fix #30: allow optional "stop" field for oai

[Mardak]

Fix #30 r? @ggerganov

Previously omitting "stop" asserts:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
  "messages": [{
    "role": "user",
    "content": "Write a limerick about python exceptions"
  }]
}'

llama.cpp/server/json.h:21313: assert(it != m_value.object->end()) failed

[Mardak]

This avoids the assert, but some .choices[0].message.content include <|im_sep|> (which also happens without this fix but requesting with "stop": null)

[Mardak]

Oh, I see the original commit 401dd08 cherry-picked ggml-org/llama.cpp#4198 so maybe I should fix this upstream?

[Doublespaced1]

UnsubscribeOn Dec 1, 2023, at 5:09 PM, Ed Lee ***@***.***> wrote: Oh, I see the original commit 401dd08 cherry-picked ggml-org/llama.cpp#4198 so maybe I should fix this upstream? —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[jart]

I checked the OpenAI API documentation. https://platform.openai.com/docs/api-reference/chat/create I believe your fix is correct. I've confirmed the issue is also present ggerganov/llama.cpp too. It'd make me very happy to see you send a PR to the upstream project too. Thanks!