Sth wrong with using Ollama +qdrant：Vector dimension error: expected dim: 1536, got 768

[AI-Beans]

🐛 Describe the bug

I use ollama embedingmodel and chatmodel，get right response 。 But response form Qdrant：Vector dimension error: expected dim: 1536, got 768 where can I config the para ？

[teatimekon]

Me,too，after fixing bugs in mem0/llms/ollama.py and correctly make m = Memory.from_config(config)，i meet this bug while using result = m.add()

[teatimekon]

🐛 Describe the bug

I use ollama embedingmodel and chatmodel，get right response 。 But response form Qdrant：Vector dimension error: expected dim: 1536, got 768 where can I config the para ？

just restart the qdrant Docker can fix this bug ,see #712

[AI-Beans]

Rebooted, but still having the same issue！

[teatimekon]

Rebooted, but still having the same issue！

try edit mem0/embeddings/ollama.py in line 9

    def __init__(self, model="nomic-embed-text"):
        self.model = model
        self._ensure_model_exists()
        self.dims = 768  #this is mine
        ```

[AI-Beans]

config = { "vector_store": { "provider": "qdrant", "config": { "host": "localhost", "port": 6333,
} }, have you changed this config？

[teatimekon]

nope

[AI-Beans]

Is there anything else that needs to be set up?

[teatimekon]

Is there anything else that needs to be set up?

I modified a very small portion of the source code based on the error message, but the fact is that mem0 currently does not support local ollama operation

[AI-Beans]

Is there anything else that needs to be set up?

I modified a very small portion of the source code based on the error message, but the fact is that mem0 currently does not support local ollama operation

OK Then I won't take the time to make it happen. I used openai first, and the effect was not bad！

Could you offer you email or wechat account, We can discuss it together

[yslion]

just create a new 'collection_name' ， by default, the 'collection_name' is 'mem0' (see : class MemoryItem ) , dims is 1536 . different config ,different collection_name :)

[johnwlockwood]

The docs for the embedding model nomic-embed-text:latest shows it's dims are 768. So config the vector_store's embedding_model_dims to 768.

"vector_store": {
        "provider": "qdrant",
        "config": {
            "collection_name": "test",
            "host": "localhost",
            "port": 6333,
            "embedding_model_dims": 768,  # Change this according to your local model's dimensions
        },
    },

If you have already created the collection, which you have by the time you got this error, then either delete the collection in the Qdrant UI or change the collection_name so it will make a new one during Memory.add

[parshvadaftari]

Closing as fixed

[nerdyabhi]

getting same issue with typescript sdk :

const config = {
  llm: {
    provider: "groq",
    config: {
      model: "llama-3.1-8b-instant",
      temperature: 0.1,
      max_tokens: 1000,
    },
  },
  embedder: {
    provider: "google",
    config: {
      apiKey: process.env.GEMINI_API_KEY!,
      model: "gemini-embedding-001",
      embeddingDims: 768,
      embedding_dims: 768,
      embedding_model_dims: 768,
    },
  },
  vectorStore: {
    provider: "qdrant",
    config: {
      collectionName: "abhi-new-test",
      embeddingModelDims: 768,
      host: "localhost",
      port: 6333,
    },
  },
  version: "v1.1",
};

tried whole day figuring this out !!

             throw new fun.Error(err);
                      ^
ApiError: Bad Request
    at Object.fun [as searchPoints] (file:///home/abhi/abhi-projects/genAI/node_modules/.pnpm/@[email protected]/node_modules/@qdrant/openapi-typescript-fetch/dist/esm/fetcher.js:169:23)
    at processTicksAndRejections (node:internal/process/task_queues:105:5)
    at async QdrantClient.search (file:///home/abhi/abhi-projects/genAI/node_modules/.pnpm/@[email protected][email protected]/node_modules/@qdrant/js-client-rest/dist/esm/qdrant-client.js:167:26)
    at async Qdrant.search (file:///home/abhi/abhi-projects/genAI/node_modules/.pnpm/[email protected]_@[email protected][email protected]__@[email protected]_73eb170a988c6a467084e35c8e96c8d9/node_modules/mem0ai/src/oss/src/vector_stores/qdrant.ts:124:21)
    at async _Memory.addToVectorStore (file:///home/abhi/abhi-projects/genAI/node_modules/.pnpm/[email protected]_@[email protected][email protected]__@[email protected]_73eb170a988c6a467084e35c8e96c8d9/node_modules/mem0ai/src/oss/src/memory/index.ts:278:32)
    at async _Memory.add (file:///home/abhi/abhi-projects/genAI/node_modules/.pnpm/[email protected]_@[email protected][email protected]__@[email protected]_73eb170a988c6a467084e35c8e96c8d9/node_modules/mem0ai/src/oss/src/memory/index.ts:191:31)
    at async file:///home/abhi/abhi-projects/genAI/src/concepts/langraph/mem0.ts:59:1 {
  headers: Headers {},
  url: 'http://localhost:6333/collections/abhi-new-test/points/search',
  status: 400,
  statusText: 'Bad Request',
  data: {
    status: {
      error: 'Wrong input: Vector dimension error: expected dim: 1536, got 768'
    },
    time: 0.000357633
  }
}