Skip to content

common: add default reranker presets #13352

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

ca1yz
Copy link

@ca1yz ca1yz commented May 7, 2025

This commit adds default reranker presets for the following models:

  • bge-reranker-v2-m3 (most frequently downloaded reranker model)
  • jina-reranker-v1-turbo-en (is provided by ggml-org on their huggingface repository)

Example: Running a server with bge-reranker-v2-m3 for reranking:

./build/bin/llama-server --rerank-bge-v2-m3-default

Inputing this example from jina-reranker-v2-base-multilingual returns a JSON response with the scores

curl -X POST http://127.0.0.1:8080/v1/rerank \
  -H "Content-Type: application/json" \
  -d '{
    "query": "Organic skincare products for sensitive skin",
    "documents": [
    "Organic skincare for sensitive skin with aloe vera and chamomile.",
    "New makeup trends focus on bold colors and innovative techniques",
    "Bio-Hautpflege für empfindliche Haut mit Aloe Vera und Kamille",
    "Neue Make-up-Trends setzen auf kräftige Farben und innovative Techniken",
    "Cuidado de la piel orgánico para piel sensible con aloe vera y manzanilla",
    "Las nuevas tendencias de maquillaje se centran en colores vivos y técnicas innovadoras",
    "针对敏感肌专门设计的天然有机护肤产品",
    "新的化妆趋势注重鲜艳的颜色和创新的技巧",
    "敏感肌のために特別に設計された天然有機スキンケア製品",
    "新しいメイクのトレンドは鮮やかな色と革新的な技術に焦点を当てています"
  ]
  }'
{"model":"gpt-3.5-turbo","object":"list","usage":{"prompt_tokens":265,"total_tokens":265},"results":
[{"index":0,"relevance_score":4.414063453674316},{"index":1,"relevance_score":-11.041223526000977},
{"index":2,"relevance_score":1.0591806173324585},{"index":3,"relevance_score":-11.04470157623291},
{"index":4,"relevance_score":3.870026111602783},{"index":5,"relevance_score":-11.040743827819824},
{"index":6,"relevance_score":5.819011688232422},{"index":7,"relevance_score":-11.039907455444336},
{"index":8,"relevance_score":6.67466926574707},{"index":9,"relevance_score":-11.036961555480957}]}%

Refs: #10932

{"--rerank-bge-v2-m3-default"},
string_format("use default bge-reranker-v2-m3 (note: can download weights from the internet)"),
[](common_params & params) {
params.model.hf_repo = "gpustack/bge-reranker-v2-m3-GGUF";
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should host this model in ggml-org. I'll make a copy when GGUF My Repo is back online.

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should host this model in ggml-org. I'll make a copy when GGUF My Repo is back online.

it is now back online

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants