WorkflowLearner
by u3588064
WorkflowLearner aims to teach LLM Agents how to work with PSR.exe's MHT files. It uses MHT files recorded by PSR.exe to enable large language models (LLMs) to learn workflows by parsing these MHT files.
Last updated: N/A
What is WorkflowLearner?
WorkflowLearner is a project that allows Large Language Models (LLMs) to learn and understand user workflows by parsing MHT files generated by Windows' Problem Steps Recorder (PSR.exe). It provides scripts to parse these files and train LLMs to recognize and replicate the recorded workflows.
How to use WorkflowLearner?
First, record user actions using PSR.exe to generate an MHT file. Then, use the parse_mht.py script to extract workflow information from the MHT file. Finally, use the learn_workflow.py script to train an LLM on the parsed data.
Key features of WorkflowLearner
Records user operations using PSR.exe and generates MHT files.
Parses MHT files and extracts workflow information.
Uses LLMs to learn and understand workflows.
Provides scripts for parsing and learning workflows.
Use cases of WorkflowLearner
Automating repetitive tasks based on recorded workflows.
Training AI assistants to perform specific procedures.
Creating interactive tutorials based on user actions.
Analyzing user behavior and identifying workflow bottlenecks.
FAQ from WorkflowLearner
What is PSR.exe?
What is PSR.exe?
PSR.exe is the Problem Steps Recorder in Windows, which records user actions and generates MHT files.
What is an MHT file?
What is an MHT file?
An MHT file is a web archive format that stores a webpage and its associated resources in a single file.
How do I install the dependencies?
How do I install the dependencies?
Run pip install -r requirements.txt after cloning the repository.
How do I parse an MHT file?
How do I parse an MHT file?
Use the command python parse_mht.py path/to/your/file.mht.
How do I use the LLM to learn the workflow?
How do I use the LLM to learn the workflow?
Use the command python learn_workflow.py path/to/parsed/data.json.