From 9d393e687af55404ff57b00be1f5d5468af16eb3 Mon Sep 17 00:00:00 2001 From: Theresa Eimer Date: Thu, 30 May 2024 17:06:48 +0200 Subject: [PATCH] finish example readme --- docs/contributing.rst | 2 +- examples/Readme.md | 10 ++++++++-- 2 files changed, 9 insertions(+), 3 deletions(-) diff --git a/docs/contributing.rst b/docs/contributing.rst index e582053ea..4fc50161c 100644 --- a/docs/contributing.rst +++ b/docs/contributing.rst @@ -1 +1 @@ -.. include:: ../CONTRIBUTING.rst +.. mdinclude:: ../CONTRIBUTING.md diff --git a/examples/Readme.md b/examples/Readme.md index e45c9e04f..77c60a42d 100644 --- a/examples/Readme.md +++ b/examples/Readme.md @@ -159,7 +159,7 @@ As you can see, most of the defaults are decided by the environment and algorith In the current config, we do a single training interval consisting of the total number of environment steps suggested for our target domain. If we want to instead run a schedule of length 10 with each schedule segment taking 10e4 steps, we can change the configuration like this: ```bash -python run_heuristic_schedule.py autorl.n_steps=10 autorl.total_timesteps=10000 +python run_heuristic_schedule.py autorl.n_steps=10 autorl.n_total_timesteps=10000 ``` ## 3. Reactive Schedules @@ -170,4 +170,10 @@ Lastly, we can also adjust the hyperparameters based on algorithm statistics. In python run_reactive_schedule.py ``` -To actually configure to w \ No newline at end of file +To actually configure to what information ARLBench returns about the RL algorithm's internal state, we can use the 'state' features key - in this case, we want to add the gradient norm and variance like this: + +```bash +python run_reactive_schedule.py "autorl.state_features=['grad_info']" +``` + +Now we can build a schedule that takes the gradient information into account. \ No newline at end of file