logprobs, best_of and echo parameters are not available on gpt-35-turbo model

[loshammo]

logprobs, best_of and echo parameters are not available on gpt-35-turbo models How to solve this problem? I am using azure openai

[feiyun0112]

you use wrong service, please see

#650

[loshammo]

you use wrong service, please see

#650

大神，怎么切换到ChatCompletionService ？

我是执行这个例子时报的错

[evchaki]

@loshammo , good call out here. The Simple chat summary, book creator, auth and api and GitHub repo Q&A does require the use of a text completion model. You will need to use something like d-003 for the model.

[loshammo]

@loshammo , good call out here. The Simple chat summary, book creator, auth and api and GitHub repo Q&A does require the use of a text completion model. You will need to use something like d-003 for the model.

Well, I tried the text avinci-003 model and it works.

Does this repo not consider compatibility with the gpt-35 turbo model? Gpt-35 turbo has sufficient capabilities in most scenarios and is relatively inexpensive

[feiyun0112]

，好在这里呼唤。简单聊天摘要、书籍创建者、身份验证和 API 以及 GitHub 存储库问答确实需要使用文本完成模型。您需要为模型使用 d-003 之类的东西。

好吧，我尝试了文本 avinci-003 模型，它可以工作。

这个回购不考虑与 gpt-35 涡轮增压模型的兼容性吗？GPT-35涡轮增压器在大多数情况下具有足够的能力，并且相对便宜

you can use this sample Copilot Chat Sample Application

[JadynWong]

您使用了错误的服务，请参阅。 #650

大神，怎么切换到ChatCompletionService ？

我是执行这个例子时报的错

For gpt-35-turbo or gpt4, replace AddAzureTextCompletionService with AddAzureChatCompletionService.

See

https://github.com/microsoft/semantic-kernel/blob/6afcbe25b9aa63d9952f63936962784ee4434510/samples/dotnet/kernel-syntax-examples/Example27_SemanticFunctionsUsingChatGPT.cs#L19-L28

[loshammo]

Thank you very much. I understand the problem now