Azure OpenAI config.toml keeps running into these errors~!!!

[LeoYoungChina]

this is the content of my config.toml file：

this is the errors keeps coming up:

[enyst]

The error says to check docker ps. Write docker ps in console and see what it says. Does it give an error or what does it say?

[LeoYoungChina]

The error says to check docker ps. Write docker ps in console and see what it says. Does it give an error or what does it say?

then i docker ps , it comes into 404 error, is there anything wrong with my Azure settings ? :

[enyst]

It does look like something is wrong with model settings. I think "deployment name" is wrong, it should be the name of the embeddings deployment in Azure, not chat GPT deployment. Please try to find that, and change it.

[LeoYoungChina]

It does look like something is wrong with model settings. I think "deployment name" is wrong, it should be the name of the embeddings deployment in Azure, not chat GPT deployment. Please try to find that, and change it.
is there any advices or full params example for azure open ai settings for us?

[LeoYoungChina]

It does look like something is wrong with model settings. I think "deployment name" is wrong, it should be the name of the embeddings deployment in Azure, not chat GPT deployment. Please try to find that, and change it.
is there any advices or full params example for azure open ai settings for us?

still getting 404 errors coming up all the time

[enyst]

I think it is again the wrong name: the model for embeddings is "text-embedding-ada-002". But the "deployment name" in Azure is a name of an Azure resource with this model that you have access to. It's a name probably similar with this, but not the same.
I'm confused too 😆 but really, look around at what they call deployment, related to embeddings.

[VishalHaibatpure1]

@LeoYoungChina : Please see if below instruction with screenshot from AzureStudio helps you

Please note: For LLM_EMBEDDING_DEPLOYMENT_NAME to work, you will have to do few changes mentioned here: #1027 (comment)

@enyst : Mostly this confusion on the deployment name and API version is happening for everyone. May be some detailed/special instructions will be required as part of Azure config doc for OpenDevin.

[enyst]

@enyst : Mostly this confusion on the deployment name and API version is happening for everyone. May be some detailed/special instructions will be required as part of Azure config doc for OpenDevin.

We have just added a guide here:
https://github.com/OpenDevin/OpenDevin/blob/main/docs/documentation/AZURE_LLM_GUIDE.md

Please take a look and let us know what you think. Is it missing something relevant? Maybe this screenshot? I think it could help! 😄

[Umpire2018]

@he-aook I believe above pic can solve your question.

@enyst Suggest to add the screenshot to AZURE_LLM_GUIDE.md to avoid misunderstanding. And i think DEPLOYMENT_NAME is little bit confused, maybe rename it to LLM_EMBEDDING_SERVICE_NAME , and LLM_EMBEDDING_MODEL to LLM_EMBEDDING_TYPE ?

[rbren]

Deduping all the azure OpenAI issues into this: #1042