-
Notifications
You must be signed in to change notification settings - Fork 11
/
index.html
515 lines (482 loc) · 20.5 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Tutorial: Tools for Robotic Reinforcement Learning | ICRA 2022</title>
<link rel="stylesheet" href="https://bulma.io/vendor/fontawesome-free-5.15.2-web/css/all.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bulma@0.9.3/css/bulma.min.css">
</head>
<body>
<section class="hero is-light is-medium">
<!-- Hero head: will stick at the top -->
<div class="hero-head">
<nav class="navbar">
<div class="container">
<div class="navbar-brand">
<a class="navbar-item" href="https://www.icra2022.org/">
<img src="https://www.icra2022.org/sites/default/themes/icra2022/resources/img/logos/title-conf.svg"
alt="Logo"
width="300px" style="max-height:400px">
</a>
<!-- <span class="navbar-burger" data-target="navbarMenuHeroA">
<span></span>
<span></span>
<span></span>
</span> -->
</div>
<div id="navbarMenuHeroA" class="navbar-menu">
<div class="navbar-end">
<a class="navbar-item" href="#goals">
Goals
</a>
<a class="navbar-item" href="#schedule">
Schedule
</a>
<!-- <a class="navbar-item" href="#notebooks">
Notebooks
</a> -->
<a class="navbar-item" href="#speakers">
Speakers
</a>
<a class="navbar-item" href="#organizers">
Organizers
</a>
<span class="navbar-item">
<a class="button is-light is-inverted" href="https://github.com/araffin/tools-for-robotic-rl-icra2022" target="_blank">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Github Repo</span>
</a>
</span>
</div>
</div>
</div>
</nav>
</div>
<!-- Hero content: will be in the middle -->
<div class="hero-body">
<div class="container has-text-centered">
<h1 class="title">
Tutorial: Tools for Robotic Reinforcement Learning
</h1>
<p class="subtitle">
Hands-on RL for Robotics with
<a href="https://github.com/eager-dev/eagerx" target="_blank">EAGERx</a>
and
<a href="https://github.com/DLR-RM/stable-baselines3" target="_blank">Stable-Baselines3</a>
</p>
</div>
<div class="container has-text-centered">
<h3>ICRA 2022, 8:30 AM - 5:20 PM (UTC−4), May 23, 2022 Philadelphia (PA), USA - Room 115A</h3>
</div>
</div>
<!-- Hero footer: will stick at the bottom -->
<!-- <div class="hero-foot">
<nav class="tabs">
<div class="container">
<ul>
<li class="is-active"><a>Overview</a></li>
<li><a>Modifiers</a></li>
<li><a>Grid</a></li>
<li><a>Elements</a></li>
<li><a>Components</a></li>
<li><a>Layout</a></li>
</ul>
</div>
</nav>
</div> -->
</section>
<div class="container is-max-desktop">
<section class="section has-text-centered">
<h2 class="title">Motivation</h2>
<h3 class="subtitle">
Reinforcement learning (RL) methods have received much attention due to impressive results in many robotic applications.
While RL promises learning-based control of near-optimal behaviors in theory, successful learning can elude practitioners due to various implementation challenges.
Even if the best-suited learning method was selected, learning performance can nonetheless disappoint due to badly chosen hyper-parameters or an unreliable implementation of the algorithm.
Furthermore, a learning task can be made unnecessarily hard by incorrect specifications.<br>
This full-day tutorial points-out these practical pitfalls and introduces the audience to the tools for robotic RL that will aid roboticists in successfully solving robotic learning tasks, both in simulation and the real-world.
</h3>
</section>
<section class="section has-text-centered">
<h2 class="title" id="goals">Goals</h2>
<h3 class="subtitle">
We will cover the use of
<a href="https://github.com/eager-dev/eagerx" target="_blank">Engine Agnostic Gym Environment for Robotics (EAGERx)</a>
to define and create tasks that work both in simulation and on a real robot, and then learn to use the
<a href="https://github.com/DLR-RM/stable-baselines3" target="_blank">Stable-Baselines3 (SB3)</a>
library to solve it with SOTA algorithms, following best practices. <br>
This tutorial will cover: creating tasks in EAGERx, basic usage of SB3, automatic hyperparameter optimization and managing RL experiments.
</h3>
</section>
<section class="section has-text-centered">
<h2 class="title">Requirements</h2>
<h3 class="subtitle">
Basic knowledge of reinforcement learning and python programming is required.
</h3>
</section>
<section class="section has-text-centered">
<h2 class="title">Videos</h2>
<h3 class="subtitle">
<a href="https://www.youtube.com/playlist?list=PL42jkf1t1F7etDiYXWC5Q77yIuVYhXNoy" target="_blank">
Replay of the tutorial can be found on YouTube
</a>
</h3>
<div class="video">
<iframe width="560" height="315" src="https://www.youtube.com/embed/videoseries?list=PL42jkf1t1F7etDiYXWC5Q77yIuVYhXNoy" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
</div>
</section>
<section class="section">
<h2 class="title has-text-centered" id="schedule">Schedule</h2>
<h3 class="subtitle has-text-centered">
May 23, 8:30 AM - 5:20 PM (UTC−4), Room 115A
</h3>
<table class="table is-striped is-fullwidth is-hoverable is-bordered">
<thead>
<tr>
<th>Time</th>
<th>Talk</th>
<th>Comments</th>
</tr>
</thead>
<tbody>
<tr>
<td>8:30-8:35</td>
<td>Introduction</td>
<td></td>
</tr>
<tr>
<td>8:35-9:45</td>
<td>
<a href="https://araffin.github.io/slides/icra22-gym-sb3-quickstart/">
Getting Started with Gym and RL in practice
</a>
</td>
<td>Presenter: Antonin Raffin</td>
</tr>
<tr>
<td>9:30-10:30</td>
<td>
<a href="https://drive.google.com/file/d/19ImRxp8SfbTLtMDdFwYY__DH7txqlDd0/view">
Accelerating physics simulators for Robotics Reinforcement Learning
</a>
</td>
<td>
<a href="https://drive.google.com/file/d/19ImRxp8SfbTLtMDdFwYY__DH7txqlDd0/view">
Invited Speaker: Erwin Coumans
</a>
</td>
</tr>
<tr>
<td>10:45-11:15</td>
<td>
<a
href="https://colab.research.google.com/github/araffin/tools-for-robotic-rl-icra2022/blob/main/notebooks/icra_hands_on_sb3.ipynb">
Hands-on Session with Gym and SB3
</a>
</td>
<td>
<a
href="https://colab.research.google.com/github/araffin/tools-for-robotic-rl-icra2022/blob/main/notebooks/icra_hands_on_sb3.ipynb">
with colab notebooks
</a>
</td>
</tr>
<tr>
<td>11:15-11:25</td>
<td>Break</td>
<td></td>
</tr>
<tr>
<td>11:25-11:55</td>
<td>
<a href="https://docs.google.com/presentation/d/1Q66oV7YYFght82h-Poe56tKL27h7fuvC/edit#slide=id.p1">
Getting Started with EAGERx Part I
</a>
-
<a href="https://docs.google.com/presentation/d/18GMTxdoVaPjU3OoZCLNX7yKXsXGz1Sx-/edit#slide=id.p1">
Part II
</a>
</td>
<td>
Presenter: Jelle Luijkx
</td>
</tr>
<tr>
<td>11:55-12:30</td>
<td>
<a
href="https://colab.research.google.com/github/eager-dev/eagerx_tutorials/blob/master/tutorials/icra/getting_started.ipynb">
Hands-on Session with EAGERx
</a>
</td>
<td>
<a
href="https://colab.research.google.com/github/eager-dev/eagerx_tutorials/blob/master/tutorials/icra/getting_started.ipynb">
with colab notebooks
</a>
</td>
</tr>
<tr>
<td>12:30-13:30</td>
<td>Lunch Break</td>
<td></td>
</tr>
<tr>
<td>13:30-14:30</td>
<td>safe-control-gym: a Unified Benchmark Suite for Safe Learning-based Control and Reinforcement Learning</td>
<td>Invited Speaker: Angela Schoellig</td>
</tr>
<tr>
<td>14:30-15:15</td>
<td>
<a href="https://araffin.github.io/slides/icra22-hyperparam-opt/">
Automatic Hyperparameter Optimization
</a>
</td>
<td>Presenter: Antonin Raffin</td>
</tr>
<tr>
<td>15:15-16:00</td>
<td>
<a
href="https://colab.research.google.com/github/araffin/tools-for-robotic-rl-icra2022/blob/main/notebooks/optuna_lab.ipynb">
Hyperparameter Tuning with Optuna
</a>
</td>
<td>
<a
href="https://colab.research.google.com/github/araffin/tools-for-robotic-rl-icra2022/blob/main/notebooks/optuna_lab.ipynb">
with colab notebooks
</a>
</td>
</tr>
<tr>
<td>16:00-16:15</td>
<td>Break</td>
<td></td>
</tr>
<tr>
<td>16:15-17:15</td>
<td>
<a
href="https://colab.research.google.com/github/eager-dev/eagerx_tutorials/blob/master/tutorials/icra/advanced_usage.ipynb">
EAGERx Advanced usage
</a>
</td>
<td>
<a
href="https://colab.research.google.com/github/eager-dev/eagerx_tutorials/blob/master/tutorials/icra/advanced_usage.ipynb">
with colab notebooks
</a>
</td>
</tr>
<tr>
<td>17:15-17:20</td>
<td>Closing remarks</td>
<td></td>
</tr>
</tbody>
</table>
</section>
<!-- <section class="section">
<h2 class="title has-text-centered" id="notebooks">Notebooks</h2>
<h3 class="title has-text-centered">Stable-Baselines3 (SB3)</h3>
<ol>
<li>
<a
href="https://colab.research.google.com/github/araffin/tools-for-robotic-rl-icra2022/blob/main/notebooks/icra_hands_on_sb3.ipynb">
Gym/Stable
Baselines3 Getting Started
</a>
</li>
<li>
<a
href="https://colab.research.google.com/github/araffin/tools-for-robotic-rl-icra2022/blob/main/notebooks/optuna_lab.ipynb">
Hyperparameter
tuning with Optuna
</a>
</li>
</ol> -->
</section>
<section class="section">
<h2 class="title has-text-centered" id="speakers">Speakers</h2>
<article class="media">
<figure class="media-left">
<p class="image is-128x128">
<img src="https://avatars.githubusercontent.com/u/725468?v=4">
</p>
</figure>
<div class="media-content">
<div class="content">
<p>
<strong>Erwin Coumans</strong>
<small>
<a href="https://twitter.com/erwincoumans">
@erwincoumans
</a>
</small>
<br>
Erwin Coumans is creator of the Bullet physics engine,
former member of the Google Brain team,
and now works in the NVIDIA Omniverse team.
His interests include real-time physics simulation research and development,
with a focus on robotics and machine learning.
</p>
</div>
</div>
</article>
<article class="media">
<figure class="media-left">
<p class="image is-128x128">
<img src="https://scholar.googleusercontent.com/citations?view_op=medium_photo&user=QMfeRz0AAAAJ&citpid=4">
</p>
</figure>
<div class="media-content">
<div class="content">
<p>
<strong>Angela Schoellig</strong>
<small>
<a href="https://twitter.com/angelaschoellig">
@angelaschoellig
</a>
</small>
<br>
Angela Schoellig is an Associate Professor at the University of Toronto Institute for Aerospace Studies and a Faculty Member of the Vector Institute for Artificial Intelligence.
She conducts research at the intersection of robotics, controls, and machine learning. Her goal is to enhance the performance, safety, and autonomy of robots by enabling them to learn from past experiments and from each other.
</p>
</div>
</div>
</article>
</section>
<section class="section">
<h2 class="title has-text-centered" id="organizers">Organizers</h2>
<article class="media">
<figure class="media-left">
<p class="image is-128x128">
<img src="https://media-exp1.licdn.com/dms/image/C5603AQHnL2olLCKDoQ/profile-displayphoto-shrink_200_200/0/1605105302591?e=2147483647&v=beta&t=QGBBEf-qHBlHSIiobOhM5CZIiSBt4WYJsmaO_fD1FG0">
</p>
</figure>
<div class="media-content">
<div class="content">
<p>
<strong>Jelle Luijkx</strong>
<small>
<a href="https://github.com/jelledouwe">
@jelledouwe
</a>
</small>
<br>
Jelle is a PhD candidate at the Cognitive Robotics department of the Delft University of Technology.
He is working on deep learning tools for robot control within the OpenDR project and is co-creator of the Engine Agnostic Gym Environment for Robotics (EAGERx) toolkit.
</p>
</div>
</div>
</article>
<article class="media">
<figure class="media-left">
<p class="image is-128x128">
<img src="https://scholar.googleusercontent.com/citations?view_op=medium_photo&user=rLHzGRMAAAAJ&citpid=1">
</p>
</figure>
<div class="media-content">
<div class="content">
<p>
<strong>Bas Van der Heijden</strong>
<small>
<a href="https://github.com/bheijden">
@bheijden
</a>
</small>
<br>
Bas is a PhD candidate at TU Delft working on robotics and reinforcement learning.
He is co-creator of the Engine Agnostic Gym Environment for Robotics (EAGERx) toolkit.
</p>
</div>
</div>
</article>
<article class="media">
<figure class="media-left">
<p class="image is-128x128">
<img src="https://araffin.github.io/authors/admin/avatar_hu33d8f2710ea4928d295bd08cdc05f6eb_60040_270x270_fill_q90_lanczos_center.jpg">
</p>
</figure>
<div class="media-content">
<div class="content">
<p>
<strong>Antonin Raffin</strong>
<small>
<a href="https://twitter.com/araffin2">
@araffin2
</a>
</small>
<br>
Antonin Raffin is a Research Engineer in Robotics and Machine Learning at the German Aerospace Center (DLR).
He was previously working on state representation learning in the ENSTA robotics lab (U2IS) where he co-created the Stable-Baselines library with Ashley Hill. His research focus is now on applying reinforcement learning directly on real robots, for which he continues to maintain the Stable-Baselines3 library.
</p>
</div>
</div>
</article>
<article class="media">
<figure class="media-left">
<p class="image is-128x128">
<img src="https://scholar.googleusercontent.com/citations?view_op=medium_photo&user=XOWZzUcAAAAJ&citpid=8">
</p>
</figure>
<div class="media-content">
<div class="content">
<p>
<strong>Jens Kober </strong>
<br>
Jens Kober is an associate professor at the Cognitive
Robotics department, 3mE, TU Delft, Netherlands. He worked as a
postdoctoral scholar jointly at the CoR-Lab, Bielefeld University,
Germany and at the Honda Research Institute Europe, Germany. He
graduated in 2012 with a PhD Degree in Engineering from TU Darmstadt and
the MPI for Intelligent Systems. For his research he received the
annually awarded Georges Giralt PhD Award for the best PhD thesis in
robotics in Europe, the 2018 IEEE RAS Early Academic Career Award, and
has received an ERC Starting grant. His research interests include motor
skill learning, (deep) reinforcement learning, imitation learning,
interactive learning, and machine learning for control.
</p>
</div>
</div>
</article>
</section>
<section class="section has-text-centered">
<h2 class="title">Acknowledgements</h2>
<h3 class="subtitle">
The tutorial is supported by the EU H2020 projects
<i>
<a href="https://www.veridream.eu/">
VERtical Innovation in the Domain of Robotics Enabled by Artificial intelligence Methods
</a>
</i>
and
<i>
<a href="https://opendr.eu/">OpenDR</a>
</i>.
</h3>
<h3 class="subtitle">
The tutorial is also supported by the
<a href="https://www.ieee-ras.org/robot-learning">
IEEE RAS Technical Committee on Robot Learning
</a>
.
</h3>
</section>
</div>
<footer class="footer">
<div class="content has-text-centered">
<p>
Template made by <a href="https://araffin.github.io/">Antonin Raffin</a>
based on <i>Bulma</i>. The source code is licensed
<a href="http://opensource.org/licenses/mit-license.php">MIT</a>. The website content
is licensed <a href="http://creativecommons.org/licenses/by-nc-sa/4.0/">CC BY NC SA 4.0</a>.
</p>
</div>
</footer>
</body>
</html>