[Bug]: can not work with llama.cpp

[calvin2021y]

Prerequisites

• I have searched existing issues and discussions to avoid duplicates
• I am using the latest version (or have tested against main/nightly)

Description

{
  "providers": [
    {
      "name": "a",
      "keys": [
        
      ],
      "network_config": {
        "base_url": "http://127.0.0.1:8081/v1",
        "default_request_timeout_in_seconds": 300,
        "max_retries": 0,
        "retry_backoff_initial": 500,
        "retry_backoff_max": 5000
      },
      "concurrency_and_buffer_size": {
        "concurrency": 1000,
        "buffer_size": 5000
      },
      "proxy_config": null,
      "send_back_raw_response": false,
      "custom_provider_config": {
        "is_key_less": true,
        "base_provider_type": "openai",
        "allowed_requests": {
          "list_models": true,
          "text_completion": true,
          "text_completion_stream": true,
          "chat_completion": true,
          "chat_completion_stream": true,
          "responses": true,
          "responses_stream": true,
          "embedding": true,
          "speech": true,
          "speech_stream": true,
          "transcription": true,
          "transcription_stream": true
        }
      },
      "status": "active"
    }
  ],
  "total": 1
}

Steps to reproduce

curl -X POST http://localhost:8080/v1/chat/completions -v
-H "Content-Type: application/json"
-d '{
"model": "a/a",
"messages": [{"role": "user", "content": "Hello, Bifrost!"}]
}'

Expected behavior

??

Actual behavior

??

Affected area(s)

Core (Go)

Version

the new onw

Environment

macos

try with  "base_url": "http://127.0.0.1:8081/v1", and  "base_url": "http://127.0.0.1:8081", none work.

both result int request: POST /v1/v1/chat/completions 127.0.0.1 404


## The bug reports template this project is very very very bad. ##

Relevant logs/output

{"is_bifrost_error":false,"status_code":404,"error":{"type":"not_found_error","code":"","message":"File Not Found"},"extra_fields":{"provider":"a","model_requested":"a","request_type":"chat_completion"}}
* Connection #0 to host localhost left intact

Regression?

No response

Severity

Low (minor issue or cosmetic)

[calvin2021y]

"base_url": "http://127.0.0.1:8081/" work with llama.cpp.

http://127.0.0.1:8081 and http://127.0.0.1:8081/v1 will not work.

[calvin2021y]

curl work for me:

curl --location 'http://localhost:8080/v1/completions' \
  --header 'Content-Type: application/json' \
  --data '{
      "model": "a/a",
      "prompt": "Write a short haiku about the ocean",
      "stream": true
  }'

try with zed get this error:

API request to http://127.0.0.1:8080/v1 failed: model should be in provider/model format

[akshaydeo]

@calvin2021y could you elaborate the issue clearly?

What I can gather from you are sending a/a as a provider + model but the base url is not supported by your local provider. Not sure how is it Bifrost error. A little bit of context would help here 🙇