example-rl-feedback-workflow

Example RL feedback chain for training LLM. Incorportates feedback from compiler, klee, codeql in form of text files accessible by the training program.

Name		Name	Last commit message	Last commit date
generated_code		generated_code
README.md		README.md
commands-to-use-after-prerequisites.txt		commands-to-use-after-prerequisites.txt
compile-bitcode-executable.sh		compile-bitcode-executable.sh
config.json		config.json
gpu_requirements.txt		gpu_requirements.txt
install_llms.py		install_llms.py
klee_c_wrapper.py		klee_c_wrapper.py
klee_requirements.txt		klee_requirements.txt
klee_reward_calc.py		klee_reward_calc.py
klee_test.sh		klee_test.sh
prerequisites-setup.sh		prerequisites-setup.sh
run_codeql.py		run_codeql.py
run_llm.py		run_llm.py

Navigation Menu

Repository navigation

About

Repository files navigation

example-rl-feedback-workflow

About

Releases

Packages

Contributors

Languages

Latest commit

History

Folders and files

Repository files navigation

example-rl-feedback-workflow

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages