[Question]: Parse doc failed

[mobguang]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (Language Policy).
• Non-english title submitions will be closed directly ( 非英文标题的提交将会被直接关闭 ) (Language Policy).
• Please do not modify this template :) and fill in all the required fields.

Describe your problem

Hi @dosubot,

I deployed ragflow 0.21.1 on mac os by source code.

I encountered two issues:

1. I tried to parse .doc file by BAAI/bge-large-zh-v1.5 model and it listed in llm_factories.json already, but the log still displayed error as following:

‘’‘
2025-10-28 18:01:08,333 ERROR 44839 Fail to bind embedding model: Model(BAAI/bge-large-zh-v1.5@BAAI) not authorized
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 799, in do_handle_task
embedding_model = LLMBundle(task_tenant_id, LLMType.EMBEDDING, llm_name=task_embedding_id, lang=task_language)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 72, in init
super().init(tenant_id, llm_type, llm_name, lang, **kwargs)
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/tenant_llm_service.py", line 250, in init
self.mdl = TenantLLMService.model_instance(tenant_id, llm_type, llm_name, lang=lang, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/peewee.py", line 3128, in inner
return fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/tenant_llm_service.py", line 132, in model_instance
model_config = TenantLLMService.get_model_config(tenant_id, llm_type, llm_name)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/peewee.py", line 3128, in inner
return fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/tenant_llm_service.py", line 120, in get_model_config
raise LookupError(f"Model({mdlnm}@{fid}) not authorized")
LookupError: Model(BAAI/bge-large-zh-v1.5@BAAI) not authorized
2025-10-28 18:01:08,343 INFO 44839 set_progress(05f20e5ab3e511f0b9f6ea8e869d4ad9), progress: -1, progress_msg: 18:01:08 [ERROR][Exception]: Model(BAAI/bge-large-zh-v1.5@BAAI) not authorized
2025-10-28 18:01:08,344 ERROR 44839 handle_task got exception for task {"id": "05f20e5ab3e511f0b9f6ea8e869d4ad9", "doc_id": "232f8e5eb3e311f0a59aea8e869d4ad9", "from_page": 0, "to_page": 100000000, "retry_count": 0, "kb_id": "6f76425ab3e111f0a59aea8e869d4ad9", "parser_id": "naive", "parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "raptor": {"use_raptor": true, "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.", "max_token": 256, "threshold": 0.1, "max_cluster": 64, "random_seed": 0}, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}}, "name": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "type": "doc", "location": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "size": 2560931, "tenant_id": "74f31f7eb3cc11f093cd5e95a36a973e", "language": "English", "embd_id": "BAAI/bge-large-zh-v1.5@BAAI", "pagerank": 0, "kb_parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "raptor": {"use_raptor": true, "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.", "max_token": 256, "threshold": 0.1, "max_cluster": 64, "random_seed": 0}, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}}, "img2txt_id": "", "asr_id": "", "llm_id": "gpt-oss:120b@Ollama", "update_time": 1761645667725, "task_type": ""}
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 972, in handle_task
await do_handle_task(task)
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 775, in async_wrapper
return await func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 799, in do_handle_task
embedding_model = LLMBundle(task_tenant_id, LLMType.EMBEDDING, llm_name=task_embedding_id, lang=task_language)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 72, in init
super().init(tenant_id, llm_type, llm_name, lang, **kwargs)
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/tenant_llm_service.py", line 250, in init
self.mdl = TenantLLMService.model_instance(tenant_id, llm_type, llm_name, lang=lang, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/peewee.py", line 3128, in inner
return fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/tenant_llm_service.py", line 132, in model_instance
model_config = TenantLLMService.get_model_config(tenant_id, llm_type, llm_name)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/peewee.py", line 3128, in inner
return fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/tenant_llm_service.py", line 120, in get_model_config
raise LookupError(f"Model({mdlnm}@{fid}) not authorized")
LookupError: Model(BAAI/bge-large-zh-v1.5@BAAI) not authorized
’‘’

2. I tried to parse .doc file by bge-m3:latest model that launched by ollama locally, but another error displayed in log as following:
‘’‘
Please check that Ollama is downloaded, running and accessible. https://ollama.com/download
2025-10-28 18:28:25,932 ERROR 44839 Fail to bind embedding model: Failed to connect to Ollama. Please check that Ollama is downloaded, running and accessible. https://ollama.com/download
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 800, in do_handle_task
vts, _ = embedding_model.encode(["ok"])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(api.db.services.llm_service.LLMBundle.encode) at 0x3a202e700>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 84, in encode
embeddings, used_tokens = self.mdl.encode(texts)
^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(rag.llm.embedding_model.OllamaEmbed.encode) at 0x3a15fda80>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/llm/embedding_model.py", line 278, in encode
res = self.client.embeddings(prompt=txt, model=self.model_name, options={"use_mmap": True}, keep_alive=self.keep_alive)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 401, in embeddings
return self._request(
^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 189, in _request
return cls(**self._request_raw(*args, **kwargs).json())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 135, in _request_raw
raise ConnectionError(CONNECTION_ERROR_MESSAGE) from None
ConnectionError: Failed to connect to Ollama. Please check that Ollama is downloaded, running and accessible. https://ollama.com/download
2025-10-28 18:28:25,939 INFO 44839 set_progress(d65d4278b3e811f0b9f6ea8e869d4ad9), progress: -1, progress_msg: 18:28:25 [ERROR][Exception]: Failed to connect to Ollama. Please check that Ollama is downloaded, running and accessible. https://ollama.com/download
2025-10-28 18:28:25,939 ERROR 44839 handle_task got exception for task {"id": "d65d4278b3e811f0b9f6ea8e869d4ad9", "doc_id": "232f8e5eb3e311f0a59aea8e869d4ad9", "from_page": 0, "to_page": 100000000, "retry_count": 0, "kb_id": "6f76425ab3e111f0a59aea8e869d4ad9", "parser_id": "naive", "parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "raptor": {"use_raptor": true, "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.", "max_token": 256, "threshold": 0.1, "max_cluster": 64, "random_seed": 0}, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}}, "name": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "type": "doc", "location": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "size": 2560931, "tenant_id": "74f31f7eb3cc11f093cd5e95a36a973e", "language": "English", "embd_id": "bge-m3:latest@Ollama", "pagerank": 0, "kb_parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "raptor": {"use_raptor": true, "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.", "max_token": 256, "threshold": 0.1, "max_cluster": 64, "random_seed": 0}, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}}, "img2txt_id": "", "asr_id": "", "llm_id": "gpt-oss:120b@Ollama", "update_time": 1761647305884, "task_type": ""}
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 972, in handle_task
await do_handle_task(task)
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 775, in async_wrapper
return await func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 800, in do_handle_task
vts, _ = embedding_model.encode(["ok"])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(api.db.services.llm_service.LLMBundle.encode) at 0x3a202e700>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 84, in encode
embeddings, used_tokens = self.mdl.encode(texts)
^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(rag.llm.embedding_model.OllamaEmbed.encode) at 0x3a15fda80>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/llm/embedding_model.py", line 278, in encode
res = self.client.embeddings(prompt=txt, model=self.model_name, options={"use_mmap": True}, keep_alive=self.keep_alive)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 401, in embeddings
return self._request(
^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 189, in _request
return cls(**self._request_raw(*args, **kwargs).json())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 135, in _request_raw
raise ConnectionError(CONNECTION_ERROR_MESSAGE) from None
ConnectionError: Failed to connect to Ollama. Please check that Ollama is downloaded, running and accessible. https://ollama.com/download
’‘’

**I found other similar issues, you suggested that we must register model firstly, but what "register model" mean? There is no any doc explain it.

Please kindly provide the detail solution. Thanks.**

[dosubot]

In RAGFlow, "register model" means explicitly associating a model with a tenant in the backend configuration/database—not just listing it in llm_factories.json. Each tenant must have the model authorized with the correct factory, API key, and base URL. If you see Model(...@...) not authorized, it means this association is missing for your tenant, even if the model is globally available. You need to add an entry for your tenant and model in the TenantLLM table (or equivalent config), specifying the model name, factory (like "BAAI" or "Ollama"), API key, and base URL. This is a required step for every tenant-model pair you want to use, and is handled outside of the embedding model code itself (source).

For Ollama models, the error Failed to connect to Ollama means the Ollama service must be running and accessible at the configured base_url. Make sure Ollama is installed, started, and reachable from your RAGFlow environment. This is separate from model registration—both are required for successful use (source).

In summary:

1. Register/authorize the model for your tenant in the backend config/database (not just llm_factories.json).
2. For Ollama, ensure the service is running and accessible.

If you need step-by-step instructions for updating the backend config/database, let me know your setup (e.g., are you using SQLite, Postgres, or a config file), and I can help you with the exact commands or queries.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}    

[Magicbook1108]

For your second question, you should make sure your Ollama running in the backend. Follow the tutorial.

[mobguang]

@dosu, @Magicbook1108 ,

Thanks for your reply.

For the first issue, please provide the step-by-step tutorial for how to register model.

For the second issue, I executed "curl http://localhost:11434/" and the response in console was "Ollama is running". But there is no change for the parsing.

[dosubot]

Here’s how to register a model for a tenant in RAGFlow:

1. You need to create an entry in the TenantLLM table (the backend database table for tenant-model associations). This entry must include: tenant_id, llm_name (e.g., BAAI/bge-large-zh-v1.5), llm_factory (e.g., BAAI or Ollama), api_key (if needed), and api_base (the model’s base URL) [source].

2. This registration is not handled by editing llm_factories.json alone. You must explicitly associate the model with your tenant in the backend config/database. If you’re using SQLite or Postgres, you can insert a row directly into the TenantLLM table. For example (SQL syntax, adjust for your DB):

INSERT INTO TenantLLM (tenant_id, llm_name, llm_factory, api_key, api_base)
VALUES ('your-tenant-id', 'BAAI/bge-large-zh-v1.5', 'BAAI', 'your-api-key', 'your-base-url');

For Ollama, the process is the same: register the model (e.g., bge-m3:latest) for your tenant with llm_factory set to 'Ollama', and api_base set to your Ollama URL (e.g., http://localhost:11434).

If you’re not sure how to access your database or which DB you’re using, let me know and I can help with the exact commands.

For the Ollama issue: If curl http://localhost:11434/ returns “Ollama is running” but parsing still fails, double-check that the base_url in your TenantLLM entry matches the URL you’re using for curl, and that the model name in RAGFlow matches the model you’ve pulled in Ollama. Also, make sure the model is actually loaded in Ollama and available for embedding requests [source].

If you need help with the exact SQL or Python code for your environment, let me know your backend DB type.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}    

[mobguang]

@dosu,

The register operation can only be implemented by insert DB table manually? No GUI function entrance?

[dosubot]

You don't have to register models only by manually inserting into the database. RAGFlow provides a Web UI ("Model Providers" page) where you can add/register models for your user account—just select the factory, enter the model name, base URL, and API key, then save. However, this UI registration is user-specific and does not apply globally to all tenants. For global or tenant-wide registration, you still need to use backend methods like direct DB insertion or backend API calls. There is also a backend API endpoint (/add_llm) for programmatic registration if you prefer not to edit the DB directly. See more details and step-by-step examples in this discussion: source.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}    

[mobguang]

@Magicbook1108, @dosu,

I pulled qllama/bge-m3:latest model by ollama and parsed the.docx file by it (this file contains picture and text) .

The configuration as following screenshot.

There are some errors for the parsing progress as following:

'''
2025-10-28 19:52:20,587 INFO 16279 set_progress(8e32656cb3f411f088edea8e869d4ad9), progress: None, progress_msg: 19:52:20 Page(1100000001): Generate 6 chunks
2025-10-28 19:52:21,113 INFO 16279 set_progress(8e32656cb3f411f088edea8e869d4ad9), progress: -1, progress_msg: 19:52:21 Page(1100000001): [ERROR]Generate embedding error:do embedding request: Post "http://127.0.0.1:52804/embedding": EOF (status code: 500)
2025-10-28 19:52:21,114 ERROR 16279 Generate embedding error:do embedding request: Post "http://127.0.0.1:52804/embedding": EOF (status code: 500)
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 916, in do_handle_task
token_count, vector_size = await embedding(chunks, embedding_model, task_parser_config, progress_callback)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in embedding
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 449, in to_thread_run_sync
return msg_from_thread.unwrap()
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 373, in do_release_then_return_result
return result.unwrap()
^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 392, in worker_fn
ret = context.run(sync_fn, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 758, in wrapper
raise result
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 742, in target
result = func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 485, in batch_encode
return mdl.encode([truncate(c, mdl.max_length-10) for c in txts])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(api.db.services.llm_service.LLMBundle.encode) at 0x3a082e7a0>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 84, in encode
embeddings, used_tokens = self.mdl.encode(texts)
^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(rag.llm.embedding_model.OllamaEmbed.encode) at 0x387dfdb20>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/llm/embedding_model.py", line 278, in encode
res = self.client.embeddings(prompt=txt, model=self.model_name, options={"use_mmap": True}, keep_alive=self.keep_alive)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 401, in embeddings
return self._request(
^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 189, in _request
return cls(**self._request_raw(*args, **kwargs).json())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 133, in _request_raw
raise ResponseError(e.response.text, e.response.status_code) from None
ollama._types.ResponseError: do embedding request: Post "http://127.0.0.1:52804/embedding": EOF (status code: 500)
2025-10-28 19:52:21,126 INFO 16279 set_progress(8e32656cb3f411f088edea8e869d4ad9), progress: -1, progress_msg: 19:52:21 [ERROR][Exception]: do embedding request: Post "http://127.0.0.1:52804/embedding": EOF (status code: 500)
2025-10-28 19:52:21,126 ERROR 16279 handle_task got exception for task {"id": "8e32656cb3f411f088edea8e869d4ad9", "doc_id": "831d268ab3f411f088edea8e869d4ad9", "from_page": 0, "to_page": 100000000, "retry_count": 0, "kb_id": "6f76425ab3e111f0a59aea8e869d4ad9", "parser_id": "naive", "parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "raptor": {"use_raptor": true, "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.", "max_token": 256, "threshold": 0.1, "max_cluster": 64, "random_seed": 0}, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}}, "name": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "type": "doc", "location": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "size": 2560931, "tenant_id": "74f31f7eb3cc11f093cd5e95a36a973e", "language": "English", "embd_id": "qllama/bge-m3:latest@Ollama", "pagerank": 0, "kb_parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "raptor": {"use_raptor": true, "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.", "max_token": 256, "threshold": 0.1, "max_cluster": 64, "random_seed": 0}, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}}, "img2txt_id": "", "asr_id": "", "llm_id": "gpt-oss:120b@Ollama", "update_time": 1761652338770, "task_type": ""}
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 972, in handle_task
await do_handle_task(task)
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 775, in async_wrapper
return await func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 916, in do_handle_task
token_count, vector_size = await embedding(chunks, embedding_model, task_parser_config, progress_callback)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in embedding
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 449, in to_thread_run_sync
return msg_from_thread.unwrap()
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 373, in do_release_then_return_result
return result.unwrap()
^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 392, in worker_fn
ret = context.run(sync_fn, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 758, in wrapper
raise result
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 742, in target
result = func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 485, in batch_encode
return mdl.encode([truncate(c, mdl.max_length-10) for c in txts])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(api.db.services.llm_service.LLMBundle.encode) at 0x3a082e7a0>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 84, in encode
embeddings, used_tokens = self.mdl.encode(texts)
^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(rag.llm.embedding_model.OllamaEmbed.encode) at 0x387dfdb20>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/llm/embedding_model.py", line 278, in encode
res = self.client.embeddings(prompt=txt, model=self.model_name, options={"use_mmap": True}, keep_alive=self.keep_alive)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 401, in embeddings
return self._request(
^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 189, in _request
return cls(**self._request_raw(*args, **kwargs).json())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 133, in _request_raw
raise ResponseError(e.response.text, e.response.status_code) from None
ollama._types.ResponseError: do embedding request: Post "http://127.0.0.1:52804/embedding": EOF (status code: 500)
2025-10-28 19:52:23,832 INFO 16280 127.0.0.1 - - [28/Oct/2025 19:52:23] "POST /v1/document/list?kb_id=6f76425ab3e111f0a59aea8e869d4ad9&keywords=&page_size=50&page=1 HTTP/1.1" 200 -
2025-10-28 19:52:23,863 INFO 16280 127.0.0.1 - - [28/Oct/2025 19:52:23] "GET /v1/kb/detail?kb_id=6f76425ab3e111f0a59aea8e869d4ad9 HTTP/1.1" 200 -
'''

[dosubot]

The error EOF (status code: 500) when posting to http://127.0.0.1:52804/embedding usually means RAGFlow is trying to reach Ollama on the wrong port or address. By default, Ollama listens on port 11434. If you’re running RAGFlow in Docker or a similar environment, 127.0.0.1 inside the container points to the container itself, not your host machine where Ollama is running. This causes connection failures and server-side errors like the one you’re seeing.

To fix this, update your embedding model configuration in RAGFlow so the base_url for Ollama is set to http://host.docker.internal:11434 (on macOS/Windows Docker) or http://localhost:11434 if running everything natively. Make sure the port matches the one Ollama is actually using. Also, double-check that the model name in your config matches exactly what you see in ollama list (for example, bge-m3:latest).

If you still see errors, check the Ollama server logs for crashes, and try lowering EMBEDDING_BATCH_SIZE and setting MAX_CONCURRENT_TASKS=1 in your .env to reduce load. Some models may not support embedding or may have compatibility issues—if the problem persists, try switching to a different embedding model that is known to work with Ollama and RAGFlow (like nomic-embed-text:latest) reference.

Let me know if you need help updating your config or want to check model compatibility!

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}    

[mobguang]

@dosu,

Actually, the base_url is http://localhost:11434 already in configuration.

And I tried it again, but the same error exists. But the base_url port change to "55311" instead of "52804" and it is not 11434. Is it random number for every launch?

'''
2025-10-28 20:10:26,340 INFO 53329 set_progress(15f52bb8b3f711f0b5e6ea8e869d4ad9), progress: None, progress_msg: 20:10:26 Page(1100000001): Generate 6 chunks
2025-10-28 20:10:26,579 INFO 53329 set_progress(15f52bb8b3f711f0b5e6ea8e869d4ad9), progress: -1, progress_msg: 20:10:26 Page(1100000001): [ERROR]Generate embedding error:do embedding request: Post "http://127.0.0.1:55311/embedding": EOF (status code: 500)
2025-10-28 20:10:26,579 ERROR 53329 Generate embedding error:do embedding request: Post "http://127.0.0.1:55311/embedding": EOF (status code: 500)
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 916, in do_handle_task
token_count, vector_size = await embedding(chunks, embedding_model, task_parser_config, progress_callback)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in embedding
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 449, in to_thread_run_sync
return msg_from_thread.unwrap()
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 373, in do_release_then_return_result
return result.unwrap()
^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 392, in worker_fn
ret = context.run(sync_fn, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 758, in wrapper
raise result
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 742, in target
result = func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 485, in batch_encode
return mdl.encode([truncate(c, mdl.max_length-10) for c in txts])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(api.db.services.llm_service.LLMBundle.encode) at 0x3b1f2e7a0>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 84, in encode
embeddings, used_tokens = self.mdl.encode(texts)
^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(rag.llm.embedding_model.OllamaEmbed.encode) at 0x3b14fdb20>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/llm/embedding_model.py", line 278, in encode
res = self.client.embeddings(prompt=txt, model=self.model_name, options={"use_mmap": True}, keep_alive=self.keep_alive)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 401, in embeddings
return self._request(
^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 189, in _request
return cls(**self._request_raw(*args, **kwargs).json())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 133, in _request_raw
raise ResponseError(e.response.text, e.response.status_code) from None
ollama._types.ResponseError: do embedding request: Post "http://127.0.0.1:55311/embedding": EOF (status code: 500)
2025-10-28 20:10:26,595 INFO 53329 set_progress(15f52bb8b3f711f0b5e6ea8e869d4ad9), progress: -1, progress_msg: 20:10:26 [ERROR][Exception]: do embedding request: Post "http://127.0.0.1:55311/embedding": EOF (status code: 500)
2025-10-28 20:10:26,595 ERROR 53329 handle_task got exception for task {"id": "15f52bb8b3f711f0b5e6ea8e869d4ad9", "doc_id": "f7cf4470b3f611f0b5e6ea8e869d4ad9", "from_page": 0, "to_page": 100000000, "retry_count": 0, "kb_id": "6f76425ab3e111f0a59aea8e869d4ad9", "parser_id": "naive", "parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}, "task_page_size": 12, "entity_types": [], "pages": [[1, 1024]]}, "name": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "type": "doc", "location": "\u8f66\u8f86\u64cd\u4f5c\u89c4\u8303\u53ca\u5e38\u89c1\u6545\u969c\u6392\u67e5\u6307\u535720220208.docx", "size": 2560931, "tenant_id": "74f31f7eb3cc11f093cd5e95a36a973e", "language": "English", "embd_id": "qllama/bge-m3:latest@Ollama", "pagerank": 0, "kb_parser_config": {"layout_recognize": "DeepDOC", "chunk_token_num": 512, "delimiter": "\n", "auto_keywords": 0, "auto_questions": 0, "html4excel": false, "topn_tags": 3, "toc_extraction": false, "raptor": {"use_raptor": true, "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n {cluster_content}\nThe above is the content you need to summarize.", "max_token": 256, "threshold": 0.1, "max_cluster": 64, "random_seed": 0}, "graphrag": {"use_graphrag": true, "entity_types": ["organization", "person", "geo", "event", "category"], "method": "light"}}, "img2txt_id": "", "asr_id": "", "llm_id": "gpt-oss:120b@Ollama", "update_time": 1761653425542, "task_type": ""}
Traceback (most recent call last):
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 972, in handle_task
await do_handle_task(task)
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 775, in async_wrapper
return await func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 916, in do_handle_task
token_count, vector_size = await embedding(chunks, embedding_model, task_parser_config, progress_callback)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in embedding
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 449, in to_thread_run_sync
return msg_from_thread.unwrap()
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 373, in do_release_then_return_result
return result.unwrap()
^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/outcome/_impl.py", line 213, in unwrap
raise captured_error
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/trio/_threads.py", line 392, in worker_fn
ret = context.run(sync_fn, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 490, in
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(cnts[i: i + EMBEDDING_BATCH_SIZE]))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 758, in wrapper
raise result
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/utils/api_utils.py", line 742, in target
result = func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/svr/task_executor.py", line 485, in batch_encode
return mdl.encode([truncate(c, mdl.max_length-10) for c in txts])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(api.db.services.llm_service.LLMBundle.encode) at 0x3b1f2e7a0>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/api/db/services/llm_service.py", line 84, in encode
embeddings, used_tokens = self.mdl.encode(texts)
^^^^^^^^^^^^^^^^^^^^^^
File "<@beartype(rag.llm.embedding_model.OllamaEmbed.encode) at 0x3b14fdb20>", line 31, in encode
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/rag/llm/embedding_model.py", line 278, in encode
res = self.client.embeddings(prompt=txt, model=self.model_name, options={"use_mmap": True}, keep_alive=self.keep_alive)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 401, in embeddings
return self._request(
^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 189, in _request
return cls(**self._request_raw(*args, **kwargs).json())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Volumes/M2SSD/Space/Workspace/llm/ragflow_0211/.venv/lib/python3.11/site-packages/ollama/_client.py", line 133, in _request_raw
raise ResponseError(e.response.text, e.response.status_code) from None
ollama._types.ResponseError: do embedding request: Post "http://127.0.0.1:55311/embedding": EOF (status code: 500)
2025-10-28 20:10:30,604 INFO 53330 127.0.0.1 - - [28/Oct/2025 20:10:30] "POST /v1/document/list?kb_id=6f76425ab3e111f0a59aea8e869d4ad9&keywords=&page_size=50&page=1 HTTP/1.1" 200 -
2025-10-28 20:10:30,634 INFO 53330 127.0.0.1 - - [28/Oct/2025 20:10:30] "GET /v1/kb/detail?kb_id=6f76425ab3e111f0a59aea8e869d4ad9 HTTP/1.1" 200 -
2025-10-28 20:10:44,862 INFO 53329 task_executor_0 reported heartbeat: {"name": "task_executor_0", "now": "2025-10-28T20:10:44.859+08:00", "boot_at": "2025-10-28T20:08:44.737+08:00", "pending": 0, "lag": 0, "done": 1, "failed": 1, "current": {}}
'''

[dosubot]

The Ollama embedding client in RAGFlow uses the base_url you configure—such as http://localhost:11434—directly for all embedding requests. There is no logic in RAGFlow that randomly assigns ports or changes the endpoint for embedding requests; the port should always match what you set in your model config or database entry reference.

If you see requests going to random ports like 55311 or 52804, this points to a misconfiguration or an environment issue outside of RAGFlow itself. Common causes include: a proxy or local service running on those ports, a mismatch between the UI config and the backend database, or an environment variable or config file overriding the base_url. Double-check that your model registration (in the database or via the UI/API) matches the actual Ollama endpoint, and that no other service is intercepting or redirecting requests reference.

Make sure only one embedding backend is running, and verify your TenantLLM table or config for the correct base_url. If you’re running in Docker, confirm that the network settings allow access to the host’s Ollama port (11434), and that you’re not accidentally connecting to a different service.

If you need help checking your database or config files for mismatches, let me know your setup details.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}    

[mobguang]

@dosu,

I lanuched ragflow by source code instead of docker. And the corresponding value of tenant_llm table is correct too. Please refer to the screenshot.

What else configuration files I need check?

[dosubot]

Given that your tenant_llm table is correct and you’re running RAGFlow from source, here’s what else you should check:

1. Look for a .env file or any environment variables (like OLLAMA_BASE_URL, EMBEDDING_BASE_URL, or similar) in your project directory or shell session—these can override the database config and cause unexpected endpoint/port changes.
2. Check for config files such as service_conf.yaml, config.yaml, or any custom YAML/JSON in your conf/ directory. Sometimes, these files set default embedding endpoints or override database values on startup.
3. Make sure there’s no proxy, local service, or background process on your machine that could be intercepting or redirecting HTTP requests to random ports.
4. If you have multiple RAGFlow processes or background workers running, ensure only one instance is active to avoid port conflicts or race conditions.
5. Restart your backend after any config changes to ensure all settings are reloaded.

There’s no logic in RAGFlow that randomly assigns embedding ports—the port should always match your config/database entry. If you’re still seeing requests to random ports, it’s almost always due to an override from a .env file, environment variable, or an external proxy/service. Double-check these sources and try running with a clean environment (no extra env vars set) to isolate the issue.
Reference

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}