[BUG] FIM with LiteLLM: Default provider parameters persisting after adding new provider

[mrblackoutz]

Describe the bug
When using LiteLLM with FIM (auto-complete), after deleting all providers and adding a new one, the extension sends requests with default provider parameters instead of the configured ones. The workaround is to reset providers and edit the default provider instead of creating a new one.

To Reproduce
Steps to reproduce the behavior:

1. Delete all existing providers
2. Add a new provider for FIM
3. Try to use FIM auto-complete
4. Check debug logs - requests are using default provider parameters
5. Reset providers
6. Edit default provider instead of creating new one
7. FIM works correctly

Expected behavior
FIM should work correctly with newly added providers without requiring the workaround of editing the default provider.

API Provider

• Provider: LiteLLM Proxy
• Setup command:

docker run --name LiteLLMProxy --platform linux/arm64/v8 \
  -v <any>:/app/config.yaml \
  -e GEMINI_API_KEY=AIXXXXXXXXXXXXXX \
  -p 4000:4000 -d ghcr.io/berriai/litellm:main-latest \
  --config /app/config.yaml --detailed_debug

Twinny Configuration Used:

Label: Ollama FIM
Type: fim
FIM Template: custom-template
Provider: litellm
Protocol: http
Model Name: gemini-2.0-flash-lite-preview-02-05
Hostname: localhost
Port: 4000
API Path: /chat/completions
API Key: 
[ X ] Repository level

Chat or Auto Complete?
FIM (auto-complete)

Model Name
gemini-2.0-flash-lite-preview-02-05

Desktop:

• OS: Windows_NT x64 24H2 10.0.26100.2454
• VSCode Version: 1.97.1