Hi, I'm using LlamaIndex 0.10.18 with

ssti_uuuuu_its

Hi, I'm using LlamaIndex 0.10.18 with FastAPI. Every module that I use with use_async=True return errors. Am I doing anything wrong? Here is the example error when I use SubQuestionQueryEngine:

Plain Text

File "/opt/anaconda3/envs/talking-resume/lib/python3.11/site-packages/llama_index/core/async_utils.py", line 49, in run_async_tasks
    outputs: List[Any] = asyncio.run(_gather())
                         ^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/anaconda3/envs/talking-resume/lib/python3.11/asyncio/runners.py", line 186, in run
    raise RuntimeError(
RuntimeError: asyncio.run() cannot be called from a running event loop
/opt/anaconda3/envs/talking-resume/lib/python3.11/site-packages/uvicorn/protocols/http/httptools_impl.py:-1: RuntimeWarning: coroutine 'run_async_tasks.<locals>._gather' was never awaited
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
/opt/anaconda3/envs/talking-resume/lib/python3.11/site-packages/uvicorn/protocols/http/httptools_impl.py:-1: RuntimeWarning: coroutine 'SubQuestionQueryEngine._aquery_subq' was never awaited
RuntimeWarning: Enable tracemalloc to get the object allocation traceback

6 comments

LLogan M

You need to enable nested async (this happens when a sync method is running async tasks with asyncio.run()



```
import nest_asyncio
nest_asyncio.apply()
```

But in fast-api, that will error out unless you set the loop type to be asyncio

uvicorn.run(..., loop="asyncio")`

ssti_uuuuu_its

I tried but it didn't work.

ssti_uuuuu_its

Where to put that async io code? In main.py where the FastAPI starts or just include when running async task?

LLogan M

At the top of your main.py file

ssti_uuuuu_its

It works now. Thanks a lot!

ssti_uuuuu_its

Hi @Logan M I'm using SummaryIndex and VectorIndex for the same document and trying to delete the indices from /storage (docstore.json, index_store.json), like this:

Plain Text

index.delete_ref_doc(ref_doc_id=ref_doc_id, delete_from_docstore=True)

The vector indices in db is deleted but
1) The nodes in index_store.json index_store/data object is empty but the record of the indices still exists. Like this:

Plain Text

{
  "index_store/data": {
    "vector_store.5aca6f9d-5e21-40db-bc8d-c64d0cb7cb49": {
      "__type__": "vector_store",
      "__data__": "{\"index_id\": \"vector_store.5aca6f9d-5e21-40db-bc8d-c64d0cb7cb49\", \"summary\": null, \"nodes_dict\": {}, \"doc_id_dict\": {}, \"embeddings_dict\": {}}"
    },
  "list.5aca6f9d-5e21-40db-bc8d-c64d0cb7cb49": {
      "__type__": "list",
      "__data__": "{\"index_id\": \"list.5aca6f9d-5e21-40db-bc8d-c64d0cb7cb49\", \"summary\": null, \"nodes\": []}"
    },
}

2) The docstore/data is still there just the embedding is null now but docstore/metadata and docstore/ref_doc_info are also null.

How can I delete them both? Thank you in advance.

Add a reply

Find answers from the community

Hi, I'm using LlamaIndex 0.10.18 with