Add README documentation

[Copilot]

The repository lacked user-facing documentation.

Changes

Added README.md covering:

• Project description (benchmarking framework for Guided Infilling Models)
• Installation via pip (regular and dev)
• Available benchmark types: CV parsing, regex matching, MCQA, perplexity
• Command examples for running MMLU-Pro, GPQA Diamond, and GIM-SFT benchmarks
• Development commands: linting, formatting, pre-commit

Original prompt

add a simple readme

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[Ki-Seki]

• install with pip is preferred for common users.
• add command examples
• remove license sections

@copilot

[Copilot]

• install with pip is preferred for common users.
• add command examples
• remove license sections ...

Updated in commit b72eeb9:

• Installation now uses pip instead of uv
• Added command examples for MMLU-Pro, GPQA Diamond, and GIM-SFT benchmarks
• Removed License and Copyright sections

[Copilot]

Pull request overview

Adds a new top-level README to provide user-facing documentation for GIMBench (a benchmarking framework for Guided Infilling Models), including installation, benchmark types, example run commands, and basic development workflows.

Changes:

• Added project overview and high-level feature/benchmark descriptions.
• Documented installation steps (pip + dev setup) and example benchmark invocation commands.
• Documented common development commands (lint, lint-fix, pre-commit) via make.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The development install instructions rely on make install-dev, which runs uv sync (see Makefile). Consider documenting uv as a prerequisite (or provide an alternative dev install command) so the steps are reproducible for new contributors.

[Copilot]

The GPQA example uses --model_type openai with gpt-4 but does not specify --base_url. The CLI default base URL is http://localhost:8000/v1, so this command would point at a local endpoint rather than the OpenAI API unless clarified. Suggest adding an explicit OpenAI base URL (or noting that openai here means an OpenAI-compatible local server).

Suggested change

	--model_name gpt-4 \
	--model_name gpt-4 \
	--base_url https://api.openai.com/v1 \