074af118c6afa114270f678c46 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Wed, 19 Feb 2025 18:02:33 -0800
Subject: [PATCH 02/48] Ollama test end to end

---
 .github/workflows/test_ollama.yml | 21 ++++++++++++++++-----
 1 file changed, 16 insertions(+), 5 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 8af4bbb963..2473de3c2d 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -1,9 +1,9 @@
 name: test | ollama
 
 on:
-  schedule:
-    - cron: "0 0 * * *" # daily schedule
   workflow_dispatch:
+  pull_request:
+    types: [ labeled, synchronize ]
 
 jobs:
   update-insights:
@@ -45,7 +45,10 @@ jobs:
 
       - name: Pull Model
         run: |
-          curl -X POST http://localhost:11434/api/pull -d '{"name": "mistral"}'
+          curl -X POST http://localhost:11434/api/pull -d '{"name": "ollama/llama3.2"}'
+      - name: Pull Embedding Model
+        run: |
+          curl -X POST http://localhost:11434/api/pull -d '{"name": "avr/sfr-embedding-mistral:latest"}'
           
 
   run_simple_example_test:
@@ -53,7 +56,15 @@ jobs:
       with:
         example-location: ./examples/python/simple_example.py
       secrets:
-        LLM_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        LLM_API_KEY: "ollama"
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
         GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
-        GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
\ No newline at end of file
+        GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
+        LLM_PROVIDER: "ollama"
+        LLM_ENDPOINT: "http://localhost:11434"
+        LLM_MODEL: "ollama/llama3.2"
+        EMBEDDING_PROVIDER: "ollama"
+        EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
+        EMBEDDING_ENDPOINT: "http://localhost:11434/api/embeddings"
+        EMBEDDING_DIMENSIONS: 4096
+        HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
\ No newline at end of file

From 8ebd9a7d8a0db78b3058c5a24c76c67cffcfa7fb Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Wed, 19 Feb 2025 18:10:49 -0800
Subject: [PATCH 03/48] Ollama test end to end

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 2473de3c2d..0ff0517b42 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -56,10 +56,10 @@ jobs:
       with:
         example-location: ./examples/python/simple_example.py
       secrets:
-        LLM_API_KEY: "ollama"
         OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
         GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
         GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
+        LLM_API_KEY: "ollama"
         LLM_PROVIDER: "ollama"
         LLM_ENDPOINT: "http://localhost:11434"
         LLM_MODEL: "ollama/llama3.2"

From 9d0d96eb1cd5a9262c9605b35491b32d00acc358 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 13:43:04 -0800
Subject: [PATCH 04/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 83 ++++++++++++++++++++++---------
 1 file changed, 59 insertions(+), 24 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 0ff0517b42..d85de95ee6 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -6,7 +6,7 @@ on:
     types: [ labeled, synchronize ]
 
 jobs:
-  update-insights:
+  setup-ollama:
     runs-on: ubuntu-latest
     services:
       ollama:
@@ -17,8 +17,6 @@ jobs:
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
-        with:
-          token: ${{ secrets.GH_PAT }} # Use PAT instead of default token
 
       - name: Set up Python
         uses: actions/setup-python@v4
@@ -39,32 +37,69 @@ jobs:
       - name: Wait for Ollama to be ready
         run: |
           for i in {1..30}; do
-            curl -s http://localhost:11434/api/tags && break
+            if curl -s http://localhost:11434/api/tags > /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 Model
+      - name: Pull Models
         run: |
           curl -X POST http://localhost:11434/api/pull -d '{"name": "ollama/llama3.2"}'
-      - name: Pull Embedding Model
-        run: |
           curl -X POST http://localhost:11434/api/pull -d '{"name": "avr/sfr-embedding-mistral:latest"}'
-          
 
   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 }}
-        LLM_API_KEY: "ollama"
-        LLM_PROVIDER: "ollama"
-        LLM_ENDPOINT: "http://localhost:11434"
-        LLM_MODEL: "ollama/llama3.2"
-        EMBEDDING_PROVIDER: "ollama"
-        EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-        EMBEDDING_ENDPOINT: "http://localhost:11434/api/embeddings"
-        EMBEDDING_DIMENSIONS: 4096
-        HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
\ No newline at end of file
+    needs: setup-ollama
+    runs-on: ubuntu-latest
+    services:
+      ollama:
+        image: ollama/ollama
+        ports:
+          - 11434:11434
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: "3.x"
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install -r requirements.txt
+
+      - 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
+              echo "Ollama is ready"
+              exit 0
+            fi
+            echo "Waiting for Ollama... attempt $i"
+            sleep 2
+          done
+          echo "Ollama failed to start"
+          exit 1
+
+      - name: Run example test
+        env:
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+          GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
+          GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
+          LLM_API_KEY: "ollama"
+          LLM_PROVIDER: "ollama"
+          LLM_ENDPOINT: "http://localhost:11434"
+          LLM_MODEL: "ollama/llama3.2"
+          EMBEDDING_PROVIDER: "ollama"
+          EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
+          EMBEDDING_ENDPOINT: "http://localhost:11434/api/embeddings"
+          EMBEDDING_DIMENSIONS: "4096"
+          HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
+        run: python ./examples/python/simple_example.py
\ No newline at end of file

From b4088bec29d8a1881f84f73ffcabac0c88cec056 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 13:51:30 -0800
Subject: [PATCH 05/48] Fix ollama

---
 .github/workflows/upgrade_deps.yml | 23 ++++++++++++++++++++++-
 1 file changed, 22 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/upgrade_deps.yml b/.github/workflows/upgrade_deps.yml
index 934a3ed6a7..327ee08cc7 100644
--- a/.github/workflows/upgrade_deps.yml
+++ b/.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:

From b670697b8bbbef2bbebe1acbdea4dd0239c6f90a Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 13:55:06 -0800
Subject: [PATCH 06/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 1 +
 1 file changed, 1 insertion(+)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index d85de95ee6..d6e9d41755 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -74,6 +74,7 @@ jobs:
         run: |
           python -m pip install --upgrade pip
           pip install -r requirements.txt
+          pip install ollama-python requests
 
       - name: Wait for Ollama to be ready
         run: |

From bfe039d2aaa103645cfd81e86ff9b63c4b36221a Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 13:59:56 -0800
Subject: [PATCH 07/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 1 -
 1 file changed, 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index d6e9d41755..1ac8377e58 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -73,7 +73,6 @@ jobs:
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install -r requirements.txt
           pip install ollama-python requests
 
       - name: Wait for Ollama to be ready

From 6bc4f6ae24a6fb65dfe44ccdd0c1735df7510d9a Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 14:05:27 -0800
Subject: [PATCH 08/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 1ac8377e58..939d10ce0b 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -32,7 +32,9 @@ jobs:
             ${{ runner.os }}-pip-
 
       - name: Install dependencies
-        run: pip install -r requirements.txt
+        run: |
+          python -m pip install --upgrade pip
+          pip install ollama-python requests
 
       - name: Wait for Ollama to be ready
         run: |

From 96adcfbe4e74e2c38108c9406a08e7e784bef53c Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 14:16:10 -0800
Subject: [PATCH 09/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 939d10ce0b..5e60a7b8fb 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -67,10 +67,21 @@ jobs:
       - name: Checkout repository
         uses: actions/checkout@v4
 
-      - name: Set up Python
-        uses: actions/setup-python@v4
+      - name: Setup Python
+        uses: actions/setup-python@v5
         with:
-          python-version: "3.x"
+          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
 
       - name: Install dependencies
         run: |

From c06c28d7a63ce8c8fbe50eb7a92bad7e000fd901 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 14:27:35 -0800
Subject: [PATCH 10/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 5e60a7b8fb..71cedf94da 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -115,4 +115,4 @@ jobs:
           EMBEDDING_ENDPOINT: "http://localhost:11434/api/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
-        run: python ./examples/python/simple_example.py
\ No newline at end of file
+        run: poetry run python ./examples/python/simple_example.py
\ No newline at end of file

From edd681f7de2beee66ac64df6e5417eef0b18121b Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 14:46:58 -0800
Subject: [PATCH 11/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 71cedf94da..908a9eb894 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -101,6 +101,12 @@ jobs:
           echo "Ollama failed to start"
           exit 1
 
+      - 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 }}

From 02b0109d0ed9e3832cf132546e89185d485c7ae1 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 15:36:04 -0800
Subject: [PATCH 12/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 908a9eb894..5fabbb59b9 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -86,7 +86,8 @@ jobs:
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install ollama-python requests
+          poetry install ollama-python requests
+          poetry install transformers
 
       - name: Wait for Ollama to be ready
         run: |

From a91e83eafaac717509402cd62515080718138b77 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 15:48:38 -0800
Subject: [PATCH 13/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 5fabbb59b9..6456f711eb 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -85,7 +85,6 @@ jobs:
 
       - name: Install dependencies
         run: |
-          python -m pip install --upgrade pip
           poetry install ollama-python requests
           poetry install transformers
 
@@ -113,6 +112,7 @@ jobs:
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
           GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
           GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
+          PYTHONFAULTHANDLER: 1
           LLM_API_KEY: "ollama"
           LLM_PROVIDER: "ollama"
           LLM_ENDPOINT: "http://localhost:11434"

From 326c41843f281d68f68f6b4277c32fb6c09b3517 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 15:56:30 -0800
Subject: [PATCH 14/48] Fix ollama

---
 .github/workflows/test_ollama.yml           | 12 +---
 cognee/infrastructure/llm/ollama/adapter.py | 65 ++++++++++++++++++++-
 2 files changed, 64 insertions(+), 13 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 6456f711eb..b66de43861 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -21,20 +21,12 @@ jobs:
       - name: Set up Python
         uses: actions/setup-python@v4
         with:
-          python-version: "3.x"
-
-      - name: Cache pip packages
-        uses: actions/cache@v4
-        with:
-          path: ~/.cache/pip
-          key: ${{ runner.os }}-pip-${{ hashFiles('**/requirements.txt') }}
-          restore-keys: |
-            ${{ runner.os }}-pip-
+          python-version: "3.12.x"
 
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install ollama-python requests
+          pip install requests
 
       - name: Wait for Ollama to be ready
         run: |
diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index 4eb3927394..ed0ef31bde 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -1,20 +1,28 @@
+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
 
 class OllamaAPIAdapter(LLMInterface):
-    """Adapter for a Generic API LLM provider using instructor with an OpenAI backend."""
+    """Adapter for a Ollama API LLM provider using instructor with an OpenAI backend."""
+
+    api_version: str
+
+    MAX_RETRIES = 5
 
-    def __init__(self, endpoint: str, api_key: str, model: str, name: str, max_tokens: int):
+    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.aclient = instructor.from_openai(
             OpenAI(base_url=self.endpoint, api_key=self.api_key), mode=instructor.Mode.JSON
@@ -42,3 +50,54 @@ 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.")
+
+        # with open(input, 'rb') as audio_file:
+        #     audio_data = audio_file.read()
+
+        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,
+        )
+
+        return transcription
+
+    def transcribe_image(self, input) -> BaseModel:
+        with open(input, "rb") as image_file:
+            encoded_image = base64.b64encode(image_file.read()).decode("utf-8")
+
+        return self.aclient.completion(
+            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}",
+                            },
+                        },
+                    ],
+                }
+            ],
+            api_key=self.api_key,
+            api_base=self.endpoint,
+            api_version=self.api_version,
+            max_tokens=300,
+            max_retries=self.MAX_RETRIES,
+        )
\ No newline at end of file

From 92602aa34f8fa53bb5d089eaab4d82ede46bdff1 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:02:08 -0800
Subject: [PATCH 15/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index b66de43861..bd7da456f7 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -77,7 +77,7 @@ jobs:
 
       - name: Install dependencies
         run: |
-          poetry install ollama-python requests
+          poetry install requests
           poetry install transformers
 
       - name: Wait for Ollama to be ready

From f2d09090e1d96c71bb266fabca00da91949d3545 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:06:05 -0800
Subject: [PATCH 16/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index bd7da456f7..222f8cb70c 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -75,11 +75,6 @@ jobs:
         run: |
           poetry install --no-interaction --all-extras
 
-      - name: Install dependencies
-        run: |
-          poetry install requests
-          poetry install transformers
-
       - name: Wait for Ollama to be ready
         run: |
           for i in {1..30}; do
@@ -107,7 +102,7 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_API_KEY: "ollama"
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://localhost:11434"
+          LLM_ENDPOINT: "http://localhost:11434/api/generate"
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"

From 97465f168aec3189022c57419ccf83653d2ac7a7 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:06:56 -0800
Subject: [PATCH 17/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 222f8cb70c..e18b4dee60 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -19,14 +19,20 @@ jobs:
         uses: actions/checkout@v4
 
       - name: Set up Python
-        uses: actions/setup-python@v4
+        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: |
-          python -m pip install --upgrade pip
-          pip install requests
+          poetry install --no-interaction --all-extras
 
       - name: Wait for Ollama to be ready
         run: |

From 73662b8b01d64f5d54b2f1545e51eb0ee59925da Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:12:20 -0800
Subject: [PATCH 18/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index e18b4dee60..bddf708dc4 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -108,7 +108,7 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_API_KEY: "ollama"
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://localhost:11434/api/generate"
+          LLM_ENDPOINT: "http://127.0.0.1:11434"
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"

From 90d96aa6ad8129cf415d27bec87d7e4bd4b94303 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:17:50 -0800
Subject: [PATCH 19/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index bddf708dc4..edfd289104 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -49,7 +49,7 @@ jobs:
 
       - name: Pull Models
         run: |
-          curl -X POST http://localhost:11434/api/pull -d '{"name": "ollama/llama3.2"}'
+          curl -X POST http://localhost:11434/api/pull -d '{"name": "llama3.2"}'
           curl -X POST http://localhost:11434/api/pull -d '{"name": "avr/sfr-embedding-mistral:latest"}'
 
   run_simple_example_test:

From 3a88b94d0185424c007fad794475c92d65e7e640 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:25:08 -0800
Subject: [PATCH 20/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index edfd289104..35e06311b7 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -112,7 +112,7 @@ jobs:
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "http://localhost:11434/api/embeddings"
+          EMBEDDING_ENDPOINT: "http://127.0.0.1:11434/api/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py
\ No newline at end of file

From 11442df1aaed717e74da5bccbe3908a78501b1e8 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:41:42 -0800
Subject: [PATCH 21/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 50 ++++++++-----------------------
 1 file changed, 12 insertions(+), 38 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 35e06311b7..5a1af17078 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -8,44 +8,7 @@ on:
 jobs:
   setup-ollama:
     runs-on: ubuntu-latest
-    services:
-      ollama:
-        image: ollama/ollama
-        ports:
-          - 11434:11434
-
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Set up 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
-
-      - 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
-              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 Models
         run: |
@@ -53,7 +16,7 @@ jobs:
           curl -X POST http://localhost:11434/api/pull -d '{"name": "avr/sfr-embedding-mistral:latest"}'
 
   run_simple_example_test:
-    needs: setup-ollama
+
     runs-on: ubuntu-latest
     services:
       ollama:
@@ -80,6 +43,17 @@ jobs:
       - name: Install dependencies
         run: |
           poetry install --no-interaction --all-extras
+        steps:
+      - 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
+        run: |
+          curl -d '{"model": "llama3.2", "stream": false, "prompt":"Whatever I say, asnwer with Yes"}' http://localhost:11434/api/generate
 
       - name: Wait for Ollama to be ready
         run: |

From 1dfb0dd58b7dd5e8ec6a8dbd268e006a342b0b80 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:42:09 -0800
Subject: [PATCH 22/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 5a1af17078..15e1b3bdf0 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -6,14 +6,6 @@ on:
     types: [ labeled, synchronize ]
 
 jobs:
-  setup-ollama:
-    runs-on: ubuntu-latest
-
-
-      - name: Pull Models
-        run: |
-          curl -X POST http://localhost:11434/api/pull -d '{"name": "llama3.2"}'
-          curl -X POST http://localhost:11434/api/pull -d '{"name": "avr/sfr-embedding-mistral:latest"}'
 
   run_simple_example_test:
 
@@ -43,7 +35,7 @@ jobs:
       - name: Install dependencies
         run: |
           poetry install --no-interaction --all-extras
-        steps:
+
       - name: Install ollama
         run: curl -fsSL https://ollama.com/install.sh | sh
       - name: Run ollama

From 4c4723b7e0e3d4a00c4ba8c2f3ca547106e4e2a9 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 16:55:41 -0800
Subject: [PATCH 23/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 15e1b3bdf0..976220d74e 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -74,7 +74,7 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_API_KEY: "ollama"
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://127.0.0.1:11434"
+          LLM_ENDPOINT: "http://127.0.0.1:11434/v1/chat/completions"
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"

From 846c45e54dac5bb3daa23df083f265454d967356 Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 17:18:05 -0800
Subject: [PATCH 24/48] Fix ollama

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 976220d74e..f0312a81e6 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -74,7 +74,7 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_API_KEY: "ollama"
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://127.0.0.1:11434/v1/chat/completions"
+          LLM_ENDPOINT: "http://127.0.0.1:11434/api/generate"
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"

From 2c0bfc87a6d15d324527a9a3285d2254907ad0ba Mon Sep 17 00:00:00 2001
From: vasilije <vas.markovic@gmail.com>
Date: Fri, 21 Feb 2025 17:45:29 -0800
Subject: [PATCH 25/48] Fix ollama

---
 .github/workflows/test_gemini.yml | 33 +++++++++++++++++++++++++++++++
 1 file changed, 33 insertions(+)
 create mode 100644 .github/workflows/test_gemini.yml

diff --git a/.github/workflows/test_gemini.yml b/.github/workflows/test_gemini.yml
new file mode 100644
index 0000000000..e5bf0520f3
--- /dev/null
+++ b/.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"

From 5c7b4a5b1e1eea8f314ef4b5c5d74f504960c9c6 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Tue, 25 Feb 2025 12:21:44 +0100
Subject: [PATCH 26/48] Ruff it.

---
 cognee/infrastructure/llm/ollama/adapter.py | 84 +++++++++------------
 1 file changed, 37 insertions(+), 47 deletions(-)

diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index ed0ef31bde..d14c8a57f6 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -1,28 +1,33 @@
-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
 
-class OllamaAPIAdapter(LLMInterface):
-    """Adapter for a Ollama API LLM provider using instructor with an OpenAI backend."""
 
-    api_version: str
+class OllamaAPIAdapter(LLMInterface):
+    """Adapter for an Ollama API LLM provider using instructor with an OpenAI backend."""
 
     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
@@ -33,71 +38,56 @@ async def acreate_structured_output(
     ) -> BaseModel:
         """Generate a structured output from the LLM using the provided text and system prompt."""
 
-        response = self.aclient.chat.completions.create(
+        response = await self.aclient.chat.completions.create(
             model=self.model,
             messages=[
-                {
-                    "role": "user",
-                    "content": f"Use the given format to extract information from the following input: {text_input}",
-                },
-                {
-                    "role": "system",
-                    "content": system_prompt,
-                },
+                {"role": "system", "content": system_prompt},
+                {"role": "user", "content": text_input},
             ],
-            max_retries=5,
-            response_model=response_model,
+            max_tokens=self.max_tokens,
         )
 
-        return response
+        return response_model.parse_raw(response.choices[0].message.content)
 
-
-    def create_transcript(self, input):
-        """Generate a audio transcript from a user query."""
+    def create_transcript(self, input: str):
+        """Generate an audio transcript from a user query."""
 
         if not os.path.isfile(input):
             raise FileNotFoundError(f"The file {input} does not exist.")
 
-        # with open(input, 'rb') as audio_file:
-        #     audio_data = audio_file.read()
+        with open(input, "rb") as audio_file:
+            transcription = self.aclient.audio.transcriptions.create(
+                model="whisper-1",  # Ensure the correct model for transcription
+                file=audio_file,
+                language="en",
+            )
 
-        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,
-        )
+        return transcription.text
 
-        return transcription
+    def transcribe_image(self, input: str) -> str:
+        """Transcribe content from an image using base64 encoding."""
+
+        if not os.path.isfile(input):
+            raise FileNotFoundError(f"The file {input} does not exist.")
 
-    def transcribe_image(self, input) -> BaseModel:
         with open(input, "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,
-        )
\ No newline at end of file
+        )
+
+        return response.choices[0].message.content

From 0bba1f8261a91a3a9559529887f23e1104b2c721 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Tue, 25 Feb 2025 13:02:46 +0100
Subject: [PATCH 27/48] Response model fun.

---
 cognee/infrastructure/llm/ollama/adapter.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index d14c8a57f6..c41f3da7fa 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -45,9 +45,10 @@ async def acreate_structured_output(
                 {"role": "user", "content": text_input},
             ],
             max_tokens=self.max_tokens,
+            response_model=response_model,
         )
 
-        return response_model.parse_raw(response.choices[0].message.content)
+        return response
 
     def create_transcript(self, input: str):
         """Generate an audio transcript from a user query."""

From 061fbbdc5861dee2b888a57fe850a1670ce6573b Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Tue, 25 Feb 2025 13:35:42 +0100
Subject: [PATCH 28/48] OpenAI mode.

---
 .github/workflows/test_ollama.yml | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index f0312a81e6..f4534e559a 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -40,17 +40,17 @@ jobs:
         run: curl -fsSL https://ollama.com/install.sh | sh
       - name: Run ollama
         run: |
-          ollama serve &
+          ollama serve --openai &
           ollama pull llama3.2 &
           ollama pull avr/sfr-embedding-mistral:latest
       - name: Call ollama API
         run: |
-          curl -d '{"model": "llama3.2", "stream": false, "prompt":"Whatever I say, asnwer with Yes"}' http://localhost:11434/api/generate
+          curl -d '{"model": "llama3.2", "stream": false, "prompt":"Whatever I say, asnwer with Yes"}' http://localhost:11434/v1/chat/completions
 
       - 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
@@ -74,11 +74,11 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_API_KEY: "ollama"
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://127.0.0.1:11434/api/generate"
+          LLM_ENDPOINT: "http://127.0.0.1:11434/v1/chat/completions"
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "http://127.0.0.1:11434/api/embeddings"
+          EMBEDDING_ENDPOINT: "http://127.0.0.1:11434//v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
-        run: poetry run python ./examples/python/simple_example.py
\ No newline at end of file
+        run: poetry run python ./examples/python/simple_example.py

From 0ed6aa6ba4913c4abe90cb07775d2779b3a02cf2 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Tue, 25 Feb 2025 14:11:14 +0100
Subject: [PATCH 29/48] Typo.

---
 .github/workflows/test_ollama.yml | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index f4534e559a..d6d8c30747 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -45,7 +45,16 @@ jobs:
           ollama pull avr/sfr-embedding-mistral:latest
       - name: Call ollama API
         run: |
-          curl -d '{"model": "llama3.2", "stream": false, "prompt":"Whatever I say, asnwer with Yes"}' http://localhost:11434/v1/chat/completions
+          curl -X POST http://localhost:11434/v1/chat/completions \
+            -H "Content-Type: application/json" \
+            -d '{
+              "model": "llama3.2",
+              "stream": false,
+              "messages": [
+                { "role": "system", "content": "You are a helpful assistant." },
+                { "role": "user", "content": "Whatever I say, answer with Yes." }
+              ]
+            }'
 
       - name: Wait for Ollama to be ready
         run: |
@@ -78,7 +87,7 @@ jobs:
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "http://127.0.0.1:11434//v1/embeddings"
+          EMBEDDING_ENDPOINT: "http://127.0.0.1:11434/v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From 30903336ff72cd5eec5c099699dac4345e923a5f Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Tue, 25 Feb 2025 14:53:56 +0100
Subject: [PATCH 30/48] Add a call, homogenous localhost.

---
 .github/workflows/test_ollama.yml | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index d6d8c30747..16d0a91840 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -55,6 +55,13 @@ jobs:
                 { "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: Wait for Ollama to be ready
         run: |
@@ -83,11 +90,11 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_API_KEY: "ollama"
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://127.0.0.1:11434/v1/chat/completions"
+          LLM_ENDPOINT: "http://localhost:11434/v1/chat/completions"
           LLM_MODEL: "ollama/llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "http://127.0.0.1:11434/v1/embeddings"
+          EMBEDDING_ENDPOINT: "http://localhost:11434/v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From 80ccf558770d8ac1dae721cf6360f20028f821ec Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Tue, 25 Feb 2025 15:51:51 +0100
Subject: [PATCH 31/48] Should conform more.

---
 .github/workflows/test_ollama.yml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 16d0a91840..739e845986 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -88,10 +88,10 @@ jobs:
           GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
           GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
           PYTHONFAULTHANDLER: 1
-          LLM_API_KEY: "ollama"
+          LLM_API_KEY: ""
           LLM_PROVIDER: "ollama"
           LLM_ENDPOINT: "http://localhost:11434/v1/chat/completions"
-          LLM_MODEL: "ollama/llama3.2"
+          LLM_MODEL: "llama3"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
           EMBEDDING_ENDPOINT: "http://localhost:11434/v1/embeddings"

From ce8c2da4e72d3e9d02b481823dfae2067008e5c1 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Tue, 25 Feb 2025 15:59:44 +0100
Subject: [PATCH 32/48] Unset, my friend, unset.

---
 .github/workflows/test_ollama.yml | 1 -
 1 file changed, 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 739e845986..feb43a46ed 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -88,7 +88,6 @@ jobs:
           GRAPHISTRY_USERNAME: ${{ secrets.GRAPHISTRY_USERNAME }}
           GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
           PYTHONFAULTHANDLER: 1
-          LLM_API_KEY: ""
           LLM_PROVIDER: "ollama"
           LLM_ENDPOINT: "http://localhost:11434/v1/chat/completions"
           LLM_MODEL: "llama3"

From 65927b39cd2390b9ff6b7f3ea84ff6cee4a76cde Mon Sep 17 00:00:00 2001
From: Vasilije <8619304+Vasilije1990@users.noreply.github.com>
Date: Tue, 25 Feb 2025 19:41:16 +0100
Subject: [PATCH 33/48] Update test_ollama.yml

---
 .github/workflows/test_ollama.yml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index feb43a46ed..b9eea184d6 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -89,11 +89,11 @@ jobs:
           GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
           PYTHONFAULTHANDLER: 1
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "http://localhost:11434/v1/chat/completions"
+          LLM_ENDPOINT: "ollama:11434/v1/chat/completions"
           LLM_MODEL: "llama3"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "http://localhost:11434/v1/embeddings"
+          EMBEDDING_ENDPOINT: "ollama:11434/v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From 6463c2e08be8846d842656749917c544c7025596 Mon Sep 17 00:00:00 2001
From: Vasilije <8619304+Vasilije1990@users.noreply.github.com>
Date: Tue, 25 Feb 2025 19:47:57 +0100
Subject: [PATCH 34/48] Update test_ollama.yml

---
 .github/workflows/test_ollama.yml | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index b9eea184d6..fc5c498b50 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -90,6 +90,8 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_PROVIDER: "ollama"
           LLM_ENDPOINT: "ollama:11434/v1/chat/completions"
+          LLM_API_KEY: "ollama"
+          EMBEDDING_API_KEY: "ollama"
           LLM_MODEL: "llama3"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"

From 70f9b5f7c55e42c1c810d5969c0cbb897cefc80e Mon Sep 17 00:00:00 2001
From: Vasilije <8619304+Vasilije1990@users.noreply.github.com>
Date: Tue, 25 Feb 2025 19:54:48 +0100
Subject: [PATCH 35/48] Update test_ollama.yml

---
 .github/workflows/test_ollama.yml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index fc5c498b50..0acfe45b3e 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -89,13 +89,13 @@ jobs:
           GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
           PYTHONFAULTHANDLER: 1
           LLM_PROVIDER: "ollama"
-          LLM_ENDPOINT: "ollama:11434/v1/chat/completions"
+          LLM_ENDPOINT: "http://ollama:11434/v1/chat/completions"
           LLM_API_KEY: "ollama"
           EMBEDDING_API_KEY: "ollama"
           LLM_MODEL: "llama3"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "ollama:11434/v1/embeddings"
+          EMBEDDING_ENDPOINT: http://"ollama:11434/v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From 468268cc8f69e4999486a1ced004376548cb0d95 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 10:33:02 +0100
Subject: [PATCH 36/48] Docker Composish way.

---
 .github/workflows/test_ollama.yml | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 0acfe45b3e..8a778b62be 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -90,12 +90,10 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_PROVIDER: "ollama"
           LLM_ENDPOINT: "http://ollama:11434/v1/chat/completions"
-          LLM_API_KEY: "ollama"
-          EMBEDDING_API_KEY: "ollama"
           LLM_MODEL: "llama3"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: http://"ollama:11434/v1/embeddings"
+          EMBEDDING_ENDPOINT: "http://ollama:11434/v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From c72b12d7129d8767dcd86bac237e1775ee30be24 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 10:40:47 +0100
Subject: [PATCH 37/48] Let's be Pydantic.

---
 .github/workflows/test_ollama.yml | 1 +
 1 file changed, 1 insertion(+)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 8a778b62be..84e8059baf 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -89,6 +89,7 @@ jobs:
           GRAPHISTRY_PASSWORD: ${{ secrets.GRAPHISTRY_PASSWORD }}
           PYTHONFAULTHANDLER: 1
           LLM_PROVIDER: "ollama"
+          LLM_API_KEY: "ollama"
           LLM_ENDPOINT: "http://ollama:11434/v1/chat/completions"
           LLM_MODEL: "llama3"
           EMBEDDING_PROVIDER: "ollama"

From c22455675c6056f6454e20de10caf88658f4bffa Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 11:33:23 +0100
Subject: [PATCH 38/48] Launch Docker manually.

---
 .github/workflows/test_ollama.yml | 64 ++++++++++++++++++-------------
 1 file changed, 37 insertions(+), 27 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 84e8059baf..9b55ba62c9 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -10,11 +10,11 @@ jobs:
   run_simple_example_test:
 
     runs-on: ubuntu-latest
-    services:
-      ollama:
-        image: ollama/ollama
-        ports:
-          - 11434:11434
+#    services:
+#      ollama:
+#        image: ollama/ollama
+#        ports:
+#          - 11434:11434
 
     steps:
       - name: Checkout repository
@@ -36,13 +36,37 @@ jobs:
         run: |
           poetry install --no-interaction --all-extras
 
-      - name: Install ollama
-        run: curl -fsSL https://ollama.com/install.sh | sh
-      - name: Run ollama
+#      - 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 in a Docker container
+        run: |
+          docker run -d --name ollama -p 11434:11434 ollama/ollama serve --openai
+          sleep 5  # Give some time for Ollama to start
+
+      - 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: |
-          ollama serve --openai &
-          ollama pull llama3.2 &
-          ollama pull avr/sfr-embedding-mistral:latest
+          curl -X POST http://localhost:11434/api/pull -d '{"name": "llama3.2"}'
+          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 \
@@ -62,20 +86,6 @@ jobs:
             "input": "This is a test sentence to generate an embedding."
             }'
 
-
-      - 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: Dump Docker logs
         run: |
           docker ps
@@ -90,11 +100,11 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_PROVIDER: "ollama"
           LLM_API_KEY: "ollama"
-          LLM_ENDPOINT: "http://ollama:11434/v1/chat/completions"
+          LLM_ENDPOINT: "http://localhost:11434/v1/chat/completions"
           LLM_MODEL: "llama3"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "http://ollama:11434/v1/embeddings"
+          EMBEDDING_ENDPOINT: "http://localhost:11434/v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
           HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From ec9bbca79856def1a5c09dbbb594c920568b1cbe Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 11:46:24 +0100
Subject: [PATCH 39/48] Cosmetics.

---
 .github/workflows/test_ollama.yml | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 9b55ba62c9..dd3de932ca 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -44,10 +44,14 @@ jobs:
 #          ollama pull llama3.2 &
 #          ollama pull avr/sfr-embedding-mistral:latest
 
-      - name: Start Ollama in a Docker container
+      - name: Start Ollama container
         run: |
-          docker run -d --name ollama -p 11434:11434 ollama/ollama serve --openai
-          sleep 5  # Give some time for Ollama to start
+          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: |

From cabbfd628931bc45124956ef1ddbdda49824ecdd Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 11:57:17 +0100
Subject: [PATCH 40/48] Maybe we could fly without the Hugger.

---
 .github/workflows/test_ollama.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index dd3de932ca..16692ae206 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -110,5 +110,5 @@ jobs:
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
           EMBEDDING_ENDPOINT: "http://localhost:11434/v1/embeddings"
           EMBEDDING_DIMENSIONS: "4096"
-          HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
+#          HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From c329cefcda02242b6d158dece0b245017fcf8051 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 14:02:29 +0100
Subject: [PATCH 41/48] OHMY.

---
 .github/workflows/test_ollama.yml | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index 16692ae206..ceffa08b02 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -104,11 +104,11 @@ jobs:
           PYTHONFAULTHANDLER: 1
           LLM_PROVIDER: "ollama"
           LLM_API_KEY: "ollama"
-          LLM_ENDPOINT: "http://localhost:11434/v1/chat/completions"
-          LLM_MODEL: "llama3"
+          LLM_ENDPOINT: "http://localhost:11434/v1/"
+          LLM_MODEL: "llama3.2"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
-          EMBEDDING_ENDPOINT: "http://localhost:11434/v1/embeddings"
+          EMBEDDING_ENDPOINT: "http://localhost:11434/v1/"
           EMBEDDING_DIMENSIONS: "4096"
-#          HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
+          HUGGINGFACE_TOKENIZER: "Salesforce/SFR-Embedding-Mistral"
         run: poetry run python ./examples/python/simple_example.py

From 44f02df0c8a5b6ccad15fa5f453b2fe06878c217 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 14:16:10 +0100
Subject: [PATCH 42/48] Async it.

---
 .github/workflows/test_ollama.yml           |  1 +
 cognee/infrastructure/llm/ollama/adapter.py | 45 ++++++++++++++-------
 2 files changed, 31 insertions(+), 15 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index ceffa08b02..f2e33c5d47 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -35,6 +35,7 @@ jobs:
       - 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
diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index c41f3da7fa..7ec0a83f07 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -1,10 +1,9 @@
-from typing import Type
+from typing import Type, Optional
 from pydantic import BaseModel
 import instructor
 from cognee.infrastructure.llm.llm_interface import LLMInterface
-from openai import OpenAI
+from openai import AsyncOpenAI  # Use AsyncOpenAI for async compatibility
 import base64
-from pathlib import Path
 import os
 
 
@@ -20,7 +19,7 @@ def __init__(
         model: str,
         name: str,
         max_tokens: int,
-        api_version: str = None,
+        api_version: Optional[str] = None,
     ) -> None:
         self.name = name
         self.model = model
@@ -29,8 +28,9 @@ def __init__(
         self.max_tokens = max_tokens
         self.api_version = api_version
 
+        # Use AsyncOpenAI for proper async handling
         self.aclient = instructor.from_openai(
-            OpenAI(base_url=self.endpoint, api_key=self.api_key), mode=instructor.Mode.JSON
+            AsyncOpenAI(base_url=self.endpoint, api_key=self.api_key), mode=instructor.Mode.JSON
         )
 
     async def acreate_structured_output(
@@ -38,6 +38,10 @@ async def acreate_structured_output(
     ) -> BaseModel:
         """Generate a structured output from the LLM using the provided text and system prompt."""
 
+        # Ensure the function being awaited is actually async
+        if not callable(getattr(self.aclient.chat.completions, "create", None)):
+            raise TypeError("self.aclient.chat.completions.create is not callable!")
+
         response = await self.aclient.chat.completions.create(
             model=self.model,
             messages=[
@@ -45,33 +49,40 @@ async def acreate_structured_output(
                 {"role": "user", "content": text_input},
             ],
             max_tokens=self.max_tokens,
-            response_model=response_model,
         )
 
-        return response
+        # Ensure response is valid before passing to Pydantic model
+        if not isinstance(response, dict):
+            raise ValueError(f"Unexpected response format: {response}")
+
+        return response_model(**response)
 
-    def create_transcript(self, input: str):
+    def create_transcript(self, input_file: str) -> str:
         """Generate an audio transcript from a user query."""
 
-        if not os.path.isfile(input):
-            raise FileNotFoundError(f"The file {input} does not exist.")
+        if not os.path.isfile(input_file):
+            raise FileNotFoundError(f"The file {input_file} does not exist.")
 
-        with open(input, "rb") as audio_file:
+        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: str) -> str:
+    def transcribe_image(self, input_file: str) -> str:
         """Transcribe content from an image using base64 encoding."""
 
-        if not os.path.isfile(input):
-            raise FileNotFoundError(f"The file {input} does not exist.")
+        if not os.path.isfile(input_file):
+            raise FileNotFoundError(f"The file {input_file} does not exist.")
 
-        with open(input, "rb") as image_file:
+        with open(input_file, "rb") as image_file:
             encoded_image = base64.b64encode(image_file.read()).decode("utf-8")
 
         response = self.aclient.chat.completions.create(
@@ -91,4 +102,8 @@ def transcribe_image(self, input: str) -> str:
             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

From 6b490789639851d6fd392adabf4f43dca9027dbc Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 14:25:16 +0100
Subject: [PATCH 43/48] Response model.

---
 cognee/infrastructure/llm/ollama/adapter.py | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index 7ec0a83f07..fa5023a1c2 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -38,10 +38,11 @@ async def acreate_structured_output(
     ) -> BaseModel:
         """Generate a structured output from the LLM using the provided text and system prompt."""
 
-        # Ensure the function being awaited is actually async
+        # Ensure the API method is async
         if not callable(getattr(self.aclient.chat.completions, "create", None)):
             raise TypeError("self.aclient.chat.completions.create is not callable!")
 
+        # Call the API with the expected parameters
         response = await self.aclient.chat.completions.create(
             model=self.model,
             messages=[
@@ -51,11 +52,12 @@ async def acreate_structured_output(
             max_tokens=self.max_tokens,
         )
 
-        # Ensure response is valid before passing to Pydantic model
+        # Ensure the response is valid before passing to Pydantic model
         if not isinstance(response, dict):
             raise ValueError(f"Unexpected response format: {response}")
 
-        return response_model(**response)
+        # Use instructor's parse_response method instead of passing response_model in API call
+        return instructor.parse_response(response, response_model)
 
     def create_transcript(self, input_file: str) -> str:
         """Generate an audio transcript from a user query."""

From fe7da6028b3e3bf836486ce27346e101a8724bf7 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 14:41:23 +0100
Subject: [PATCH 44/48] Will graph fly.

---
 cognee/infrastructure/llm/ollama/adapter.py | 46 +++++----------------
 1 file changed, 10 insertions(+), 36 deletions(-)

diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index fa5023a1c2..1390d812ca 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -2,7 +2,7 @@
 from pydantic import BaseModel
 import instructor
 from cognee.infrastructure.llm.llm_interface import LLMInterface
-from openai import AsyncOpenAI  # Use AsyncOpenAI for async compatibility
+from openai import AsyncOpenAI
 import base64
 import os
 
@@ -28,55 +28,29 @@ def __init__(
         self.max_tokens = max_tokens
         self.api_version = api_version
 
-        # Use AsyncOpenAI for proper async handling
-        self.aclient = instructor.from_openai(
-            AsyncOpenAI(base_url=self.endpoint, api_key=self.api_key), mode=instructor.Mode.JSON
-        )
+        # Properly initialize AsyncOpenAI
+        self.client = AsyncOpenAI(base_url=self.endpoint, api_key=self.api_key)
+
+        # Apply instructor patch (this enables structured output support)
+        instructor.patch(self.client)
 
     async def acreate_structured_output(
         self, text_input: str, system_prompt: str, response_model: Type[BaseModel]
     ) -> BaseModel:
         """Generate a structured output from the LLM using the provided text and system prompt."""
 
-        # Ensure the API method is async
-        if not callable(getattr(self.aclient.chat.completions, "create", None)):
-            raise TypeError("self.aclient.chat.completions.create is not callable!")
-
-        # Call the API with the expected parameters
-        response = await self.aclient.chat.completions.create(
+        # Directly pass `response_model` inside `.create()`
+        response = await self.client.chat.completions.create(
             model=self.model,
             messages=[
                 {"role": "system", "content": system_prompt},
                 {"role": "user", "content": text_input},
             ],
             max_tokens=self.max_tokens,
+            response_model=response_model,  # This works after `instructor.patch(self.client)`
         )
 
-        # Ensure the response is valid before passing to Pydantic model
-        if not isinstance(response, dict):
-            raise ValueError(f"Unexpected response format: {response}")
-
-        # Use instructor's parse_response method instead of passing response_model in API call
-        return instructor.parse_response(response, response_model)
-
-    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
+        return response  # Already converted into `response_model`
 
     def transcribe_image(self, input_file: str) -> str:
         """Transcribe content from an image using base64 encoding."""

From a77655a7765e562b54baa062cb3b353fc05792a4 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 14:51:04 +0100
Subject: [PATCH 45/48] Oops, putting back create transcript.

---
 cognee/infrastructure/llm/ollama/adapter.py | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index 1390d812ca..7a1701bbf7 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -52,6 +52,25 @@ async def acreate_structured_output(
 
         return response  # Already converted into `response_model`
 
+    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."""
 

From cfc93e3d7aebf74f963b27dafba70a50677e8eea Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Wed, 26 Feb 2025 15:18:24 +0100
Subject: [PATCH 46/48] Clean up adapter.

---
 cognee/infrastructure/llm/ollama/adapter.py | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index 7a1701bbf7..bf79de1a15 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -5,6 +5,7 @@
 from openai import AsyncOpenAI
 import base64
 import os
+import json
 
 
 class OllamaAPIAdapter(LLMInterface):
@@ -28,18 +29,14 @@ def __init__(
         self.max_tokens = max_tokens
         self.api_version = api_version
 
-        # Properly initialize AsyncOpenAI
+        # ✅ Properly initialize AsyncOpenAI and patch with instructor
         self.client = AsyncOpenAI(base_url=self.endpoint, api_key=self.api_key)
-
-        # Apply instructor patch (this enables structured output support)
-        instructor.patch(self.client)
+        instructor.patch(self.client)  # ✅ Patch OpenAI with instructor for structured output
 
     async def acreate_structured_output(
         self, text_input: str, system_prompt: str, response_model: Type[BaseModel]
     ) -> BaseModel:
         """Generate a structured output from the LLM using the provided text and system prompt."""
-
-        # Directly pass `response_model` inside `.create()`
         response = await self.client.chat.completions.create(
             model=self.model,
             messages=[
@@ -47,10 +44,10 @@ async def acreate_structured_output(
                 {"role": "user", "content": text_input},
             ],
             max_tokens=self.max_tokens,
-            response_model=response_model,  # This works after `instructor.patch(self.client)`
+            response_model=response_model,
         )
 
-        return response  # Already converted into `response_model`
+        return response_model(response)
 
     def create_transcript(self, input_file: str) -> str:
         """Generate an audio transcript from a user query."""
@@ -60,12 +57,12 @@ def create_transcript(self, input_file: str) -> str:
 
         with open(input_file, "rb") as audio_file:
             transcription = self.aclient.audio.transcriptions.create(
-                model="whisper-1",  # ✅ Ensure the correct model for transcription
+                model="whisper-1",  # Ensure the correct model for transcription
                 file=audio_file,
                 language="en",
             )
 
-        # ✅ Ensure the response contains a valid transcript
+        # Ensure the response contains a valid transcript
         if not hasattr(transcription, "text"):
             raise ValueError("Transcription failed. No text returned.")
 

From 647d87286af5cba926bf06368de0d50f38676286 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Thu, 27 Feb 2025 12:13:04 +0100
Subject: [PATCH 47/48] Phi4 can respond reasonably.

---
 .github/workflows/test_ollama.yml           |  6 +--
 cognee/infrastructure/llm/ollama/adapter.py | 46 ++++++++++-----------
 2 files changed, 24 insertions(+), 28 deletions(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index f2e33c5d47..ada006a12b 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -69,7 +69,7 @@ jobs:
 
       - name: Pull required Ollama models
         run: |
-          curl -X POST http://localhost:11434/api/pull -d '{"name": "llama3.2"}'
+          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
@@ -77,7 +77,7 @@ jobs:
           curl -X POST http://localhost:11434/v1/chat/completions \
             -H "Content-Type: application/json" \
             -d '{
-              "model": "llama3.2",
+              "model": "phi4",
               "stream": false,
               "messages": [
                 { "role": "system", "content": "You are a helpful assistant." },
@@ -106,7 +106,7 @@ jobs:
           LLM_PROVIDER: "ollama"
           LLM_API_KEY: "ollama"
           LLM_ENDPOINT: "http://localhost:11434/v1/"
-          LLM_MODEL: "llama3.2"
+          LLM_MODEL: "phi4"
           EMBEDDING_PROVIDER: "ollama"
           EMBEDDING_MODEL: "avr/sfr-embedding-mistral:latest"
           EMBEDDING_ENDPOINT: "http://localhost:11434/v1/"
diff --git a/cognee/infrastructure/llm/ollama/adapter.py b/cognee/infrastructure/llm/ollama/adapter.py
index bf79de1a15..ac6b91af43 100644
--- a/cognee/infrastructure/llm/ollama/adapter.py
+++ b/cognee/infrastructure/llm/ollama/adapter.py
@@ -1,53 +1,49 @@
-from typing import Type, Optional
+from typing import Type
 from pydantic import BaseModel
 import instructor
 from cognee.infrastructure.llm.llm_interface import LLMInterface
-from openai import AsyncOpenAI
+from cognee.infrastructure.llm.config import get_llm_config
+from openai import OpenAI
 import base64
 import os
-import json
 
 
 class OllamaAPIAdapter(LLMInterface):
-    """Adapter for an Ollama API LLM provider using instructor with an OpenAI backend."""
-
-    MAX_RETRIES = 5
-
-    def __init__(
-        self,
-        endpoint: str,
-        api_key: str,
-        model: str,
-        name: str,
-        max_tokens: int,
-        api_version: Optional[str] = None,
-    ) -> None:
+    """Adapter for a Generic API LLM provider using instructor with an OpenAI backend."""
+
+    def __init__(self, endpoint: str, api_key: str, model: str, name: str, max_tokens: int):
         self.name = name
         self.model = model
         self.api_key = api_key
         self.endpoint = endpoint
         self.max_tokens = max_tokens
-        self.api_version = api_version
 
-        # ✅ Properly initialize AsyncOpenAI and patch with instructor
-        self.client = AsyncOpenAI(base_url=self.endpoint, api_key=self.api_key)
-        instructor.patch(self.client)  # ✅ Patch OpenAI with instructor for structured output
+        self.aclient = instructor.from_openai(
+            OpenAI(base_url=self.endpoint, api_key=self.api_key), mode=instructor.Mode.JSON
+        )
 
     async def acreate_structured_output(
         self, text_input: str, system_prompt: str, response_model: Type[BaseModel]
     ) -> BaseModel:
         """Generate a structured output from the LLM using the provided text and system prompt."""
-        response = await self.client.chat.completions.create(
+
+        response = self.aclient.chat.completions.create(
             model=self.model,
             messages=[
-                {"role": "system", "content": system_prompt},
-                {"role": "user", "content": text_input},
+                {
+                    "role": "user",
+                    "content": f"Use the given format to extract information from the following input: {text_input}",
+                },
+                {
+                    "role": "system",
+                    "content": system_prompt,
+                },
             ],
-            max_tokens=self.max_tokens,
+            max_retries=5,
             response_model=response_model,
         )
 
-        return response_model(response)
+        return response
 
     def create_transcript(self, input_file: str) -> str:
         """Generate an audio transcript from a user query."""

From 01bb8cbf0cafde849af458a23c664a69baea4f58 Mon Sep 17 00:00:00 2001
From: Daniel Molnar <soobrosa@gmail.com>
Date: Fri, 28 Feb 2025 19:06:09 +0100
Subject: [PATCH 48/48] Beefy runner.

---
 .github/workflows/test_ollama.yml | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/test_ollama.yml b/.github/workflows/test_ollama.yml
index ada006a12b..030d6e3e13 100644
--- a/.github/workflows/test_ollama.yml
+++ b/.github/workflows/test_ollama.yml
@@ -9,7 +9,8 @@ jobs:
 
   run_simple_example_test:
 
-    runs-on: ubuntu-latest
+    # needs 16 Gb RAM for phi4
+    runs-on: buildjet-4vcpu-ubuntu-2204
 #    services:
 #      ollama:
 #        image: ollama/ollama