ML-Agents - 'Making a New Learning Environment' problems and unrecognized arguments: config.yaml

Systems Settings:
Windows 10
Unity 2019.3.1f1
Barracuda '0.4.0-preview'
Anaconda - Virtual environment "ml-agents" with:
Python 3.7.6
Tensorflow version: 2.0.1
ml-agents: 0.13.1,
ml-agents-envs: 0.13.1
Communicator API: API-13

Describe the problems:
1.- I created an New unity project -New Learning Project- where I followed the instructions of 'Making a New Learning Environment' (https://github.com/Unity-Technologi...lease/docs/Learning-Environment-Create-New.md). I dragged the ML-Agents folder to the project and technically it goes smoothly, but I have a problem with the RollerAcademy script in the editor. It's different from the one in the tutorial. (See Picture) . I'm also posting the scripts I have in the scene.


2 .- When I pass

(ml-agents) D:\>mlagents-learn D:/Unity Projects/ML-agents/ml-agents/config/config.yaml --run-id=RollerBall-1 --train

in my terminal, as written in the documentation, I get the message:

Code (CSharp):

1. WARNING:tensorflow:From c:\users\dcastill\appdata\local\continuum\anaconda3\envs\ml-agents\lib\site-packages\tensorflow_core\python\compat\v2_compat.py:65: disable_resource_variables (from tensorflow.python.ops.variable_scope) is deprecated and will be removed in a future version.
2. Instructions for updating:
3. non-resource variables are not supported in the long term
4.  
5.  
6.                         ▄▄▄▓▓▓▓
7.                    ╓▓▓▓▓▓▓█▓▓▓▓▓
8.               ,▄▄▄m▀▀▀'  ,▓▓▓▀▓▓▄                           ▓▓▓  ▓▓▌
9.            ▄▓▓▓▀'      ▄▓▓▀  ▓▓▓      ▄▄     ▄▄ ,▄▄ ▄▄▄▄   ,▄▄ ▄▓▓▌▄ ▄▄▄    ,▄▄
10.           ▄▓▓▓▀        ▄▓▓▀   ▐▓▓▌     ▓▓▌   ▐▓▓ ▐▓▓▓▀▀▀▓▓▌ ▓▓▓ ▀▓▓▌▀ ^▓▓▌  ╒▓▓▌
11.         ▄▓▓▓▓▓▄▄▄▄▄▄▄▄▓▓▓      ▓▀      ▓▓▌   ▐▓▓ ▐▓▓    ▓▓▓ ▓▓▓  ▓▓▌   ▐▓▓▄ ▓▓▌
12.         ▀▓▓▓▓▀▀▀▀▀▀▀▀▀▀▓▓▄     ▓▓      ▓▓▌   ▐▓▓ ▐▓▓    ▓▓▓ ▓▓▓  ▓▓▌    ▐▓▓▐▓▓
13.           ^█▓▓▓        ▀▓▓▄   ▐▓▓▌     ▓▓▓▓▄▓▓▓▓ ▐▓▓    ▓▓▓ ▓▓▓  ▓▓▓▄    ▓▓▓▓`
14.             '▀▓▓▓▄      ^▓▓▓  ▓▓▓       └▀▀▀▀ ▀▀ ^▀▀    `▀▀ `▀▀   '▀▀    ▐▓▓▌
15.                ▀▀▀▀▓▄▄▄   ▓▓▓▓▓▓,                                      ▓▓▓▓▀
16.                    `▀█▓▓▓▓▓▓▓▓▓▌
17.                         ¬`▀▀▀█▓
18.  
19.  
20. Version information:
21.   ml-agents: 0.13.1,
22.   ml-agents-envs: 0.13.1,
23.   Communicator API: API-13,
24.   TensorFlow: 2.0.1
25. usage: mlagents-learn [-h] [--env ENV_PATH] [--curriculum CURRICULUM_FOLDER]
26.                       [--sampler SAMPLER_FILE_PATH]
27.                       [--keep-checkpoints KEEP_CHECKPOINTS] [--lesson LESSON]
28.                       [--load] [--run-id RUN_ID] [--num-runs NUM_RUNS]
29.                       [--save-freq SAVE_FREQ] [--seed SEED] [--train]
30.                       [--base-port BASE_PORT] [--num-envs NUM_ENVS]
31.                       [--docker-target-name DOCKER_TARGET_NAME]
32.                       [--no-graphics] [--debug] [--multi-gpu] [--env-args ...]
33.                       [--cpu] [--version] [--width WIDTH] [--height HEIGHT]
34.                       [--quality-level QUALITY_LEVEL]
35.                       [--time-scale TIME_SCALE]
36.                       [--target-frame-rate TARGET_FRAME_RATE]
37.                       trainer_config_path
38. mlagents-learn: error: unrecognized arguments: Projects/ML-agents/ml-agents/config/config.yaml

and the config.yaml file has

Code (CSharp):

1. default:
2.     trainer: ppo
3.     batch_size: 10
4.     beta: 5.0e-3
5.     buffer_size: 100
6.     epsilon: 0.2
7.     hidden_units: 128
8.     lambd: 0.95
9.     learning_rate: 3.0e-4
10.     learning_rate_schedule: linear
11.     max_steps: 5.0e4
12.     memory_size: 256
13.     normalize: false
14.     num_epoch: 3
15.     num_layers: 2
16.     time_horizon: 64
17.     sequence_length: 64
18.     summary_freq: 1000
19.     use_recurrent: false
20.     vis_encode_type: simple
21.     reward_signals:
22.         extrinsic:
23.             strength: 1.0
24.             gamma: 0.99

I don't know what to do. Any help provided will be appreciated.

 

Try putting the path in quotes "D:/Unity Projects/ML-agents/ml-agents/config/config.yaml", because it contains a blank space. Engine configuration settings are now (as of 0.13 I think) command line options and were removed from the Academy class. The tutorial was probably made before hat change.
https://github.com/Unity-Technologi...ng-ML-Agents.md#command-line-training-options

 

Instead of this try this -> mlagents-learn D:/Unity Projects/ML-agents/ml-agents/config/trainerconfig.yaml --run-id=default --train


As there is no file named config.yaml in config folder and there is no run-id named "RollerBall-1" in trainerconfig.yaml, unless it is declared manually.