host.today/ht-docker-ai
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(paddleocr-vl): add PaddleOCR-VL full pipeline Docker image and API server, plus integration tests and docker helpers

Browse Source
This commit is contained in:
Jürgen Kunz
2026-01-17 20:22:23 +00:00
parent addae20cbd
commit 80e6866442
12 changed files with 2414 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
test/test.bankstatements.combined.ts
View File

@@ -1,15 +1,23 @@
/**
* Bank statement extraction test using MiniCPM-V (visual) + PaddleOCR-VL (table recognition)
*
* This is the combined/dual-VLM approach that uses both models for consensus:
* - MiniCPM-V for visual extraction
* - PaddleOCR-VL for table recognition
*/
import { tap, expect } from '@git.zone/tstest/tapbundle';
import * as fs from 'fs';
import * as path from 'path';
import { execSync } from 'child_process';
import * as os from 'os';
import { ensurePaddleOcrVl, ensureMiniCpm } from './helpers/docker.js';
// Service URLs
const OLLAMA_URL = 'http://localhost:11434';
const PADDLEOCR_VL_URL = 'http://localhost:8000';
// Models
const MINICPM_MODEL = 'openbmb/minicpm-v4.5:q8_0';
const MINICPM_MODEL = 'minicpm-v:latest';
const PADDLEOCR_VL_MODEL = 'paddleocr-vl';
// Prompt for MiniCPM-V visual extraction
@@ -477,11 +485,18 @@ function findTestCases(): Array<{ name: string; pdfPath: string; jsonPath: strin
// Tests
tap.test('should connect to Ollama API', async () => {
const response = await fetch(`${OLLAMA_URL}/api/tags`);
expect(response.ok).toBeTrue();
const data = await response.json();
expect(data.models).toBeArray();
tap.test('setup: ensure Docker containers are running', async () => {
console.log('\n[Setup] Checking Docker containers...\n');
// Ensure PaddleOCR-VL is running (auto-detects GPU/CPU)
const paddleOk = await ensurePaddleOcrVl();
expect(paddleOk).toBeTrue();
// Ensure MiniCPM is running
const minicpmOk = await ensureMiniCpm();
expect(minicpmOk).toBeTrue();
console.log('\n[Setup] All containers ready!\n');
});
tap.test('should have MiniCPM-V 4.5 model loaded', async () => {
@@ -494,8 +509,7 @@ tap.test('should have MiniCPM-V 4.5 model loaded', async () => {
tap.test('should check PaddleOCR-VL availability', async () => {
const available = await isPaddleOCRVLAvailable();
console.log(`PaddleOCR-VL available: ${available}`);
// This test passes regardless - PaddleOCR-VL is optional
expect(true).toBeTrue();
expect(available).toBeTrue();
});
// Dynamic test for each PDF/JSON pair