Hi @Stephanieewelu,
Came across AgenticLearnPro on pypi. I was exploring RL frameworks that could be adapted beyond academic environments, and the way you structured the environment-agent loop here feels refreshingly modular.
Use Case I’m Exploring:
I’m prototyping a CI/CD tuning agent — basically using RL to dynamically adjust deployment schedules and pipeline configurations. Your QLearningAgent class was clean enough to slot into early-stage tests, with minor environment adaptations.
What stood out:
• Reward handling is transparent enough for fast debugging.
• The SimpleEnv design made it easy to replace states/actions with a DevOps domain.
• The Streamlit demo made testing concepts feel intuitive before hard-coding anything.
Suggestions (if you’re considering extensions):
• Optional logging hooks for custom metrics or model checkpointing.
• A plug-in environment registry to make it easier to swap domains.
• Maybe a tiny README section on custom environments?
Hi @Stephanieewelu,
Came across AgenticLearnPro on pypi. I was exploring RL frameworks that could be adapted beyond academic environments, and the way you structured the environment-agent loop here feels refreshingly modular.
Use Case I’m Exploring:
I’m prototyping a CI/CD tuning agent — basically using RL to dynamically adjust deployment schedules and pipeline configurations. Your QLearningAgent class was clean enough to slot into early-stage tests, with minor environment adaptations.
What stood out:
• Reward handling is transparent enough for fast debugging.
• The SimpleEnv design made it easy to replace states/actions with a DevOps domain.
• The Streamlit demo made testing concepts feel intuitive before hard-coding anything.
Suggestions (if you’re considering extensions):
• Optional logging hooks for custom metrics or model checkpointing.
• A plug-in environment registry to make it easier to swap domains.
• Maybe a tiny README section on custom environments?