Add Omni Reader Project#187

Merged

marwan37 merged 131 commits intomainfrom

Apr 9, 2025

Commits on Mar 28, 2025

Initial project setup with dependencies and license
marwan37
committed
Add project README with documentation and usage instructions
marwan37
committed
Add core utility functions for integrations
marwan37
committed
Add data schemas for the OCR pipeline
marwan37
committed
Add OCR comparison pipeline framework
marwan37
committed
Add model evaluation step for OCR comparison
marwan37
committed
Add main execution script for OCR pipeline
marwan37
committed
Add image encoding, metrics calculation and prompt utilities
marwan37
committed
Add OCR implementation steps for Gemma3 and Mistral models
marwan37
committed
Add Streamlit web interface for interactive OCR comparison
marwan37
committed
Add sample images for OCR testing
marwan37
committed
update README
marwan37
committed
Add configuration settings for OCR pipeline
marwan37
committed
Add standalone script for quick OCR comparison without ZenML
marwan37
committed
Add poetry.lock file to lock dependencies
marwan37
committed
Add pip requirements file for non-Poetry installations
marwan37
committed
Add ImageDescription pydantic model for the OCR pipeline
marwan37
committed
update readme
marwan37
committed
update assets dir structure
marwan37
committed

Commits on Mar 29, 2025

update prompt.py: add confidence for extracted text
marwan37
committed
add confidence field to ImageDescription
marwan37
committed
organize assets dir
marwan37
committed
remove integrations utils: remove mlflow, and docker for now
marwan37
committed
add pipelines __init__ file
marwan37
committed
add other init files
marwan37
committed
add run_ocr.py file: reusable ocr step for various models
marwan37
committed
delete run_gemma3_ocr and run_mistral_ocr in favor of unified run_ocr step
marwan37
committed
update ocr_pipeline to save ocr results and visualizations
marwan37
committed
update save ocr results step
marwan37
committed
add configs dir with ocr_config yaml file
marwan37
committed
add config util file
marwan37
committed
add io utils file for loading images/files and for saving results
marwan37
committed
add ocr model utils file for processing chat completions for various models
marwan37
committed
add step for loading images/files
marwan37
committed
add run.py file
marwan37
committed
pass each result dict to save_ocr_results step to resolve StepInterfaceError
marwan37
committed
update ocr_pipeline to pass model_names in addition to results when saving
marwan37
committed
refactor error_analysis and metrics under 1 file and remove confidence extraction method
marwan37
committed
add detailed html string containing ocr results, error analysis and other metrics
marwan37
committed
update config file with default image folder
marwan37
committed
update app.py
marwan37
committed
Add detailed metrics to streamlit UI app
marwan37
committed
Update README.md: fix broken links, and add project organization
marwan37
committed
remove pyproject.toml and poetry.lock in favor of single requirements.txt
marwan37
committed
update requirements.txt
marwan37
committed
remove mlflow from readme
marwan37
committed

Commits on Mar 30, 2025

delete unused assets
marwan37
committed
update prompt to clarify entitiy output
marwan37
committed

Commits on Mar 31, 2025

refactor app.py: simplify and remove entities/description
marwan37
committed
update assets
marwan37
committed
add model_info util forgetting model metadata throughout the app
marwan37
committed
update save results to dynamically render model info
marwan37
committed
move ImageDescription schema into prompt
marwan37
committed
update steps to integrate changes
marwan37
committed
update ocr pipeline
marwan37
committed
update run and run_compare_ocr entrypoint files
marwan37
committed
refactor main.py -- remove ground-trith-texts flag
marwan37
committed
update README.md
marwan37
committed
update ocr_config.yaml: remove image_patterns, update keys
marwan37
committed
update assets
marwan37
committed
delete config.yaml
marwan37
committed
update ocr model utils to process images with ollama api rather than client
marwan37
committed
remove running_from_ui param
marwan37
committed
add run_ollama_ocr_from_ui function for streamlit app
marwan37
committed
add docker settings for ocr pipeline
marwan37
committed
update run_compare_ocr
marwan37
committed
refactor run_ocr
marwan37
committed
add Dockerfile and .dockerignore
marwan37
committed
update requirements.txt
marwan37
committed
refactor: revert to use litellm+instructor for ollama models
marwan37
committed
edit api_key access for OpenAI client
marwan37
committed

Commits on Apr 1, 2025

add logos for different ocr models
marwan37
committed
update utils for multi-model ocr
marwan37
committed
update steps for multi-model ocr changes
marwan37
committed
add model_configs util for centralized model configuration and client fetching
marwan37
committed
rename ocr_model_utils to ocr_processing
marwan37
committed
update pipeline for multi-model ocr and update docker settings
marwan37
committed
update entrypoint run files
marwan37
committed
refactor streamlit app for readability and allow any number of models to compare
marwan37
committed
update README
marwan37
committed
update ocr_config.yaml to integrate new changes
marwan37
committed
update assets
marwan37
committed
update assets
marwan37
committed
improve UI for comparing and displaying multiple models
marwan37
committed
add extract json util
marwan37
committed
add .env.example
marwan37
committed
update pipeline
marwan37
committed
update ocr processing, metrics, and model configs
marwan37
committed
update utils and revamp evaluate_models
marwan37
committed
pass MODEL_CONFIGS in streamlit ocr processing
marwan37
committed
set defaults for ground_truth_model for project
marwan37
committed
add images for README
marwan37
committed
update requirements.txt
marwan37
committed
remove demo_models from utils
marwan37
committed
update README: add images, links, docs
marwan37
committed

Commits on Apr 3, 2025

simplify assets dir
marwan37
committed
delete unused assets
marwan37
committed
update steps
marwan37
committed
add loaders.py
marwan37
committed
update utils
marwan37
committed
separate batch ocr from evaluation into 2 pipelines
marwan37
committed
add ground_truth_texts samples matching images provided
marwan37
committed
remove run_compare_ocr and main.py in favor of run.py
marwan37
committed
update config.yaml to align with zenml config file definition
marwan37
committed
update README
marwan37
committed
update .env.example
marwan37
committed

Commits on Apr 4, 2025

re-integrate docker settings and cleanup config.yaml
marwan37
committed
update image links in README
marwan37
committed
update visualization img
marwan37
committed
update assets
marwan37
committed

Commits on Apr 7, 2025

refactor utils
marwan37
committed
move visualization logic into util file
marwan37
committed
refactor steps: remove local saving/loading and integrate artifacts
marwan37
committed
refactor pipelines: remove save/load from local dirs + split config files
marwan37
committed
split config into dedicated config files for each pipeline
marwan37
committed
add schemas dir
marwan37
committed
update requirements.txt
marwan37
committed
update run.py: simplify args, and integrate new config structure
marwan37
committed
delete main.py
marwan37
committed
update configs
marwan37
committed
add docker settings in pipeline definitions
marwan37
committed
update README
marwan37
committed
add analyse and Labour to .typos.toml
marwan37
committed

Commits on Apr 8, 2025