Add more pre-built images for t2v-trasformers

Support
1

Hello, how can I modify this docker-compose.yml file to add another pre-built image like semitechnologies/transformers-inference:sentence-transformers-all-MiniLM-L12-v2?

---
version: '3.4'
services:
  weaviate:
    command:
      - --host
      - 0.0.0.0
      - --port
      - '8080'
      - --scheme
      - http
    image: semitechnologies/weaviate:1.20.4
    ports:
      - 8080:8080
    restart: on-failure:0
    environment:
      TRANSFORMERS_INFERENCE_API: 'http://t2v-transformers:8080'
      QUERY_DEFAULTS_LIMIT: 25
      AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED: 'true'
      PERSISTENCE_DATA_PATH: '/var/lib/weaviate'
      DEFAULT_VECTORIZER_MODULE: 'text2vec-transformers'
      ENABLE_MODULES: 'text2vec-transformers'
      CLUSTER_HOSTNAME: 'node1'
  t2v-transformers:
    image: semitechnologies/transformers-inference:sentence-transformers-multi-qa-mpnet-base-cos-v1
    environment:
      ENABLE_CUDA: '1'
...
2

Hi Paolo - have you seen the Docker-compose install page here (Docker Compose | Weaviate - vector database). I think it has what you’re after, showing you what a different image line looks like.

The list of images can be found here: text2vec-transformers | Weaviate - vector database :slight_smile:

3

Hi, in the page you linked me, I didn’t find anything that could answer my question. I only find examples using a single image, or am I mistaken?

4

Oh, maybe I misunderstood your question then. Are you looking to see if you can use multiple images?

5

Yes, exactly. I’m sorry if I didn’t explain myself well.

6

Right. I don’t think we support that, but I will separately confirm.

I’m curious - what might be the reason for having multiple images? Is it to allow switching between vectorizers?

7

Exactly, switching between vectorizers.