File size: 790 Bytes
8cc48e6
5ddd335
8cc48e6
2a01ff5
 
 
 
 
 
8cc48e6
 
 
2a01ff5
8cc48e6
 
2a01ff5
5ddd335
 
 
 
2a01ff5
 
 
 
8cc48e6
 
 
 
 
 
 
 
 
5ddd335
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
name: multimodal-rag-executor
version: 0.1.0
description: >
  description: >
  Production-ready Jina Executor for multimodal RAG with research papers. 
  Upload PDFs, extract text/images, analyze with BLIP vision AI, index with FAISS, 
  and answer questions using FLAN-T5. 100% FREE models, no API costs required.
  Supports both local inference and optional HuggingFace API.

keywords:
  - jina
  - rag
  - multimodal
  - pdf
  - faiss
  - research-papers
  - huggingface
  - free-ai
  - sentence-transformers
  - blip
  - flan-t5
  - image-captioning
  - question-answering
  - vector-search
author:
  name: Girinath V
  email: [email protected]
license: MIT
type: executor
platform: python
runtime: "python>=3.10"
entrypoint: executor.py
install:
  - pip install -r requirements.txt