textflow
[NAACL 2025] Beyond End-to-End VLMs: Leveraging Intermediate Text Representations for Superior Flowchart Understanding
Learn to log and monitor TextFlow pipeline execution using its built-in logger. Track VQA, Textualizer, Reasoner, and Evaluation stages with timestamped files and colorized console output.
Understanding the Differences Between Reasoner Models in TextFlow's Textual ReasonerExplore TextFlow's Textual Reasoner: API vs local models. Understand tool-use differences between Claude 3.5-Sonnet, GPT-4o, Llama-3.1, and Mixtral for your NLP needs.
How to Integrate TextFlow with Your Own Applications: Complete Developer GuideIntegrate TextFlow into your Python apps with this developer guide. Programmatically convert flowchart images to text and perform visual question answering using modular pipeline components.
TextFlow Performance Optimization: 6 Best Practices for High-Throughput Flowchart ProcessingBoost TextFlow performance with 6 best practices. Learn to reduce processing time 2-3x by pre-encoding images and parallelizing API calls. Optimize your flowchart processing today.
How TextFlow Handles Different Flowchart Structures and Edge CasesDiscover how TextFlow handles diverse flowchart structures and edge cases like cycles and duplicates. Learn about its robust rendering strategies in flowchart.py for safe processing.
How to Visualize Intermediate Text Representations for Debugging in TextFlowDebug TextFlow effectively. Visualize intermediate text representations using JSON files, the built-in logger, or external diagram viewers like Mermaid and Graphviz.
How to Run Baseline VQA Comparison Against the TextFlow PipelineCompare VQA models with the TextFlow pipeline. Learn how to run the baseline VQA and TextFlow comparison for accurate results using junyiye/textflow.
How TextFlow's Modular Architecture Enables Swapping ComponentsDiscover how TextFlow's modular architecture lets you swap components like vision-language models and encoders easily. Edit one file, no refactoring needed, to customize your pipeline.
Performance Difference Between API Models and Local Models in TextFlow: A Complete AnalysisAnalyze TextFlow performance: API models offer 80%+ accuracy with tool use, while local models reach 60-70% lacking graph queries. Understand the gap for visual reasoning.
How to Debug Vision Textualizer Failures When Generated Text Representations Are InvalidFix Vision Textualizer invalid text generation. Debug prompt misconfigurations, missing code blocks, or image encoding issues by examining raw responses and extraction logic.
Prompt Templates for the Vision Textualizer and Textual Reasoner in TextFlowDiscover TextFlow's prompt templates for Vision Textualizer and Textual Reasoner. Generate diagrams and perform visual reasoning with these powerful LLM tools.
How to Use Ground Truth Text Representations for Benchmarking in TextFlowBenchmark your LLM using ground truth text representations in TextFlow. Pass --textualizer Ground-Truth and provide ideal text descriptions for accurate evaluation.
Have a question about this repo?
These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:
curl -s "https://instagit.com/install.md" 
Maintain an open-source project? Get it listed too →