theovincent

import matplotlib.pyplot as plt
import numpy as np

from matplotlib.patches import Circle, RegularPolygon
from matplotlib.path import Path
from matplotlib.projections import register_projection
from matplotlib.projections.polar import PolarAxes
from matplotlib.spines import Spine
from matplotlib.transforms import Affine2D

theovincent

1. Feedback from the supervisee to the supervisor

Feel free to leave some questions empty if you don't feel like answering them.

General Feedback

a. If you could keep only one thing from your supervisor that you wish your next supervisor would have, what would it be?

b. If you could change only one thing of your supervisor that you wish your next supervisor would not have, what would it be?

c. Did you feel like your supervisor was motivated and committed during your thesis / project?

theovincent

Things to care about when replicating Dopamine's Atari pipeline:

I. Environmnent

$a.$ gym environment: full_action_space=False, frameskip = $1$, repeat_action_probability = $0.25$.

$b.$ reset: Stack $4$ frames. The three first are zeros, the last one is the first image.

$c.$ step: Generate $4$ frames for the internal environment. Sum up the reward. Stop if a terminal state is reached. For the last $2$ frames, take the gray scale version, then, apply a max pooling and then resize the image to $84$ by $84$ with the cv2.INTER_AREA interpolation.

Tip: use env.ale.getScreenGrayscale to get the frame from the ALE. It is faster.

theovincent

srun --gres gpu --job-name "requirements" --cpus-per-task 4 --mem-per-cpu 4000 --time 24:00:00 --pty bash

theovincent

export SQUEUE_FORMAT="%.8u %.14j %.13i %.2t %.11r %.10M %.10l %.4C %.4m %.9P %N"
export SINFO_FORMAT="%.10P %.4c %.6m %.13G %.5D"
export SPRIO_FORMAT="%.8u %.7i %.8Y %.5A %.9F %.7J %.9P %.3Q %.4N"
export SACCT_FORMAT="JobName%40,AveCPU,MaxVMSize,Submit,Start,State"