[Bug]: aembedding parameter "api_base" is not working when using model with "ollama"

[johnlxi]

What happened?

litellm and ollama in different docker container while running sample code below, param api_base is not working with error print "litellm.exceptions.APIConnectionError: Cannot connect to host localhost:11434 ssl:default [Connect call failed ('127.0.0.1', 11434)]"

import asyncio
#from litellm import embedding
from litellm import aembedding

async def test_gen_embedding():
    return await aembedding(
            model="ollama/nomic-embed-text:v1.5",
            input=["test litellm ollama nomic-embed-text v1.5"],
            api_base="http://ollama:11434",
            aembedding=True
            )

if __name__ == "__main__":
    #resp = embedding(
    #        model="ollama/nomic-embed-text:v1.5",
    #        input=["test litellm ollama nomic-embed-text v1.5"],
    #        api_base="http://ollama:11434",
    #        aembedding=True
    #        )

    resp = asyncio.run(test_gen_embedding())

    print(resp)
    print("-- done --")

seems ollama.ollama_aembeddings(...) is called without passing parameter "api_base" , check file litellm/main.py:2782

            if aembedding == True:
                response = ollama.ollama_aembeddings(
                    model=model,
                    prompt=ollama_input,
                    encoding=encoding,
                    logging_obj=logging,
                    optional_params=optional_params,
                    model_response=EmbeddingResponse(),
                )

Relevant log output

Traceback (most recent call last):
  File "/tmp/share/test-litellm-2/test_litellm_embed.py", line 24, in <module>
    resp = asyncio.run(test_gen_embedding())
  File "/usr/lib/python3.11/asyncio/runners.py", line 190, in run
    return runner.run(main)
  File "/usr/lib/python3.11/asyncio/runners.py", line 118, in run
    return self._loop.run_until_complete(task)
  File "/usr/lib/python3.11/asyncio/base_events.py", line 653, in run_until_complete
    return future.result()
  File "/tmp/share/test-litellm-2/test_litellm_embed.py", line 9, in test_gen_embedding
    return await aembedding(
  File "/tmp/share/test-litellm-2/lib/python3.11/site-packages/litellm/utils.py", line 3238, in wrapper_async
    raise e
  File "/tmp/share/test-litellm-2/lib/python3.11/site-packages/litellm/utils.py", line 3069, in wrapper_async
    result = await original_function(*args, **kwargs)
  File "/tmp/share/test-litellm-2/lib/python3.11/site-packages/litellm/main.py", line 2430, in aembedding
    raise exception_type(
  File "/tmp/share/test-litellm-2/lib/python3.11/site-packages/litellm/utils.py", line 8265, in exception_type
    raise e
  File "/tmp/share/test-litellm-2/lib/python3.11/site-packages/litellm/utils.py", line 8240, in exception_type
    raise APIConnectionError(
litellm.exceptions.APIConnectionError: Cannot connect to host localhost:11434 ssl:default [Connect call failed ('127.0.0.1', 11434)]

Twitter / LinkedIn details

No response

[onukura]

This must be fixed now after #2675.

[edwinjosegeorge]

@krrishdholakia Seems like this issue is already fixed and ready to be closed