TheBloke

/

Mistral-7B-Instruct-v0.1-GGUF

like 510

Hi, thank you for your work. I tried embedding endpoint and I got an error.
Query:

curl -X 'POST' \
  'https://limcheekin-mistral-7b-instruct-v0-1-gguf.hf.space/v1/embeddings' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "input": "The food was delicious and the waiter..."
}'

Error reponse:

{
  "error": {
    "message": "Llama model must be created with embedding=True to call this method",
    "type": "internal_server_error",
    "param": null,
    "code": null
  }
}

Hi, thank you for your work. I tried embedding endpoint and I got an error.
Query:

curl -X 'POST' \
  'https://limcheekin-mistral-7b-instruct-v0-1-gguf.hf.space/v1/embeddings' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "input": "The food was delicious and the waiter..."
}'

Error reponse:

{
  "error": {
    "message": "Llama model must be created with embedding=True to call this method",
    "type": "internal_server_error",
    "param": null,
    "code": null
  }
}

Yeah, I should stated clearly in the doc the embeddings endpoint has been disabled on purpose as I tested that the embeddings created by Llama models is NOT better than other open-source text embeddings models such as BAAI/bge-large-en, intfloat/e5-large-v2, sentence-transformers/all-MiniLM-L6-v2, sentence-transformers/all-mpnet-base-v2, etc. Hence, I created the Python package at https://github.com/limcheekin/open-text-embeddings.

Anyway, that's just my experience of few months ago and my current understanding, I just enabled (turn on) the embeddings endpoint and go ahead and test it out yourself and appreciate you share the result here.

Thank you.