Implementation:FMInference FlexLLMGen Data Wrangling Install

Field	Value
Sources	FlexLLMGen
Domains	Environment_Setup, Data_Wrangling
Last Updated	2026-02-09 00:00 GMT

Overview

External tool documentation for the data wrangling dependency installation script provided by FlexLLMGen.

Description

install.sh installs Python packages (pandas, sentence-transformers, rich, pyarrow) via pip and clones the HazyResearch fm_data_tasks repository which contains benchmark datasets for entity matching, data imputation, and error detection tasks.

Code Reference

• Source: flexllmgen/apps/data_wrangle/install.sh, Lines: 1-8
• Import: No Python import (shell script)

Script content:

#!/bin/bash
pip install pandas sentence-transformers rich pyarrow
git clone https://github.com/HazyResearch/fm_data_tasks.git

I/O Contract

Direction	Name	Description
Input	Python environment	Python environment with FlexLLMGen installed
Input	Internet access	Required for PyPI and GitHub
Output	Python packages	pandas, sentence-transformers, rich, pyarrow installed
Output	Dataset directory	fm_data_tasks/ directory with benchmark datasets

Usage Examples

# Run from the data_wrangle directory
cd flexllmgen/apps/data_wrangle
bash install.sh

# Verify
python -c "import pandas; import sentence_transformers; print('OK')"
ls fm_data_tasks/data/

Related Pages

• Principle:FMInference_FlexLLMGen_Data_Wrangling_Setup