Update README.md

README.md

@@ -48,7 +48,7 @@ Coming soon: a more detailed explanation of the integral projection model used!
 
 # Training
 
-We use YAML files to specify training instances, examples of these files can be found on `hyperpars`. 
-Key components there are: specifying the RL algorithm (e.g. PPO or TQC), specifying the environment trained on (e.g. `GreenCrab-v2`) and the training duration (6M time steps in our examples).
-To train using this framework, use the command `python scripts/train.py path/to/file.yml`.
+`python scripts/train.py -f ../hyperpars/ppo-gcse.yml`
+
+Alternatively, `bash scripts/train_algos.sh` trains using several algorithms in parallel.