yashash045 commited on
Commit
b86a8f9
·
verified ·
1 Parent(s): db00b75

Kaggle frontier eval: llama33_70b

Browse files
Files changed (1) hide show
  1. eval_frontier_llama33_70b.json +760 -0
eval_frontier_llama33_70b.json ADDED
@@ -0,0 +1,760 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "meta-llama/Llama-3.3-70B-Instruct",
3
+ "tasks": {
4
+ "clean_deploy": {
5
+ "seed_0": {
6
+ "task": "clean_deploy",
7
+ "seed_offset": 0,
8
+ "steps": 15,
9
+ "reward_sum": 0.6367,
10
+ "rewards_per_step": [
11
+ 0.03,
12
+ 0.0277,
13
+ 0.01,
14
+ -0.01,
15
+ 0.07,
16
+ 0.1497,
17
+ 0.07,
18
+ 0.1499,
19
+ 0.0198,
20
+ 0.0175,
21
+ 0.0211,
22
+ 0.0217,
23
+ 0.0107,
24
+ 0.0258,
25
+ 0.0228
26
+ ],
27
+ "roles_used": [
28
+ "ops",
29
+ "sre"
30
+ ],
31
+ "num_roles_used": 2,
32
+ "all_3_modes_used": false,
33
+ "done": true,
34
+ "success": true,
35
+ "healthy_ratio": 1.0,
36
+ "steps_to_recovery": 0,
37
+ "initial_health": 100.0
38
+ },
39
+ "seed_1": {
40
+ "task": "clean_deploy",
41
+ "seed_offset": 1,
42
+ "steps": 20,
43
+ "reward_sum": -1.8686,
44
+ "rewards_per_step": [
45
+ 0.0432,
46
+ 0.026,
47
+ -0.33,
48
+ 0.0435,
49
+ -0.0185,
50
+ -0.0182,
51
+ -0.0182,
52
+ -0.0175,
53
+ -0.0169,
54
+ 0.041,
55
+ 0.0072,
56
+ -0.017,
57
+ -0.0247,
58
+ -0.0205,
59
+ -0.016,
60
+ -0.016,
61
+ 0.008,
62
+ 0.008,
63
+ -0.016,
64
+ -1.516
65
+ ],
66
+ "roles_used": [
67
+ "sre"
68
+ ],
69
+ "num_roles_used": 1,
70
+ "all_3_modes_used": false,
71
+ "done": true,
72
+ "success": false,
73
+ "healthy_ratio": 0.8,
74
+ "steps_to_recovery": 0,
75
+ "initial_health": 100.0
76
+ },
77
+ "seed_2": {
78
+ "task": "clean_deploy",
79
+ "seed_offset": 2,
80
+ "steps": 20,
81
+ "reward_sum": -1.9204,
82
+ "rewards_per_step": [
83
+ 0.0312,
84
+ 0.026,
85
+ -0.33,
86
+ 0.0483,
87
+ -0.0185,
88
+ -0.0182,
89
+ -0.0182,
90
+ -0.0175,
91
+ -0.0169,
92
+ -0.0169,
93
+ -0.0168,
94
+ -0.017,
95
+ -0.0247,
96
+ -0.0205,
97
+ -0.016,
98
+ -0.016,
99
+ -0.016,
100
+ 0.0453,
101
+ 0.008,
102
+ -1.516
103
+ ],
104
+ "roles_used": [
105
+ "sre"
106
+ ],
107
+ "num_roles_used": 1,
108
+ "all_3_modes_used": false,
109
+ "done": true,
110
+ "success": false,
111
+ "healthy_ratio": 0.8,
112
+ "steps_to_recovery": 0,
113
+ "initial_health": 100.0
114
+ }
115
+ },
116
+ "broken_pipeline": {
117
+ "seed_0": {
118
+ "task": "broken_pipeline",
119
+ "seed_offset": 0,
120
+ "steps": 20,
121
+ "reward_sum": -1.8584,
122
+ "rewards_per_step": [
123
+ 0.0312,
124
+ 0.026,
125
+ -0.05,
126
+ -0.33,
127
+ 0.0483,
128
+ -0.0185,
129
+ -0.0182,
130
+ -0.0182,
131
+ -0.0175,
132
+ 0.0443,
133
+ 0.0071,
134
+ -0.0168,
135
+ 0.007,
136
+ -0.0007,
137
+ -0.0205,
138
+ -0.016,
139
+ -0.016,
140
+ -0.016,
141
+ 0.008,
142
+ -1.492
143
+ ],
144
+ "roles_used": [
145
+ "sre"
146
+ ],
147
+ "num_roles_used": 1,
148
+ "all_3_modes_used": false,
149
+ "done": true,
150
+ "success": false,
151
+ "healthy_ratio": 0.8,
152
+ "steps_to_recovery": 0,
153
+ "initial_health": 100.0
154
+ },
155
+ "seed_1": {
156
+ "task": "broken_pipeline",
157
+ "seed_offset": 1,
158
+ "steps": 20,
159
+ "reward_sum": -1.8244,
160
+ "rewards_per_step": [
161
+ 0.0312,
162
+ 0.026,
163
+ -0.33,
164
+ 0.0483,
165
+ -0.0185,
166
+ -0.0182,
167
+ -0.0182,
168
+ -0.0175,
169
+ -0.0169,
170
+ -0.0169,
171
+ -0.0168,
172
+ 0.0443,
173
+ -0.0007,
174
+ -0.0205,
175
+ -0.016,
176
+ 0.008,
177
+ 0.008,
178
+ 0.008,
179
+ 0.008,
180
+ -1.516
181
+ ],
182
+ "roles_used": [
183
+ "sre"
184
+ ],
185
+ "num_roles_used": 1,
186
+ "all_3_modes_used": false,
187
+ "done": true,
188
+ "success": false,
189
+ "healthy_ratio": 0.8,
190
+ "steps_to_recovery": 0,
191
+ "initial_health": 100.0
192
+ },
193
+ "seed_2": {
194
+ "task": "broken_pipeline",
195
+ "seed_offset": 2,
196
+ "steps": 20,
197
+ "reward_sum": -1.8584,
198
+ "rewards_per_step": [
199
+ 0.0312,
200
+ 0.026,
201
+ -0.05,
202
+ -0.33,
203
+ 0.0483,
204
+ -0.0185,
205
+ -0.0182,
206
+ -0.0182,
207
+ -0.0175,
208
+ -0.0169,
209
+ -0.0169,
210
+ -0.0168,
211
+ 0.0443,
212
+ -0.0007,
213
+ -0.0205,
214
+ -0.016,
215
+ 0.008,
216
+ 0.008,
217
+ 0.008,
218
+ -1.492
219
+ ],
220
+ "roles_used": [
221
+ "sre"
222
+ ],
223
+ "num_roles_used": 1,
224
+ "all_3_modes_used": false,
225
+ "done": true,
226
+ "success": false,
227
+ "healthy_ratio": 0.8,
228
+ "steps_to_recovery": 0,
229
+ "initial_health": 100.0
230
+ }
231
+ },
232
+ "judgment_call": {
233
+ "seed_0": {
234
+ "task": "judgment_call",
235
+ "seed_offset": 0,
236
+ "steps": 20,
237
+ "reward_sum": -1.8244,
238
+ "rewards_per_step": [
239
+ 0.0312,
240
+ 0.026,
241
+ -0.33,
242
+ 0.0483,
243
+ -0.0185,
244
+ -0.0182,
245
+ -0.0182,
246
+ -0.0175,
247
+ -0.0169,
248
+ -0.0169,
249
+ 0.0445,
250
+ 0.007,
251
+ -0.0247,
252
+ -0.0205,
253
+ -0.016,
254
+ 0.008,
255
+ 0.008,
256
+ 0.008,
257
+ 0.008,
258
+ -1.516
259
+ ],
260
+ "roles_used": [
261
+ "sre"
262
+ ],
263
+ "num_roles_used": 1,
264
+ "all_3_modes_used": false,
265
+ "done": true,
266
+ "success": false,
267
+ "healthy_ratio": 0.8,
268
+ "steps_to_recovery": 0,
269
+ "initial_health": 100.0
270
+ },
271
+ "seed_1": {
272
+ "task": "judgment_call",
273
+ "seed_offset": 1,
274
+ "steps": 20,
275
+ "reward_sum": -1.8104,
276
+ "rewards_per_step": [
277
+ 0.0312,
278
+ 0.026,
279
+ -0.05,
280
+ -0.33,
281
+ 0.0483,
282
+ -0.0185,
283
+ 0.0431,
284
+ 0.0058,
285
+ -0.0175,
286
+ -0.0169,
287
+ 0.0071,
288
+ 0.0072,
289
+ -0.017,
290
+ -0.0247,
291
+ 0.0035,
292
+ 0.008,
293
+ 0.008,
294
+ 0.008,
295
+ -0.016,
296
+ -1.516
297
+ ],
298
+ "roles_used": [
299
+ "sre"
300
+ ],
301
+ "num_roles_used": 1,
302
+ "all_3_modes_used": false,
303
+ "done": true,
304
+ "success": false,
305
+ "healthy_ratio": 0.8,
306
+ "steps_to_recovery": 0,
307
+ "initial_health": 100.0
308
+ },
309
+ "seed_2": {
310
+ "task": "judgment_call",
311
+ "seed_offset": 2,
312
+ "steps": 20,
313
+ "reward_sum": -1.8104,
314
+ "rewards_per_step": [
315
+ 0.0312,
316
+ 0.026,
317
+ -0.05,
318
+ -0.33,
319
+ 0.0483,
320
+ 0.0428,
321
+ 0.0058,
322
+ -0.0182,
323
+ 0.0065,
324
+ 0.0071,
325
+ -0.0169,
326
+ -0.0168,
327
+ 0.007,
328
+ -0.0007,
329
+ -0.0205,
330
+ -0.016,
331
+ 0.008,
332
+ 0.008,
333
+ -0.016,
334
+ -1.516
335
+ ],
336
+ "roles_used": [
337
+ "sre"
338
+ ],
339
+ "num_roles_used": 1,
340
+ "all_3_modes_used": false,
341
+ "done": true,
342
+ "success": false,
343
+ "healthy_ratio": 0.8,
344
+ "steps_to_recovery": 0,
345
+ "initial_health": 100.0
346
+ }
347
+ },
348
+ "cascading_failure": {
349
+ "seed_0": {
350
+ "task": "cascading_failure",
351
+ "seed_offset": 0,
352
+ "steps": 20,
353
+ "reward_sum": -1.8004,
354
+ "rewards_per_step": [
355
+ 0.0312,
356
+ 0.026,
357
+ -0.33,
358
+ 0.0483,
359
+ -0.0185,
360
+ -0.0182,
361
+ -0.0182,
362
+ -0.0175,
363
+ -0.0169,
364
+ -0.0169,
365
+ 0.0445,
366
+ 0.007,
367
+ -0.0247,
368
+ -0.0205,
369
+ -0.016,
370
+ 0.008,
371
+ 0.008,
372
+ 0.008,
373
+ 0.008,
374
+ -1.492
375
+ ],
376
+ "roles_used": [
377
+ "sre"
378
+ ],
379
+ "num_roles_used": 1,
380
+ "all_3_modes_used": false,
381
+ "done": true,
382
+ "success": false,
383
+ "healthy_ratio": 0.8,
384
+ "steps_to_recovery": 0,
385
+ "initial_health": 100.0
386
+ },
387
+ "seed_1": {
388
+ "task": "cascading_failure",
389
+ "seed_offset": 1,
390
+ "steps": 20,
391
+ "reward_sum": -1.8584,
392
+ "rewards_per_step": [
393
+ 0.0312,
394
+ 0.026,
395
+ -0.05,
396
+ -0.33,
397
+ 0.0483,
398
+ -0.0185,
399
+ -0.0182,
400
+ -0.0182,
401
+ -0.0175,
402
+ -0.0169,
403
+ -0.0169,
404
+ 0.0445,
405
+ 0.007,
406
+ -0.0247,
407
+ -0.0205,
408
+ -0.016,
409
+ 0.008,
410
+ 0.008,
411
+ 0.008,
412
+ -1.492
413
+ ],
414
+ "roles_used": [
415
+ "sre"
416
+ ],
417
+ "num_roles_used": 1,
418
+ "all_3_modes_used": false,
419
+ "done": true,
420
+ "success": false,
421
+ "healthy_ratio": 0.8,
422
+ "steps_to_recovery": 0,
423
+ "initial_health": 100.0
424
+ },
425
+ "seed_2": {
426
+ "task": "cascading_failure",
427
+ "seed_offset": 2,
428
+ "steps": 15,
429
+ "reward_sum": -1.3768,
430
+ "rewards_per_step": [
431
+ 0.0886,
432
+ 0.0629,
433
+ 0.0084,
434
+ -0.0177,
435
+ 0.0082,
436
+ 0.008,
437
+ 0.0079,
438
+ 0.0078,
439
+ -0.0182,
440
+ -0.0183,
441
+ -0.0184,
442
+ -0.0185,
443
+ 0.0075,
444
+ 0.0074,
445
+ -1.4926
446
+ ],
447
+ "roles_used": [
448
+ "sre"
449
+ ],
450
+ "num_roles_used": 1,
451
+ "all_3_modes_used": false,
452
+ "done": true,
453
+ "success": false,
454
+ "healthy_ratio": 0.6,
455
+ "steps_to_recovery": 21,
456
+ "initial_health": 60.0
457
+ }
458
+ },
459
+ "capacity_crisis": {
460
+ "seed_0": {
461
+ "task": "capacity_crisis",
462
+ "seed_offset": 0,
463
+ "steps": 20,
464
+ "reward_sum": -1.9552,
465
+ "rewards_per_step": [
466
+ 0.0308,
467
+ 0.026,
468
+ -0.05,
469
+ -0.33,
470
+ 0.0479,
471
+ -0.0177,
472
+ 0.0432,
473
+ 0.0058,
474
+ -0.0175,
475
+ -0.0176,
476
+ -0.0172,
477
+ -0.0171,
478
+ -0.0171,
479
+ -0.0163,
480
+ -0.0165,
481
+ -0.0242,
482
+ -0.0199,
483
+ -0.016,
484
+ -0.016,
485
+ -1.516
486
+ ],
487
+ "roles_used": [
488
+ "sre"
489
+ ],
490
+ "num_roles_used": 1,
491
+ "all_3_modes_used": false,
492
+ "done": true,
493
+ "success": false,
494
+ "healthy_ratio": 0.8,
495
+ "steps_to_recovery": 0,
496
+ "initial_health": 100.0
497
+ },
498
+ "seed_1": {
499
+ "task": "capacity_crisis",
500
+ "seed_offset": 1,
501
+ "steps": 20,
502
+ "reward_sum": -1.8592,
503
+ "rewards_per_step": [
504
+ 0.0308,
505
+ 0.026,
506
+ -0.05,
507
+ -0.33,
508
+ 0.0479,
509
+ -0.0177,
510
+ -0.018,
511
+ -0.0182,
512
+ -0.0175,
513
+ -0.0176,
514
+ -0.0172,
515
+ -0.0171,
516
+ 0.0069,
517
+ 0.0077,
518
+ -0.0165,
519
+ 0.0371,
520
+ 0.0041,
521
+ -0.016,
522
+ 0.008,
523
+ -1.492
524
+ ],
525
+ "roles_used": [
526
+ "sre"
527
+ ],
528
+ "num_roles_used": 1,
529
+ "all_3_modes_used": false,
530
+ "done": true,
531
+ "success": false,
532
+ "healthy_ratio": 0.8,
533
+ "steps_to_recovery": 0,
534
+ "initial_health": 100.0
535
+ },
536
+ "seed_2": {
537
+ "task": "capacity_crisis",
538
+ "seed_offset": 2,
539
+ "steps": 20,
540
+ "reward_sum": -1.9072,
541
+ "rewards_per_step": [
542
+ 0.0308,
543
+ 0.026,
544
+ -0.05,
545
+ -0.33,
546
+ 0.0479,
547
+ -0.0177,
548
+ -0.018,
549
+ -0.0182,
550
+ -0.0175,
551
+ -0.0176,
552
+ 0.0441,
553
+ 0.0069,
554
+ -0.0171,
555
+ -0.0163,
556
+ -0.0165,
557
+ -0.0242,
558
+ 0.0041,
559
+ 0.008,
560
+ -0.016,
561
+ -1.516
562
+ ],
563
+ "roles_used": [
564
+ "sre"
565
+ ],
566
+ "num_roles_used": 1,
567
+ "all_3_modes_used": false,
568
+ "done": true,
569
+ "success": false,
570
+ "healthy_ratio": 0.8,
571
+ "steps_to_recovery": 0,
572
+ "initial_health": 100.0
573
+ }
574
+ },
575
+ "random_incident": {
576
+ "seed_0": {
577
+ "task": "random_incident",
578
+ "seed_offset": 0,
579
+ "steps": 20,
580
+ "reward_sum": -1.8592,
581
+ "rewards_per_step": [
582
+ 0.0308,
583
+ 0.026,
584
+ -0.05,
585
+ -0.33,
586
+ 0.0479,
587
+ -0.0177,
588
+ -0.018,
589
+ -0.0182,
590
+ -0.0175,
591
+ -0.0176,
592
+ -0.0172,
593
+ -0.0171,
594
+ 0.0442,
595
+ 0.0077,
596
+ -0.0165,
597
+ -0.0002,
598
+ 0.0041,
599
+ -0.016,
600
+ 0.008,
601
+ -1.492
602
+ ],
603
+ "roles_used": [
604
+ "sre"
605
+ ],
606
+ "num_roles_used": 1,
607
+ "all_3_modes_used": false,
608
+ "done": true,
609
+ "success": false,
610
+ "healthy_ratio": 0.8,
611
+ "steps_to_recovery": 0,
612
+ "initial_health": 100.0
613
+ },
614
+ "seed_1": {
615
+ "task": "random_incident",
616
+ "seed_offset": 1,
617
+ "steps": 20,
618
+ "reward_sum": -1.8592,
619
+ "rewards_per_step": [
620
+ 0.0308,
621
+ 0.026,
622
+ -0.05,
623
+ -0.33,
624
+ 0.0479,
625
+ 0.0436,
626
+ 0.006,
627
+ -0.0182,
628
+ -0.0175,
629
+ -0.0176,
630
+ -0.0172,
631
+ -0.0171,
632
+ 0.0069,
633
+ 0.0077,
634
+ 0.0075,
635
+ -0.0002,
636
+ -0.0199,
637
+ -0.016,
638
+ -0.016,
639
+ -1.516
640
+ ],
641
+ "roles_used": [
642
+ "sre"
643
+ ],
644
+ "num_roles_used": 1,
645
+ "all_3_modes_used": false,
646
+ "done": true,
647
+ "success": false,
648
+ "healthy_ratio": 0.8,
649
+ "steps_to_recovery": 0,
650
+ "initial_health": 100.0
651
+ },
652
+ "seed_2": {
653
+ "task": "random_incident",
654
+ "seed_offset": 2,
655
+ "steps": 20,
656
+ "reward_sum": -1.9072,
657
+ "rewards_per_step": [
658
+ 0.0308,
659
+ 0.026,
660
+ -0.05,
661
+ -0.33,
662
+ 0.0479,
663
+ -0.0177,
664
+ -0.018,
665
+ -0.0182,
666
+ -0.0175,
667
+ -0.0176,
668
+ -0.0172,
669
+ 0.0442,
670
+ 0.0069,
671
+ -0.0163,
672
+ -0.0165,
673
+ -0.0242,
674
+ -0.0199,
675
+ -0.016,
676
+ 0.008,
677
+ -1.492
678
+ ],
679
+ "roles_used": [
680
+ "sre"
681
+ ],
682
+ "num_roles_used": 1,
683
+ "all_3_modes_used": false,
684
+ "done": true,
685
+ "success": false,
686
+ "healthy_ratio": 0.8,
687
+ "steps_to_recovery": 0,
688
+ "initial_health": 100.0
689
+ }
690
+ }
691
+ },
692
+ "summary": {
693
+ "clean_deploy": {
694
+ "avg_reward": -1.0508,
695
+ "std_reward": 1.1934,
696
+ "avg_steps": 18.33,
697
+ "avg_roles_used": 1.33,
698
+ "success_rate": 0.333,
699
+ "all_3_modes_hit_rate": 0,
700
+ "avg_steps_to_recovery": 0.0,
701
+ "recovery_episodes_count": 0,
702
+ "n": 3
703
+ },
704
+ "broken_pipeline": {
705
+ "avg_reward": -1.8471,
706
+ "std_reward": 0.016,
707
+ "avg_steps": 20,
708
+ "avg_roles_used": 1,
709
+ "success_rate": 0,
710
+ "all_3_modes_hit_rate": 0,
711
+ "avg_steps_to_recovery": 0.0,
712
+ "recovery_episodes_count": 0,
713
+ "n": 3
714
+ },
715
+ "judgment_call": {
716
+ "avg_reward": -1.8151,
717
+ "std_reward": 0.0066,
718
+ "avg_steps": 20,
719
+ "avg_roles_used": 1,
720
+ "success_rate": 0,
721
+ "all_3_modes_hit_rate": 0,
722
+ "avg_steps_to_recovery": 0.0,
723
+ "recovery_episodes_count": 0,
724
+ "n": 3
725
+ },
726
+ "cascading_failure": {
727
+ "avg_reward": -1.6785,
728
+ "std_reward": 0.2147,
729
+ "avg_steps": 18.33,
730
+ "avg_roles_used": 1,
731
+ "success_rate": 0,
732
+ "all_3_modes_hit_rate": 0,
733
+ "avg_steps_to_recovery": 21,
734
+ "recovery_episodes_count": 1,
735
+ "n": 3
736
+ },
737
+ "capacity_crisis": {
738
+ "avg_reward": -1.9072,
739
+ "std_reward": 0.0392,
740
+ "avg_steps": 20,
741
+ "avg_roles_used": 1,
742
+ "success_rate": 0,
743
+ "all_3_modes_hit_rate": 0,
744
+ "avg_steps_to_recovery": 0.0,
745
+ "recovery_episodes_count": 0,
746
+ "n": 3
747
+ },
748
+ "random_incident": {
749
+ "avg_reward": -1.8752,
750
+ "std_reward": 0.0226,
751
+ "avg_steps": 20,
752
+ "avg_roles_used": 1,
753
+ "success_rate": 0,
754
+ "all_3_modes_hit_rate": 0,
755
+ "avg_steps_to_recovery": 0.0,
756
+ "recovery_episodes_count": 0,
757
+ "n": 3
758
+ }
759
+ }
760
+ }