Quickstart setup step 4 (not working)

[zaynpatel]

Discussed in #235

^{Originally posted by zaynpatel December 4, 2023}
I'm currently running the following command, copied from the documentation with the exception of a new localhost address:

url http://localhost:3000/inferences/llamacpp/loadmodel \
-H 'Content-Type: application/json' \
-d '{
    "llama_model_path": "/model/model/llama-2-7b-model.gguf",
    "ctx_len": 512,
    "ngl": 100
}'

I'm getting a 404 error which references that the inferences/llamacpp/loadmodel is not an available route.

{"url":"/inferences/llamacpp/loadmodel","statusCode":404,"statusMessage":"Cannot find any route matching /inferences/llamacpp/loadmodel.","message":"Cannot find any route matching /inferences/llamacpp/loadmodel."}% 

I'm curious about how to proceed and wonder how I can test what other load model links might be correct?

[fscardua]

Same problem here.

[0xSage]

@zaynpatel , @fscardua sorry about this. Cortex is in private preview and has breaking changes daily.

Can you both try the API playground: http://localhost:1337/api
for easy curling?

[louis-jan]

The URL has been changed; now, Cortex will serve the chat/completions endpoint at http://localhost:3000/inferences/server/loadmodel.

[0xSage]

closing as solved