When I try to switch ollama_embedder, I get an error in the system #853

Describe the bug
When I try to switch ollama_embedder, I get an error in the system, and finally i set EMBEDDING_MODEL_DIMENSION=3072 still error

2024-10-31 20:43:01 app.state.service_container = create_service_container(
2024-10-31 20:43:01 ^^^^^^^^^^^^^^^^^^^^^^^^^
2024-10-31 20:43:01 File "/src/globals.py", line 74, in create_service_container
2024-10-31 20:43:01 "indexing": indexing.Indexing(
2024-10-31 20:43:01 ^^^^^^^^^^^^^^^^^^
2024-10-31 20:43:01 File "/src/pipelines/indexing/indexing.py", line 630, in init
2024-10-31 20:43:01 dbschema_store = document_store_provider.get_store()
2024-10-31 20:43:01 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2024-10-31 20:43:01 File "/src/providers/document_store/qdrant.py", line 364, in get_store
2024-10-31 20:43:01 return AsyncQdrantDocumentStore(
2024-10-31 20:43:01 ^^^^^^^^^^^^^^^^^^^^^^^^^
2024-10-31 20:43:01 File "/src/providers/document_store/qdrant.py", line 161, in init
2024-10-31 20:43:01 self.client.create_payload_index(
2024-10-31 20:43:01 ^^^^^^^^^^^
2024-10-31 20:43:01 File "/app/.venv/lib/python3.12/site-packages/haystack_integrations/document_stores/qdrant/document_store.py", line 282, in client
2024-10-31 20:43:01 self._set_up_collection(
2024-10-31 20:43:01 File "/app/.venv/lib/python3.12/site-packages/haystack_integrations/document_stores/qdrant/document_store.py", line 833, in _set_up_collection
2024-10-31 20:43:01 raise ValueError(msg)
2024-10-31 20:43:01 ValueError: Collection 'Document' already exists in Qdrant, but it is configured with a vector size '3072'. If you want to use that collection, but with a different vector size, please set recreate_collection=True argument.
2024-10-31 20:43:01
2024-10-31 20:43:01 ERROR: Application startup failed. Exiting.
2024-10-31 20:43:31 This module is deprecated and will be removed in Hamilton 2.0 Please use hamilton.async_driver instead.
2024-10-31 20:43:32 INFO: Started server process [7]
2024-10-31 20:43:32 INFO: Waiting for application startup.

To Reproduce
Steps to reproduce the behavior:

1. Go to '.env.ai' file. and set EMBEDDER_PROVIDER=ollama_embedder EMBEDDING_MODEL=nomic-embed-text
2. ./wren-launcher-darwin

Expected behavior
wrenai-wren-ai-service-1 should start normally.

Screenshots

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]

Wren AI Information

• Version: 0.9.1
• LLM_PROVIDER=#ollama_llm
• GENERATION_MODEL= # qwen2:1.5b, llama3:70b, etc.

Additional context
Add any other context about the problem here.