Using llama.cpp not working

[rtpt-romankarwacik]

My config looks like this:

name: Local Config
version: 1.0.0
schema: v1
models:
  - name: gpt-oss-20b-Q4_K_M.gguf
    model: gpt-oss-20b-Q4_K_M.gguf
    apiBase: http://192.168.124.12:8033
    roles:
      - chat
    provider: llama.cpp

Like described in https://docs.continue.dev/customize/model-providers/more/llamacpp .
If I send any chat message I get You must either implement templateMessages or _streamChat:

[Extension Host] [@continuedev] error: You must either implement templateMessages or _streamChat {"context":"llm_stream_chat","model":"gpt-oss-20b-Q4_K_M.gguf","provider":"llama.cpp","useOpenAIAdapter":false,"streamEnabled":true,"templateMessages":false}
console.ts:137 [Extension Host] Error: You must either implement templateMessages or _streamChat
	at LlamaCpp._streamChat (/usr/share/code/resources/app/extensions/continue.continue-1.2.10-linux-x64/out/extension.js:244121:17)
	at _streamChat.next (<anonymous>)
	at LlamaCpp.streamChat (/usr/share/code/resources/app/extensions/continue.continue-1.2.10-linux-x64/out/extension.js:244000:32)
	at streamChat.next (<anonymous>)
	at llmStreamChat (/usr/share/code/resources/app/extensions/continue.continue-1.2.10-linux-x64/out/extension.js:722159:29)
	at async od.handleMessage [as value] (/usr/share/code/resources/app/extensions/continue.continue-1.2.10-linux-x64/out/extension.js:760625:29)

No request is made to the server.

Anyone got an I idea how to fix this?

[rtpt-romankarwacik]

Ok using provider: openai instead solved this. This should probably be updated in the documentation.