vincent-stragier/llm-text-classifier

How vincent-stragier/llm-text-classifier works

This repository is a script-based LLM text-classification experiment. The main workflow downloads synthetic request datasets, splits them and renders model-specific prompts, then runs each prompt through a configured local LLM and stores class scores alongside ground truth. Model configurations and prompt templates parameterize preprocessing and inference; model weights can also be downloaded separately. The classifier evaluates candidate class-token continuations using logits, while shared utilities aggregate scores and map prompt paths to result paths. Edges follow the README workflow and shown calls; the dataset, model-config, and prompt-template contents were not sampled.

Main entry points: do_classification.py Classification runner · do_preprocess.py Prompt generation · download_datasets.py Dataset downloader · classifiers.py LLM classifier

Components 17

Data acquisition

Synthetic datasets
Dataset downloader download_datasets.py

Prompt preparation

Prompt generation do_preprocess.py
Prompt templates prompts
Generated prompts prompts

Classification

Classification runner do_classification.py
LLM classifier classifiers.py

Configuration and outputs

Model configurations models_configs
Model downloader download_all_llms.py
LLM weights
Score aggregation utils.py
Classification results results
Result path mapping utils.py

Other components

Experimenter
Dataset host
Hugging Face Hub
llama.cpp runtime

Connections 22

Experimenter → Dataset downloader: runs
Dataset downloader → Dataset host: downloads from download_datasets.py
Dataset downloader → Synthetic datasets: writes datasets download_datasets.py
Experimenter → Prompt generation: runs
Prompt generation → Synthetic datasets: loads and splits do_preprocess.py
Prompt generation → Prompt templates: renders do_preprocess.py
Prompt generation → Model configurations: loads do_preprocess.py
Prompt generation → Generated prompts: writes prompts do_preprocess.py
Experimenter → Model downloader: runs optionally download_all_llms.py
Model downloader → Model configurations: loads download_all_llms.py
Model downloader → Hugging Face Hub: downloads from download_all_llms.py
Hugging Face Hub → LLM weights: provides weights
Experimenter → Classification runner: runs
Classification runner → Model configurations: loads do_classification.py
Classification runner → Generated prompts: reads prompts do_classification.py
Classification runner → Synthetic datasets: loads classes do_classification.py
Classification runner → Hugging Face Hub: fetches model do_classification.py
Classification runner → LLM classifier: calls classify do_classification.py
LLM classifier → llama.cpp runtime: decodes tokens classifiers.py
LLM classifier → Score aggregation: aggregates logits classifiers.py
Classification runner → Result path mapping: maps result path do_classification.py
Classification runner → Classification results: writes scores do_classification.py

Drawn by GitDiagram on November 11, 2026 from the repository's file tree, README and source files. Read as Markdown ·