dsh-eval-guard 0.7.4 → 0.7.6

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (31) hide show
  1. package/ISSUES.md +4 -2
  2. package/README.md +3 -3
  3. package/assets/neural-pendulum/SPEC.md +40 -42
  4. package/assets/neural-pendulum/TASK.md +34 -47
  5. package/assets/neural-pendulum/sim.cpp +97 -45
  6. package/assets/neural-pendulum/sim_agent +0 -0
  7. package/assets/neural-pendulum/sim_eval +0 -0
  8. package/assets/neural-pendulum/workspace.zip +0 -0
  9. package/lib/index.js +27 -10
  10. package/neural-pendulum-eval/README.md +4 -3
  11. package/neural-pendulum-eval/run_eval.py +133 -70
  12. package/neural-pendulum-eval/sim_eval +0 -0
  13. package/package.json +1 -1
  14. package/assets/neural-pendulum/rl-feasibility/README.md +0 -60
  15. package/assets/neural-pendulum/rl-feasibility/RESULTS.md +0 -370
  16. package/assets/neural-pendulum/rl-feasibility/evaluate.py +0 -107
  17. package/assets/neural-pendulum/rl-feasibility/finalize_run.py +0 -168
  18. package/assets/neural-pendulum/rl-feasibility/models.py +0 -124
  19. package/assets/neural-pendulum/rl-feasibility/obs.py +0 -205
  20. package/assets/neural-pendulum/rl-feasibility/plot_selected.py +0 -55
  21. package/assets/neural-pendulum/rl-feasibility/reference_controller.py +0 -112
  22. package/assets/neural-pendulum/rl-feasibility/results/curves_selected.png +0 -0
  23. package/assets/neural-pendulum/rl-feasibility/results/reference_controller.json +0 -417
  24. package/assets/neural-pendulum/rl-feasibility/results/summary.json +0 -1599
  25. package/assets/neural-pendulum/rl-feasibility/results/table.md +0 -39
  26. package/assets/neural-pendulum/rl-feasibility/sim.py +0 -324
  27. package/assets/neural-pendulum/rl-feasibility/summarize.py +0 -175
  28. package/assets/neural-pendulum/rl-feasibility/task.py +0 -342
  29. package/assets/neural-pendulum/rl-feasibility/test_contract.py +0 -188
  30. package/assets/neural-pendulum/rl-feasibility/train_ppo.py +0 -403
  31. package/assets/neural-pendulum/rl-feasibility/train_sac.py +0 -406
package/assets/neural-pendulum/rl-feasibility/results/summary.json DELETED
@@ -1,1599 +0,0 @@
1
- {
2
- "n_runs": 15,
3
- "runs": [
4
- {
5
- "exp": "ppo_angle_base_rand",
6
- "algo": "ppo",
7
- "init": "rand",
8
- "obs": "angle",
9
- "obs_dim": 24,
10
- "reward": "base",
11
- "envs": 64,
12
- "global_steps": 31756800,
13
- "wall_min": 40.56691735982895,
14
- "steps_per_s": 13047.084532090059,
15
- "final_t_can": 0.040000000000000036,
16
- "best_t_can": 0.9699999999999793,
17
- "final_success": 0.0,
18
- "final_ever_up": 1.0,
19
- "final_return": 2697.0237712207436,
20
- "best_return": 1032.1841924034118,
21
- "final_ang_err": 0.9750630243194713,
22
- "first_success_steps": null,
23
- "robust_success": 0.0,
24
- "robust_mean_t_can": 0.06499999999999995,
25
- "robust_best_success": 0.0,
26
- "robust_best_mean_t_can": 0.24374999999999625,
27
- "best_ckpt": "ppo_angle_base_rand_best.pt",
28
- "final_ckpt": "ppo_angle_base_rand_final.pt",
29
- "tb_tags": 62,
30
- "tb_missing": []
31
- },
32
- {
33
- "exp": "ppo_angle_h1_rand",
34
- "algo": "ppo",
35
- "init": "rand",
36
- "obs": "angle_h1",
37
- "obs_dim": 6,
38
- "reward": "base",
39
- "envs": 64,
40
- "global_steps": 20966400,
41
- "wall_min": 30.40433213710785,
42
- "steps_per_s": 11493.09902365906,
43
- "final_t_can": 0.0,
44
- "best_t_can": 0.0,
45
- "final_success": 0.0,
46
- "final_ever_up": 0.0,
47
- "final_return": -3526.6859360849385,
48
- "best_return": -2076.88790994392,
49
- "final_ang_err": 2.9346269115974835,
50
- "first_success_steps": null,
51
- "robust_success": 0.0,
52
- "robust_mean_t_can": 0.0,
53
- "robust_best_success": 0.0,
54
- "robust_best_mean_t_can": 0.0,
55
- "best_ckpt": "ppo_angle_h1_rand_best.pt",
56
- "final_ckpt": "ppo_angle_h1_rand_final.pt",
57
- "tb_tags": 62,
58
- "tb_missing": []
59
- },
60
- {
61
- "exp": "ppo_angle_h8_rand",
62
- "algo": "ppo",
63
- "init": "rand",
64
- "obs": "angle_h8",
65
- "obs_dim": 48,
66
- "reward": "base",
67
- "envs": 64,
68
- "global_steps": 30105600,
69
- "wall_min": 40.408548669020334,
70
- "steps_per_s": 12417.174497154854,
71
- "final_t_can": 0.049999999999998934,
72
- "best_t_can": 1.3099999999999787,
73
- "final_success": 0.0,
74
- "final_ever_up": 1.0,
75
- "final_return": 2629.48041920133,
76
- "best_return": 3031.972129344827,
77
- "final_ang_err": 0.8636340246767595,
78
- "first_success_steps": null,
79
- "robust_success": 0.0,
80
- "robust_mean_t_can": 0.05000000000000365,
81
- "robust_best_success": 0.0,
82
- "robust_best_mean_t_can": 0.1512500000000112,
83
- "best_ckpt": "ppo_angle_h8_rand_best.pt",
84
- "final_ckpt": "ppo_angle_h8_rand_final.pt",
85
- "tb_tags": 62,
86
- "tb_missing": []
87
- },
88
- {
89
- "exp": "ppo_angle_mix",
90
- "algo": "ppo",
91
- "init": "mix",
92
- "obs": "angle",
93
- "obs_dim": 24,
94
- "reward": "base",
95
- "envs": 64,
96
- "global_steps": 23500800,
97
- "wall_min": 30.489037291208902,
98
- "steps_per_s": 12846.584700558438,
99
- "final_t_can": 0.0,
100
- "best_t_can": 0.2500000000000002,
101
- "final_success": 0.0,
102
- "final_ever_up": 1.0,
103
- "final_return": -13092.395823817951,
104
- "best_return": 3938.495077872893,
105
- "final_ang_err": 2.8925450260300742,
106
- "first_success_steps": null,
107
- "robust_success": 0.0,
108
- "robust_mean_t_can": 0.0,
109
- "robust_best_success": 0.0,
110
- "robust_best_mean_t_can": 0.10125000000000683,
111
- "best_ckpt": "ppo_angle_mix_best.pt",
112
- "final_ckpt": "ppo_angle_mix_final.pt",
113
- "tb_tags": 62,
114
- "tb_missing": []
115
- },
116
- {
117
- "exp": "ppo_full_base_fixed",
118
- "algo": "ppo",
119
- "init": "fixed",
120
- "obs": "full",
121
- "obs_dim": 36,
122
- "reward": "base",
123
- "envs": 64,
124
- "global_steps": 21926400,
125
- "wall_min": 30.510263828436532,
126
- "steps_per_s": 11977.608651794035,
127
- "final_t_can": 0.0,
128
- "best_t_can": 1.189999999999975,
129
- "final_success": 0.0,
130
- "final_ever_up": 0.0,
131
- "final_return": 260.6949070506481,
132
- "best_return": 7014.763525996907,
133
- "final_ang_err": 0.7154439202317739,
134
- "first_success_steps": null,
135
- "robust_success": 0.0,
136
- "robust_mean_t_can": 0.0,
137
- "robust_best_success": 0.0,
138
- "robust_best_mean_t_can": 1.0337500000000097,
139
- "best_ckpt": "ppo_full_base_fixed_best.pt",
140
- "final_ckpt": "ppo_full_base_fixed_final.pt",
141
- "tb_tags": 62,
142
- "tb_missing": []
143
- },
144
- {
145
- "exp": "ppo_full_base_rand",
146
- "algo": "ppo",
147
- "init": "rand",
148
- "obs": "full",
149
- "obs_dim": 36,
150
- "reward": "base",
151
- "envs": 64,
152
- "global_steps": 35980800,
153
- "wall_min": 45.064537048339844,
154
- "steps_per_s": 13307.137702462915,
155
- "final_t_can": 26.610000000001918,
156
- "best_t_can": 26.850000000001913,
157
- "final_success": 1.0,
158
- "final_ever_up": 1.0,
159
- "final_return": 16772.52675626403,
160
- "best_return": 16214.52153535605,
161
- "final_ang_err": 0.021967002704261773,
162
- "first_success_steps": 11330000,
163
- "robust_success": 1.0,
164
- "robust_mean_t_can": 25.56500000000194,
165
- "robust_best_success": 1.0,
166
- "robust_best_mean_t_can": 26.862500000001912,
167
- "best_ckpt": "ppo_full_base_rand_best.pt",
168
- "final_ckpt": "ppo_full_base_rand_final.pt",
169
- "tb_tags": 53,
170
- "tb_missing": []
171
- },
172
- {
173
- "exp": "ppo_full_base_rand_s1",
174
- "algo": "ppo",
175
- "init": "rand",
176
- "obs": "full",
177
- "obs_dim": 36,
178
- "reward": "base",
179
- "envs": 64,
180
- "global_steps": 29030400,
181
- "wall_min": 40.753157977263136,
182
- "steps_per_s": 11872.454160974283,
183
- "final_t_can": 0.23000000000003595,
184
- "best_t_can": 27.170000000001906,
185
- "final_success": 0.0,
186
- "final_ever_up": 1.0,
187
- "final_return": 5034.6407313699965,
188
- "best_return": 17260.29960622796,
189
- "final_ang_err": 0.2765316418028725,
190
- "first_success_steps": 13632000,
191
- "robust_success": 0.0,
192
- "robust_mean_t_can": 0.20875000000002694,
193
- "robust_best_success": 1.0,
194
- "robust_best_mean_t_can": 26.776250000001916,
195
- "best_ckpt": "ppo_full_base_rand_s1_best.pt",
196
- "final_ckpt": "ppo_full_base_rand_s1_final.pt",
197
- "tb_tags": 62,
198
- "tb_missing": []
199
- },
200
- {
201
- "exp": "ppo_full_base_rand_s2",
202
- "algo": "ppo",
203
- "init": "rand",
204
- "obs": "full",
205
- "obs_dim": 36,
206
- "reward": "base",
207
- "envs": 64,
208
- "global_steps": 34214400,
209
- "wall_min": 40.342886273066206,
210
- "steps_per_s": 14134.833986350272,
211
- "final_t_can": 0.0,
212
- "best_t_can": 28.460000000001887,
213
- "final_success": 0.0,
214
- "final_ever_up": 0.0,
215
- "final_return": 4804.403071234418,
216
- "best_return": 17905.68045466417,
217
- "final_ang_err": 0.08200567245533463,
218
- "first_success_steps": 14400000,
219
- "robust_success": 0.0,
220
- "robust_mean_t_can": 0.0,
221
- "robust_best_success": 1.0,
222
- "robust_best_mean_t_can": 28.031250000001897,
223
- "best_ckpt": "ppo_full_base_rand_s2_best.pt",
224
- "final_ckpt": "ppo_full_base_rand_s2_final.pt",
225
- "tb_tags": 62,
226
- "tb_missing": []
227
- },
228
- {
229
- "exp": "ppo_full_energy_rand",
230
- "algo": "ppo",
231
- "init": "rand",
232
- "obs": "full",
233
- "obs_dim": 36,
234
- "reward": "energy",
235
- "envs": 64,
236
- "global_steps": 24499200,
237
- "wall_min": 35.49630283117294,
238
- "steps_per_s": 11503.169835519106,
239
- "final_t_can": 0.0,
240
- "best_t_can": 0.0,
241
- "final_success": 0.0,
242
- "final_ever_up": 0.0,
243
- "final_return": -1948.337742957759,
244
- "best_return": -1871.980578962857,
245
- "final_ang_err": 2.81246009200542,
246
- "first_success_steps": null,
247
- "robust_success": 0.0,
248
- "robust_mean_t_can": 0.0,
249
- "robust_best_success": 0.0,
250
- "robust_best_mean_t_can": 0.0,
251
- "best_ckpt": "ppo_full_energy_rand_best.pt",
252
- "final_ckpt": "ppo_full_energy_rand_final.pt",
253
- "tb_tags": 62,
254
- "tb_missing": []
255
- },
256
- {
257
- "exp": "ppo_full_h1_mix",
258
- "algo": "ppo",
259
- "init": "mix",
260
- "obs": "full_h1",
261
- "obs_dim": 9,
262
- "reward": "base",
263
- "envs": 64,
264
- "global_steps": 21388800,
265
- "wall_min": 30.461654384930927,
266
- "steps_per_s": 11702.58172768013,
267
- "final_t_can": 0.0,
268
- "best_t_can": 0.020000000000003126,
269
- "final_success": 0.0,
270
- "final_ever_up": 0.0,
271
- "final_return": -552.58611006131,
272
- "best_return": -1191.360424535703,
273
- "final_ang_err": 2.2632595731556897,
274
- "first_success_steps": null,
275
- "robust_success": 0.0,
276
- "robust_mean_t_can": 0.0,
277
- "robust_best_success": 0.0,
278
- "robust_best_mean_t_can": 0.0,
279
- "best_ckpt": "ppo_full_h1_mix_best.pt",
280
- "final_ckpt": "ppo_full_h1_mix_final.pt",
281
- "tb_tags": 62,
282
- "tb_missing": []
283
- },
284
- {
285
- "exp": "ppo_full_h1_rand",
286
- "algo": "ppo",
287
- "init": "rand",
288
- "obs": "full_h1",
289
- "obs_dim": 9,
290
- "reward": "base",
291
- "envs": 64,
292
- "global_steps": 25420800,
293
- "wall_min": 35.34266792138418,
294
- "steps_per_s": 11987.776388087876,
295
- "final_t_can": 0.0,
296
- "best_t_can": 0.06000000000000938,
297
- "final_success": 0.0,
298
- "final_ever_up": 0.0,
299
- "final_return": -1701.8825200860647,
300
- "best_return": 1787.3970086135619,
301
- "final_ang_err": 2.6674287227959432,
302
- "first_success_steps": null,
303
- "robust_success": 0.0,
304
- "robust_mean_t_can": 0.0,
305
- "robust_best_success": 0.0,
306
- "robust_best_mean_t_can": 0.061250000000008215,
307
- "best_ckpt": "ppo_full_h1_rand_best.pt",
308
- "final_ckpt": "ppo_full_h1_rand_final.pt",
309
- "tb_tags": 62,
310
- "tb_missing": []
311
- },
312
- {
313
- "exp": "ppo_full_sparse_rand",
314
- "algo": "ppo",
315
- "init": "rand",
316
- "obs": "full",
317
- "obs_dim": 36,
318
- "reward": "sparse",
319
- "envs": 64,
320
- "global_steps": 19123200,
321
- "wall_min": 30.514096240202587,
322
- "steps_per_s": 10445.008677008878,
323
- "final_t_can": 0.0,
324
- "best_t_can": 0.0,
325
- "final_success": 0.0,
326
- "final_ever_up": 0.0,
327
- "final_return": -0.39284150736598517,
328
- "best_return": -0.015385193831112047,
329
- "final_ang_err": 3.102104613352429,
330
- "first_success_steps": null,
331
- "robust_success": 0.0,
332
- "robust_mean_t_can": 0.0,
333
- "robust_best_success": 0.0,
334
- "robust_best_mean_t_can": 0.0,
335
- "best_ckpt": "ppo_full_sparse_rand_best.pt",
336
- "final_ckpt": "ppo_full_sparse_rand_final.pt",
337
- "tb_tags": 61,
338
- "tb_missing": []
339
- },
340
- {
341
- "exp": "sac_angle_base_rand",
342
- "algo": "sac",
343
- "init": "rand",
344
- "obs": "angle",
345
- "obs_dim": 24,
346
- "reward": "base",
347
- "envs": 32,
348
- "global_steps": 1423296,
349
- "wall_min": 50.590588768323265,
350
- "steps_per_s": 468.893534895032,
351
- "final_t_can": 0.0,
352
- "best_t_can": 0.03999999999999915,
353
- "final_success": 0.0,
354
- "final_ever_up": 0.0,
355
- "final_return": -581.3757746325291,
356
- "best_return": -601.8944783462481,
357
- "final_ang_err": 0.612447697828653,
358
- "first_success_steps": null,
359
- "robust_success": 0.0,
360
- "robust_mean_t_can": 0.009999999999999787,
361
- "robust_best_success": 0.0,
362
- "robust_best_mean_t_can": 0.0,
363
- "best_ckpt": "sac_angle_base_rand_best.pt",
364
- "final_ckpt": "sac_angle_base_rand_final.pt",
365
- "tb_tags": 66,
366
- "tb_missing": []
367
- },
368
- {
369
- "exp": "sac_full_base_fixed",
370
- "algo": "sac",
371
- "init": "fixed",
372
- "obs": "full",
373
- "obs_dim": 36,
374
- "reward": "base",
375
- "envs": 32,
376
- "global_steps": 1407936,
377
- "wall_min": 50.38465542793274,
378
- "steps_per_s": 465.7290955093227,
379
- "final_t_can": 27.620000000001898,
380
- "best_t_can": 28.86000000000189,
381
- "final_success": 1.0,
382
- "final_ever_up": 1.0,
383
- "final_return": 17697.113760150296,
384
- "best_return": 17714.29070651932,
385
- "final_ang_err": 0.03242419670103658,
386
- "first_success_steps": 520000,
387
- "robust_success": 1.0,
388
- "robust_mean_t_can": 27.485000000001904,
389
- "robust_best_success": 1.0,
390
- "robust_best_mean_t_can": 23.44500000000199,
391
- "best_ckpt": "sac_full_base_fixed_best.pt",
392
- "final_ckpt": "sac_full_base_fixed_final.pt",
393
- "tb_tags": 66,
394
- "tb_missing": []
395
- },
396
- {
397
- "exp": "sac_full_base_rand",
398
- "algo": "sac",
399
- "init": "rand",
400
- "obs": "full",
401
- "obs_dim": 36,
402
- "reward": "base",
403
- "envs": 32,
404
- "global_steps": 1424320,
405
- "wall_min": 50.55221416552862,
406
- "steps_per_s": 469.5870805764623,
407
- "final_t_can": 28.210000000001887,
408
- "best_t_can": 28.75000000000189,
409
- "final_success": 1.0,
410
- "final_ever_up": 1.0,
411
- "final_return": 17607.226018245365,
412
- "best_return": 17556.411753349887,
413
- "final_ang_err": 0.02929141220558762,
414
- "first_success_steps": 700000,
415
- "robust_success": 1.0,
416
- "robust_mean_t_can": 26.390000000001926,
417
- "robust_best_success": 0.75,
418
- "robust_best_mean_t_can": 20.886250000001425,
419
- "best_ckpt": "sac_full_base_rand_best.pt",
420
- "final_ckpt": "sac_full_base_rand_final.pt",
421
- "tb_tags": 66,
422
- "tb_missing": []
423
- }
424
- ],
425
- "tb_coverage": [
426
- {
427
- "exp": "ppo_angle_base_rand",
428
- "exists": true,
429
- "tags": [
430
- "eval/action_std",
431
- "eval/ep_return",
432
- "eval/ever_up",
433
- "eval/final_ang_err",
434
- "eval/first_up_t",
435
- "eval/mean_abs_action",
436
- "eval/mean_u1",
437
- "eval/mean_u3",
438
- "eval/steps",
439
- "eval/success",
440
- "eval/t_can",
441
- "eval/up_frac",
442
- "eval_final/action_std",
443
- "eval_final/ep_return",
444
- "eval_final/ever_up",
445
- "eval_final/final_ang_err",
446
- "eval_final/first_up_t",
447
- "eval_final/mean_abs_action",
448
- "eval_final/mean_u1",
449
- "eval_final/mean_u3",
450
- "eval_final/steps",
451
- "eval_final/success",
452
- "eval_final/t_can",
453
- "eval_final/up_frac",
454
- "loss/entropy",
455
- "loss/policy",
456
- "loss/value",
457
- "robust/ever_up_rate",
458
- "robust/max_t_can",
459
- "robust/mean_final_ang_err",
460
- "robust/mean_return",
461
- "robust/mean_t_can",
462
- "robust/median_t_can",
463
- "robust/min_t_can",
464
- "robust/n",
465
- "robust/success_rate",
466
- "robust_best/ever_up_rate",
467
- "robust_best/max_t_can",
468
- "robust_best/mean_final_ang_err",
469
- "robust_best/mean_return",
470
- "robust_best/mean_t_can",
471
- "robust_best/median_t_can",
472
- "robust_best/min_t_can",
473
- "robust_best/n",
474
- "robust_best/success_rate",
475
- "time/elapsed_min",
476
- "train/action_abs_mean",
477
- "train/action_std",
478
- "train/approx_kl",
479
- "train/clip_frac",
480
- "train/episode_length",
481
- "train/episode_return",
482
- "train/ever_up_frac",
483
- "train/first_up_time",
484
- "train/global_steps",
485
- "train/lr",
486
- "train/mean_step_reward",
487
- "train/std",
488
- "train/success_rate",
489
- "train/t_can",
490
- "train/t_can_max",
491
- "train/up_frac"
492
- ],
493
- "missing": [],
494
- "n_points": 62,
495
- "last": {
496
- "train/global_steps": 31756800.0,
497
- "time/elapsed_min": 40.041133880615234,
498
- "train/t_can": 0.027656808495521545,
499
- "train/success_rate": 0.0,
500
- "eval/t_can": 0.0,
501
- "eval/success": 0.0
502
- }
503
- },
504
- {
505
- "exp": "ppo_angle_h1_rand",
506
- "exists": true,
507
- "tags": [
508
- "eval/action_std",
509
- "eval/ep_return",
510
- "eval/ever_up",
511
- "eval/final_ang_err",
512
- "eval/first_up_t",
513
- "eval/mean_abs_action",
514
- "eval/mean_u1",
515
- "eval/mean_u3",
516
- "eval/steps",
517
- "eval/success",
518
- "eval/t_can",
519
- "eval/up_frac",
520
- "eval_final/action_std",
521
- "eval_final/ep_return",
522
- "eval_final/ever_up",
523
- "eval_final/final_ang_err",
524
- "eval_final/first_up_t",
525
- "eval_final/mean_abs_action",
526
- "eval_final/mean_u1",
527
- "eval_final/mean_u3",
528
- "eval_final/steps",
529
- "eval_final/success",
530
- "eval_final/t_can",
531
- "eval_final/up_frac",
532
- "loss/entropy",
533
- "loss/policy",
534
- "loss/value",
535
- "robust/ever_up_rate",
536
- "robust/max_t_can",
537
- "robust/mean_final_ang_err",
538
- "robust/mean_return",
539
- "robust/mean_t_can",
540
- "robust/median_t_can",
541
- "robust/min_t_can",
542
- "robust/n",
543
- "robust/success_rate",
544
- "robust_best/ever_up_rate",
545
- "robust_best/max_t_can",
546
- "robust_best/mean_final_ang_err",
547
- "robust_best/mean_return",
548
- "robust_best/mean_t_can",
549
- "robust_best/median_t_can",
550
- "robust_best/min_t_can",
551
- "robust_best/n",
552
- "robust_best/success_rate",
553
- "time/elapsed_min",
554
- "train/action_abs_mean",
555
- "train/action_std",
556
- "train/approx_kl",
557
- "train/clip_frac",
558
- "train/episode_length",
559
- "train/episode_return",
560
- "train/ever_up_frac",
561
- "train/first_up_time",
562
- "train/global_steps",
563
- "train/lr",
564
- "train/mean_step_reward",
565
- "train/std",
566
- "train/success_rate",
567
- "train/t_can",
568
- "train/t_can_max",
569
- "train/up_frac"
570
- ],
571
- "missing": [],
572
- "n_points": 62,
573
- "last": {
574
- "train/global_steps": 20966400.0,
575
- "time/elapsed_min": 30.015865325927734,
576
- "train/t_can": 0.0,
577
- "train/success_rate": 0.0,
578
- "eval/t_can": 0.0,
579
- "eval/success": 0.0
580
- }
581
- },
582
- {
583
- "exp": "ppo_angle_h8_rand",
584
- "exists": true,
585
- "tags": [
586
- "eval/action_std",
587
- "eval/ep_return",
588
- "eval/ever_up",
589
- "eval/final_ang_err",
590
- "eval/first_up_t",
591
- "eval/mean_abs_action",
592
- "eval/mean_u1",
593
- "eval/mean_u3",
594
- "eval/steps",
595
- "eval/success",
596
- "eval/t_can",
597
- "eval/up_frac",
598
- "eval_final/action_std",
599
- "eval_final/ep_return",
600
- "eval_final/ever_up",
601
- "eval_final/final_ang_err",
602
- "eval_final/first_up_t",
603
- "eval_final/mean_abs_action",
604
- "eval_final/mean_u1",
605
- "eval_final/mean_u3",
606
- "eval_final/steps",
607
- "eval_final/success",
608
- "eval_final/t_can",
609
- "eval_final/up_frac",
610
- "loss/entropy",
611
- "loss/policy",
612
- "loss/value",
613
- "robust/ever_up_rate",
614
- "robust/max_t_can",
615
- "robust/mean_final_ang_err",
616
- "robust/mean_return",
617
- "robust/mean_t_can",
618
- "robust/median_t_can",
619
- "robust/min_t_can",
620
- "robust/n",
621
- "robust/success_rate",
622
- "robust_best/ever_up_rate",
623
- "robust_best/max_t_can",
624
- "robust_best/mean_final_ang_err",
625
- "robust_best/mean_return",
626
- "robust_best/mean_t_can",
627
- "robust_best/median_t_can",
628
- "robust_best/min_t_can",
629
- "robust_best/n",
630
- "robust_best/success_rate",
631
- "time/elapsed_min",
632
- "train/action_abs_mean",
633
- "train/action_std",
634
- "train/approx_kl",
635
- "train/clip_frac",
636
- "train/episode_length",
637
- "train/episode_return",
638
- "train/ever_up_frac",
639
- "train/first_up_time",
640
- "train/global_steps",
641
- "train/lr",
642
- "train/mean_step_reward",
643
- "train/std",
644
- "train/success_rate",
645
- "train/t_can",
646
- "train/t_can_max",
647
- "train/up_frac"
648
- ],
649
- "missing": [],
650
- "n_points": 62,
651
- "last": {
652
- "train/global_steps": 30105600.0,
653
- "time/elapsed_min": 40.043148040771484,
654
- "train/t_can": 0.061251282691955566,
655
- "train/success_rate": 0.0,
656
- "eval/t_can": 0.0,
657
- "eval/success": 0.0
658
- }
659
- },
660
- {
661
- "exp": "ppo_angle_mix",
662
- "exists": true,
663
- "tags": [
664
- "eval/action_std",
665
- "eval/ep_return",
666
- "eval/ever_up",
667
- "eval/final_ang_err",
668
- "eval/first_up_t",
669
- "eval/mean_abs_action",
670
- "eval/mean_u1",
671
- "eval/mean_u3",
672
- "eval/steps",
673
- "eval/success",
674
- "eval/t_can",
675
- "eval/up_frac",
676
- "eval_final/action_std",
677
- "eval_final/ep_return",
678
- "eval_final/ever_up",
679
- "eval_final/final_ang_err",
680
- "eval_final/first_up_t",
681
- "eval_final/mean_abs_action",
682
- "eval_final/mean_u1",
683
- "eval_final/mean_u3",
684
- "eval_final/steps",
685
- "eval_final/success",
686
- "eval_final/t_can",
687
- "eval_final/up_frac",
688
- "loss/entropy",
689
- "loss/policy",
690
- "loss/value",
691
- "robust/ever_up_rate",
692
- "robust/max_t_can",
693
- "robust/mean_final_ang_err",
694
- "robust/mean_return",
695
- "robust/mean_t_can",
696
- "robust/median_t_can",
697
- "robust/min_t_can",
698
- "robust/n",
699
- "robust/success_rate",
700
- "robust_best/ever_up_rate",
701
- "robust_best/max_t_can",
702
- "robust_best/mean_final_ang_err",
703
- "robust_best/mean_return",
704
- "robust_best/mean_t_can",
705
- "robust_best/median_t_can",
706
- "robust_best/min_t_can",
707
- "robust_best/n",
708
- "robust_best/success_rate",
709
- "time/elapsed_min",
710
- "train/action_abs_mean",
711
- "train/action_std",
712
- "train/approx_kl",
713
- "train/clip_frac",
714
- "train/episode_length",
715
- "train/episode_return",
716
- "train/ever_up_frac",
717
- "train/first_up_time",
718
- "train/global_steps",
719
- "train/lr",
720
- "train/mean_step_reward",
721
- "train/std",
722
- "train/success_rate",
723
- "train/t_can",
724
- "train/t_can_max",
725
- "train/up_frac"
726
- ],
727
- "missing": [],
728
- "n_points": 62,
729
- "last": {
730
- "train/global_steps": 23500800.0,
731
- "time/elapsed_min": 30.01533317565918,
732
- "train/t_can": 0.022656245157122612,
733
- "train/success_rate": 0.0,
734
- "eval/t_can": 0.009999999776482582,
735
- "eval/success": 0.0
736
- }
737
- },
738
- {
739
- "exp": "ppo_full_base_fixed",
740
- "exists": true,
741
- "tags": [
742
- "eval/action_std",
743
- "eval/ep_return",
744
- "eval/ever_up",
745
- "eval/final_ang_err",
746
- "eval/first_up_t",
747
- "eval/mean_abs_action",
748
- "eval/mean_u1",
749
- "eval/mean_u3",
750
- "eval/steps",
751
- "eval/success",
752
- "eval/t_can",
753
- "eval/up_frac",
754
- "eval_final/action_std",
755
- "eval_final/ep_return",
756
- "eval_final/ever_up",
757
- "eval_final/final_ang_err",
758
- "eval_final/first_up_t",
759
- "eval_final/mean_abs_action",
760
- "eval_final/mean_u1",
761
- "eval_final/mean_u3",
762
- "eval_final/steps",
763
- "eval_final/success",
764
- "eval_final/t_can",
765
- "eval_final/up_frac",
766
- "loss/entropy",
767
- "loss/policy",
768
- "loss/value",
769
- "robust/ever_up_rate",
770
- "robust/max_t_can",
771
- "robust/mean_final_ang_err",
772
- "robust/mean_return",
773
- "robust/mean_t_can",
774
- "robust/median_t_can",
775
- "robust/min_t_can",
776
- "robust/n",
777
- "robust/success_rate",
778
- "robust_best/ever_up_rate",
779
- "robust_best/max_t_can",
780
- "robust_best/mean_final_ang_err",
781
- "robust_best/mean_return",
782
- "robust_best/mean_t_can",
783
- "robust_best/median_t_can",
784
- "robust_best/min_t_can",
785
- "robust_best/n",
786
- "robust_best/success_rate",
787
- "time/elapsed_min",
788
- "train/action_abs_mean",
789
- "train/action_std",
790
- "train/approx_kl",
791
- "train/clip_frac",
792
- "train/episode_length",
793
- "train/episode_return",
794
- "train/ever_up_frac",
795
- "train/first_up_time",
796
- "train/global_steps",
797
- "train/lr",
798
- "train/mean_step_reward",
799
- "train/std",
800
- "train/success_rate",
801
- "train/t_can",
802
- "train/t_can_max",
803
- "train/up_frac"
804
- ],
805
- "missing": [],
806
- "n_points": 62,
807
- "last": {
808
- "train/global_steps": 21926400.0,
809
- "time/elapsed_min": 30.01462173461914,
810
- "train/t_can": 0.0,
811
- "train/success_rate": 0.0,
812
- "eval/t_can": 0.0,
813
- "eval/success": 0.0
814
- }
815
- },
816
- {
817
- "exp": "ppo_full_base_rand",
818
- "exists": true,
819
- "tags": [
820
- "eval/action_std",
821
- "eval/ep_return",
822
- "eval/ever_up",
823
- "eval/final_ang_err",
824
- "eval/first_up_t",
825
- "eval/mean_abs_action",
826
- "eval/mean_u1",
827
- "eval/mean_u3",
828
- "eval/steps",
829
- "eval/success",
830
- "eval/t_can",
831
- "eval/up_frac",
832
- "eval_final/action_std",
833
- "eval_final/ep_return",
834
- "eval_final/ever_up",
835
- "eval_final/final_ang_err",
836
- "eval_final/first_up_t",
837
- "eval_final/mean_abs_action",
838
- "eval_final/mean_u1",
839
- "eval_final/mean_u3",
840
- "eval_final/steps",
841
- "eval_final/success",
842
- "eval_final/t_can",
843
- "eval_final/up_frac",
844
- "loss/entropy",
845
- "loss/policy",
846
- "loss/value",
847
- "robust/ever_up_rate",
848
- "robust/max_t_can",
849
- "robust/mean_final_ang_err",
850
- "robust/mean_return",
851
- "robust/mean_t_can",
852
- "robust/median_t_can",
853
- "robust/min_t_can",
854
- "robust/n",
855
- "robust/success_rate",
856
- "time/elapsed_min",
857
- "train/action_abs_mean",
858
- "train/action_std",
859
- "train/approx_kl",
860
- "train/clip_frac",
861
- "train/episode_length",
862
- "train/episode_return",
863
- "train/ever_up_frac",
864
- "train/first_up_time",
865
- "train/global_steps",
866
- "train/lr",
867
- "train/mean_step_reward",
868
- "train/std",
869
- "train/success_rate",
870
- "train/t_can",
871
- "train/t_can_max",
872
- "train/up_frac"
873
- ],
874
- "missing": [],
875
- "n_points": 53,
876
- "last": {
877
- "train/global_steps": 35980800.0,
878
- "time/elapsed_min": 45.064537048339844,
879
- "train/t_can": 19.041841506958008,
880
- "train/success_rate": 0.984375,
881
- "eval/t_can": 26.3700008392334,
882
- "eval/success": 1.0
883
- }
884
- },
885
- {
886
- "exp": "ppo_full_base_rand_s1",
887
- "exists": true,
888
- "tags": [
889
- "eval/action_std",
890
- "eval/ep_return",
891
- "eval/ever_up",
892
- "eval/final_ang_err",
893
- "eval/first_up_t",
894
- "eval/mean_abs_action",
895
- "eval/mean_u1",
896
- "eval/mean_u3",
897
- "eval/steps",
898
- "eval/success",
899
- "eval/t_can",
900
- "eval/up_frac",
901
- "eval_final/action_std",
902
- "eval_final/ep_return",
903
- "eval_final/ever_up",
904
- "eval_final/final_ang_err",
905
- "eval_final/first_up_t",
906
- "eval_final/mean_abs_action",
907
- "eval_final/mean_u1",
908
- "eval_final/mean_u3",
909
- "eval_final/steps",
910
- "eval_final/success",
911
- "eval_final/t_can",
912
- "eval_final/up_frac",
913
- "loss/entropy",
914
- "loss/policy",
915
- "loss/value",
916
- "robust/ever_up_rate",
917
- "robust/max_t_can",
918
- "robust/mean_final_ang_err",
919
- "robust/mean_return",
920
- "robust/mean_t_can",
921
- "robust/median_t_can",
922
- "robust/min_t_can",
923
- "robust/n",
924
- "robust/success_rate",
925
- "robust_best/ever_up_rate",
926
- "robust_best/max_t_can",
927
- "robust_best/mean_final_ang_err",
928
- "robust_best/mean_return",
929
- "robust_best/mean_t_can",
930
- "robust_best/median_t_can",
931
- "robust_best/min_t_can",
932
- "robust_best/n",
933
- "robust_best/success_rate",
934
- "time/elapsed_min",
935
- "train/action_abs_mean",
936
- "train/action_std",
937
- "train/approx_kl",
938
- "train/clip_frac",
939
- "train/episode_length",
940
- "train/episode_return",
941
- "train/ever_up_frac",
942
- "train/first_up_time",
943
- "train/global_steps",
944
- "train/lr",
945
- "train/mean_step_reward",
946
- "train/std",
947
- "train/success_rate",
948
- "train/t_can",
949
- "train/t_can_max",
950
- "train/up_frac"
951
- ],
952
- "missing": [],
953
- "n_points": 62,
954
- "last": {
955
- "train/global_steps": 29030400.0,
956
- "time/elapsed_min": 40.0202751159668,
957
- "train/t_can": 0.26484936475753784,
958
- "train/success_rate": 0.0,
959
- "eval/t_can": 0.0,
960
- "eval/success": 0.0
961
- }
962
- },
963
- {
964
- "exp": "ppo_full_base_rand_s2",
965
- "exists": true,
966
- "tags": [
967
- "eval/action_std",
968
- "eval/ep_return",
969
- "eval/ever_up",
970
- "eval/final_ang_err",
971
- "eval/first_up_t",
972
- "eval/mean_abs_action",
973
- "eval/mean_u1",
974
- "eval/mean_u3",
975
- "eval/steps",
976
- "eval/success",
977
- "eval/t_can",
978
- "eval/up_frac",
979
- "eval_final/action_std",
980
- "eval_final/ep_return",
981
- "eval_final/ever_up",
982
- "eval_final/final_ang_err",
983
- "eval_final/first_up_t",
984
- "eval_final/mean_abs_action",
985
- "eval_final/mean_u1",
986
- "eval_final/mean_u3",
987
- "eval_final/steps",
988
- "eval_final/success",
989
- "eval_final/t_can",
990
- "eval_final/up_frac",
991
- "loss/entropy",
992
- "loss/policy",
993
- "loss/value",
994
- "robust/ever_up_rate",
995
- "robust/max_t_can",
996
- "robust/mean_final_ang_err",
997
- "robust/mean_return",
998
- "robust/mean_t_can",
999
- "robust/median_t_can",
1000
- "robust/min_t_can",
1001
- "robust/n",
1002
- "robust/success_rate",
1003
- "robust_best/ever_up_rate",
1004
- "robust_best/max_t_can",
1005
- "robust_best/mean_final_ang_err",
1006
- "robust_best/mean_return",
1007
- "robust_best/mean_t_can",
1008
- "robust_best/median_t_can",
1009
- "robust_best/min_t_can",
1010
- "robust_best/n",
1011
- "robust_best/success_rate",
1012
- "time/elapsed_min",
1013
- "train/action_abs_mean",
1014
- "train/action_std",
1015
- "train/approx_kl",
1016
- "train/clip_frac",
1017
- "train/episode_length",
1018
- "train/episode_return",
1019
- "train/ever_up_frac",
1020
- "train/first_up_time",
1021
- "train/global_steps",
1022
- "train/lr",
1023
- "train/mean_step_reward",
1024
- "train/std",
1025
- "train/success_rate",
1026
- "train/t_can",
1027
- "train/t_can_max",
1028
- "train/up_frac"
1029
- ],
1030
- "missing": [],
1031
- "n_points": 62,
1032
- "last": {
1033
- "train/global_steps": 34214400.0,
1034
- "time/elapsed_min": 40.01681137084961,
1035
- "train/t_can": 0.0,
1036
- "train/success_rate": 0.0,
1037
- "eval/t_can": 0.0,
1038
- "eval/success": 0.0
1039
- }
1040
- },
1041
- {
1042
- "exp": "ppo_full_energy_rand",
1043
- "exists": true,
1044
- "tags": [
1045
- "eval/action_std",
1046
- "eval/ep_return",
1047
- "eval/ever_up",
1048
- "eval/final_ang_err",
1049
- "eval/first_up_t",
1050
- "eval/mean_abs_action",
1051
- "eval/mean_u1",
1052
- "eval/mean_u3",
1053
- "eval/steps",
1054
- "eval/success",
1055
- "eval/t_can",
1056
- "eval/up_frac",
1057
- "eval_final/action_std",
1058
- "eval_final/ep_return",
1059
- "eval_final/ever_up",
1060
- "eval_final/final_ang_err",
1061
- "eval_final/first_up_t",
1062
- "eval_final/mean_abs_action",
1063
- "eval_final/mean_u1",
1064
- "eval_final/mean_u3",
1065
- "eval_final/steps",
1066
- "eval_final/success",
1067
- "eval_final/t_can",
1068
- "eval_final/up_frac",
1069
- "loss/entropy",
1070
- "loss/policy",
1071
- "loss/value",
1072
- "robust/ever_up_rate",
1073
- "robust/max_t_can",
1074
- "robust/mean_final_ang_err",
1075
- "robust/mean_return",
1076
- "robust/mean_t_can",
1077
- "robust/median_t_can",
1078
- "robust/min_t_can",
1079
- "robust/n",
1080
- "robust/success_rate",
1081
- "robust_best/ever_up_rate",
1082
- "robust_best/max_t_can",
1083
- "robust_best/mean_final_ang_err",
1084
- "robust_best/mean_return",
1085
- "robust_best/mean_t_can",
1086
- "robust_best/median_t_can",
1087
- "robust_best/min_t_can",
1088
- "robust_best/n",
1089
- "robust_best/success_rate",
1090
- "time/elapsed_min",
1091
- "train/action_abs_mean",
1092
- "train/action_std",
1093
- "train/approx_kl",
1094
- "train/clip_frac",
1095
- "train/episode_length",
1096
- "train/episode_return",
1097
- "train/ever_up_frac",
1098
- "train/first_up_time",
1099
- "train/global_steps",
1100
- "train/lr",
1101
- "train/mean_step_reward",
1102
- "train/std",
1103
- "train/success_rate",
1104
- "train/t_can",
1105
- "train/t_can_max",
1106
- "train/up_frac"
1107
- ],
1108
- "missing": [],
1109
- "n_points": 62,
1110
- "last": {
1111
- "train/global_steps": 24499200.0,
1112
- "time/elapsed_min": 35.0157356262207,
1113
- "train/t_can": 0.0,
1114
- "train/success_rate": 0.0,
1115
- "eval/t_can": 0.0,
1116
- "eval/success": 0.0
1117
- }
1118
- },
1119
- {
1120
- "exp": "ppo_full_h1_mix",
1121
- "exists": true,
1122
- "tags": [
1123
- "eval/action_std",
1124
- "eval/ep_return",
1125
- "eval/ever_up",
1126
- "eval/final_ang_err",
1127
- "eval/first_up_t",
1128
- "eval/mean_abs_action",
1129
- "eval/mean_u1",
1130
- "eval/mean_u3",
1131
- "eval/steps",
1132
- "eval/success",
1133
- "eval/t_can",
1134
- "eval/up_frac",
1135
- "eval_final/action_std",
1136
- "eval_final/ep_return",
1137
- "eval_final/ever_up",
1138
- "eval_final/final_ang_err",
1139
- "eval_final/first_up_t",
1140
- "eval_final/mean_abs_action",
1141
- "eval_final/mean_u1",
1142
- "eval_final/mean_u3",
1143
- "eval_final/steps",
1144
- "eval_final/success",
1145
- "eval_final/t_can",
1146
- "eval_final/up_frac",
1147
- "loss/entropy",
1148
- "loss/policy",
1149
- "loss/value",
1150
- "robust/ever_up_rate",
1151
- "robust/max_t_can",
1152
- "robust/mean_final_ang_err",
1153
- "robust/mean_return",
1154
- "robust/mean_t_can",
1155
- "robust/median_t_can",
1156
- "robust/min_t_can",
1157
- "robust/n",
1158
- "robust/success_rate",
1159
- "robust_best/ever_up_rate",
1160
- "robust_best/max_t_can",
1161
- "robust_best/mean_final_ang_err",
1162
- "robust_best/mean_return",
1163
- "robust_best/mean_t_can",
1164
- "robust_best/median_t_can",
1165
- "robust_best/min_t_can",
1166
- "robust_best/n",
1167
- "robust_best/success_rate",
1168
- "time/elapsed_min",
1169
- "train/action_abs_mean",
1170
- "train/action_std",
1171
- "train/approx_kl",
1172
- "train/clip_frac",
1173
- "train/episode_length",
1174
- "train/episode_return",
1175
- "train/ever_up_frac",
1176
- "train/first_up_time",
1177
- "train/global_steps",
1178
- "train/lr",
1179
- "train/mean_step_reward",
1180
- "train/std",
1181
- "train/success_rate",
1182
- "train/t_can",
1183
- "train/t_can_max",
1184
- "train/up_frac"
1185
- ],
1186
- "missing": [],
1187
- "n_points": 62,
1188
- "last": {
1189
- "train/global_steps": 21388800.0,
1190
- "time/elapsed_min": 30.037633895874023,
1191
- "train/t_can": 0.0017187498742714524,
1192
- "train/success_rate": 0.0,
1193
- "eval/t_can": 0.0,
1194
- "eval/success": 0.0
1195
- }
1196
- },
1197
- {
1198
- "exp": "ppo_full_h1_rand",
1199
- "exists": true,
1200
- "tags": [
1201
- "eval/action_std",
1202
- "eval/ep_return",
1203
- "eval/ever_up",
1204
- "eval/final_ang_err",
1205
- "eval/first_up_t",
1206
- "eval/mean_abs_action",
1207
- "eval/mean_u1",
1208
- "eval/mean_u3",
1209
- "eval/steps",
1210
- "eval/success",
1211
- "eval/t_can",
1212
- "eval/up_frac",
1213
- "eval_final/action_std",
1214
- "eval_final/ep_return",
1215
- "eval_final/ever_up",
1216
- "eval_final/final_ang_err",
1217
- "eval_final/first_up_t",
1218
- "eval_final/mean_abs_action",
1219
- "eval_final/mean_u1",
1220
- "eval_final/mean_u3",
1221
- "eval_final/steps",
1222
- "eval_final/success",
1223
- "eval_final/t_can",
1224
- "eval_final/up_frac",
1225
- "loss/entropy",
1226
- "loss/policy",
1227
- "loss/value",
1228
- "robust/ever_up_rate",
1229
- "robust/max_t_can",
1230
- "robust/mean_final_ang_err",
1231
- "robust/mean_return",
1232
- "robust/mean_t_can",
1233
- "robust/median_t_can",
1234
- "robust/min_t_can",
1235
- "robust/n",
1236
- "robust/success_rate",
1237
- "robust_best/ever_up_rate",
1238
- "robust_best/max_t_can",
1239
- "robust_best/mean_final_ang_err",
1240
- "robust_best/mean_return",
1241
- "robust_best/mean_t_can",
1242
- "robust_best/median_t_can",
1243
- "robust_best/min_t_can",
1244
- "robust_best/n",
1245
- "robust_best/success_rate",
1246
- "time/elapsed_min",
1247
- "train/action_abs_mean",
1248
- "train/action_std",
1249
- "train/approx_kl",
1250
- "train/clip_frac",
1251
- "train/episode_length",
1252
- "train/episode_return",
1253
- "train/ever_up_frac",
1254
- "train/first_up_time",
1255
- "train/global_steps",
1256
- "train/lr",
1257
- "train/mean_step_reward",
1258
- "train/std",
1259
- "train/success_rate",
1260
- "train/t_can",
1261
- "train/t_can_max",
1262
- "train/up_frac"
1263
- ],
1264
- "missing": [],
1265
- "n_points": 62,
1266
- "last": {
1267
- "train/global_steps": 25420800.0,
1268
- "time/elapsed_min": 35.00432586669922,
1269
- "train/t_can": 0.0,
1270
- "train/success_rate": 0.0,
1271
- "eval/t_can": 0.0,
1272
- "eval/success": 0.0
1273
- }
1274
- },
1275
- {
1276
- "exp": "ppo_full_sparse_rand",
1277
- "exists": true,
1278
- "tags": [
1279
- "eval/action_std",
1280
- "eval/ep_return",
1281
- "eval/ever_up",
1282
- "eval/final_ang_err",
1283
- "eval/first_up_t",
1284
- "eval/mean_abs_action",
1285
- "eval/mean_u1",
1286
- "eval/mean_u3",
1287
- "eval/steps",
1288
- "eval/success",
1289
- "eval/t_can",
1290
- "eval/up_frac",
1291
- "eval_final/action_std",
1292
- "eval_final/ep_return",
1293
- "eval_final/ever_up",
1294
- "eval_final/final_ang_err",
1295
- "eval_final/first_up_t",
1296
- "eval_final/mean_abs_action",
1297
- "eval_final/mean_u1",
1298
- "eval_final/mean_u3",
1299
- "eval_final/steps",
1300
- "eval_final/success",
1301
- "eval_final/t_can",
1302
- "eval_final/up_frac",
1303
- "loss/entropy",
1304
- "loss/policy",
1305
- "loss/value",
1306
- "robust/ever_up_rate",
1307
- "robust/max_t_can",
1308
- "robust/mean_final_ang_err",
1309
- "robust/mean_return",
1310
- "robust/mean_t_can",
1311
- "robust/median_t_can",
1312
- "robust/min_t_can",
1313
- "robust/n",
1314
- "robust/success_rate",
1315
- "robust_best/ever_up_rate",
1316
- "robust_best/max_t_can",
1317
- "robust_best/mean_final_ang_err",
1318
- "robust_best/mean_return",
1319
- "robust_best/mean_t_can",
1320
- "robust_best/median_t_can",
1321
- "robust_best/min_t_can",
1322
- "robust_best/n",
1323
- "robust_best/success_rate",
1324
- "time/elapsed_min",
1325
- "train/action_abs_mean",
1326
- "train/action_std",
1327
- "train/approx_kl",
1328
- "train/clip_frac",
1329
- "train/episode_length",
1330
- "train/episode_return",
1331
- "train/ever_up_frac",
1332
- "train/global_steps",
1333
- "train/lr",
1334
- "train/mean_step_reward",
1335
- "train/std",
1336
- "train/success_rate",
1337
- "train/t_can",
1338
- "train/t_can_max",
1339
- "train/up_frac"
1340
- ],
1341
- "missing": [],
1342
- "n_points": 61,
1343
- "last": {
1344
- "train/global_steps": 19123200.0,
1345
- "time/elapsed_min": 30.050100326538086,
1346
- "train/t_can": 0.0,
1347
- "train/success_rate": 0.0,
1348
- "eval/t_can": 0.0,
1349
- "eval/success": 0.0
1350
- }
1351
- },
1352
- {
1353
- "exp": "sac_angle_base_rand",
1354
- "exists": true,
1355
- "tags": [
1356
- "eval/action_std",
1357
- "eval/ep_return",
1358
- "eval/ever_up",
1359
- "eval/final_ang_err",
1360
- "eval/first_up_t",
1361
- "eval/mean_abs_action",
1362
- "eval/mean_u1",
1363
- "eval/mean_u3",
1364
- "eval/steps",
1365
- "eval/success",
1366
- "eval/t_can",
1367
- "eval/up_frac",
1368
- "eval_final/action_std",
1369
- "eval_final/ep_return",
1370
- "eval_final/ever_up",
1371
- "eval_final/final_ang_err",
1372
- "eval_final/first_up_t",
1373
- "eval_final/mean_abs_action",
1374
- "eval_final/mean_u1",
1375
- "eval_final/mean_u3",
1376
- "eval_final/steps",
1377
- "eval_final/success",
1378
- "eval_final/t_can",
1379
- "eval_final/up_frac",
1380
- "loss/actor",
1381
- "loss/alpha",
1382
- "loss/critic",
1383
- "loss/entropy",
1384
- "loss/policy",
1385
- "loss/value",
1386
- "robust/ever_up_rate",
1387
- "robust/max_t_can",
1388
- "robust/mean_final_ang_err",
1389
- "robust/mean_return",
1390
- "robust/mean_t_can",
1391
- "robust/median_t_can",
1392
- "robust/min_t_can",
1393
- "robust/n",
1394
- "robust/success_rate",
1395
- "robust_best/ever_up_rate",
1396
- "robust_best/max_t_can",
1397
- "robust_best/mean_final_ang_err",
1398
- "robust_best/mean_return",
1399
- "robust_best/mean_t_can",
1400
- "robust_best/median_t_can",
1401
- "robust_best/min_t_can",
1402
- "robust_best/n",
1403
- "robust_best/success_rate",
1404
- "time/elapsed_min",
1405
- "train/action_abs_mean",
1406
- "train/action_std",
1407
- "train/alpha",
1408
- "train/entropy",
1409
- "train/episode_length",
1410
- "train/episode_return",
1411
- "train/ever_up_frac",
1412
- "train/first_up_time",
1413
- "train/global_steps",
1414
- "train/q_mean",
1415
- "train/q_target_mean",
1416
- "train/replay_size",
1417
- "train/steps_per_s",
1418
- "train/success_rate",
1419
- "train/t_can",
1420
- "train/t_can_max",
1421
- "train/up_frac"
1422
- ],
1423
- "missing": [],
1424
- "n_points": 66,
1425
- "last": {
1426
- "train/global_steps": 1423040.0,
1427
- "time/elapsed_min": 49.990970611572266,
1428
- "train/t_can": 0.0,
1429
- "train/success_rate": 0.0,
1430
- "eval/t_can": 0.0,
1431
- "eval/success": 0.0
1432
- }
1433
- },
1434
- {
1435
- "exp": "sac_full_base_fixed",
1436
- "exists": true,
1437
- "tags": [
1438
- "eval/action_std",
1439
- "eval/ep_return",
1440
- "eval/ever_up",
1441
- "eval/final_ang_err",
1442
- "eval/first_up_t",
1443
- "eval/mean_abs_action",
1444
- "eval/mean_u1",
1445
- "eval/mean_u3",
1446
- "eval/steps",
1447
- "eval/success",
1448
- "eval/t_can",
1449
- "eval/up_frac",
1450
- "eval_final/action_std",
1451
- "eval_final/ep_return",
1452
- "eval_final/ever_up",
1453
- "eval_final/final_ang_err",
1454
- "eval_final/first_up_t",
1455
- "eval_final/mean_abs_action",
1456
- "eval_final/mean_u1",
1457
- "eval_final/mean_u3",
1458
- "eval_final/steps",
1459
- "eval_final/success",
1460
- "eval_final/t_can",
1461
- "eval_final/up_frac",
1462
- "loss/actor",
1463
- "loss/alpha",
1464
- "loss/critic",
1465
- "loss/entropy",
1466
- "loss/policy",
1467
- "loss/value",
1468
- "robust/ever_up_rate",
1469
- "robust/max_t_can",
1470
- "robust/mean_final_ang_err",
1471
- "robust/mean_return",
1472
- "robust/mean_t_can",
1473
- "robust/median_t_can",
1474
- "robust/min_t_can",
1475
- "robust/n",
1476
- "robust/success_rate",
1477
- "robust_best/ever_up_rate",
1478
- "robust_best/max_t_can",
1479
- "robust_best/mean_final_ang_err",
1480
- "robust_best/mean_return",
1481
- "robust_best/mean_t_can",
1482
- "robust_best/median_t_can",
1483
- "robust_best/min_t_can",
1484
- "robust_best/n",
1485
- "robust_best/success_rate",
1486
- "time/elapsed_min",
1487
- "train/action_abs_mean",
1488
- "train/action_std",
1489
- "train/alpha",
1490
- "train/entropy",
1491
- "train/episode_length",
1492
- "train/episode_return",
1493
- "train/ever_up_frac",
1494
- "train/first_up_time",
1495
- "train/global_steps",
1496
- "train/q_mean",
1497
- "train/q_target_mean",
1498
- "train/replay_size",
1499
- "train/steps_per_s",
1500
- "train/success_rate",
1501
- "train/t_can",
1502
- "train/t_can_max",
1503
- "train/up_frac"
1504
- ],
1505
- "missing": [],
1506
- "n_points": 66,
1507
- "last": {
1508
- "train/global_steps": 1407680.0,
1509
- "time/elapsed_min": 49.99128341674805,
1510
- "train/t_can": 0.5096970796585083,
1511
- "train/success_rate": 0.0,
1512
- "eval/t_can": 27.979999542236328,
1513
- "eval/success": 1.0
1514
- }
1515
- },
1516
- {
1517
- "exp": "sac_full_base_rand",
1518
- "exists": true,
1519
- "tags": [
1520
- "eval/action_std",
1521
- "eval/ep_return",
1522
- "eval/ever_up",
1523
- "eval/final_ang_err",
1524
- "eval/first_up_t",
1525
- "eval/mean_abs_action",
1526
- "eval/mean_u1",
1527
- "eval/mean_u3",
1528
- "eval/steps",
1529
- "eval/success",
1530
- "eval/t_can",
1531
- "eval/up_frac",
1532
- "eval_final/action_std",
1533
- "eval_final/ep_return",
1534
- "eval_final/ever_up",
1535
- "eval_final/final_ang_err",
1536
- "eval_final/first_up_t",
1537
- "eval_final/mean_abs_action",
1538
- "eval_final/mean_u1",
1539
- "eval_final/mean_u3",
1540
- "eval_final/steps",
1541
- "eval_final/success",
1542
- "eval_final/t_can",
1543
- "eval_final/up_frac",
1544
- "loss/actor",
1545
- "loss/alpha",
1546
- "loss/critic",
1547
- "loss/entropy",
1548
- "loss/policy",
1549
- "loss/value",
1550
- "robust/ever_up_rate",
1551
- "robust/max_t_can",
1552
- "robust/mean_final_ang_err",
1553
- "robust/mean_return",
1554
- "robust/mean_t_can",
1555
- "robust/median_t_can",
1556
- "robust/min_t_can",
1557
- "robust/n",
1558
- "robust/success_rate",
1559
- "robust_best/ever_up_rate",
1560
- "robust_best/max_t_can",
1561
- "robust_best/mean_final_ang_err",
1562
- "robust_best/mean_return",
1563
- "robust_best/mean_t_can",
1564
- "robust_best/median_t_can",
1565
- "robust_best/min_t_can",
1566
- "robust_best/n",
1567
- "robust_best/success_rate",
1568
- "time/elapsed_min",
1569
- "train/action_abs_mean",
1570
- "train/action_std",
1571
- "train/alpha",
1572
- "train/entropy",
1573
- "train/episode_length",
1574
- "train/episode_return",
1575
- "train/ever_up_frac",
1576
- "train/first_up_time",
1577
- "train/global_steps",
1578
- "train/q_mean",
1579
- "train/q_target_mean",
1580
- "train/replay_size",
1581
- "train/steps_per_s",
1582
- "train/success_rate",
1583
- "train/t_can",
1584
- "train/t_can_max",
1585
- "train/up_frac"
1586
- ],
1587
- "missing": [],
1588
- "n_points": 66,
1589
- "last": {
1590
- "train/global_steps": 1424320.0,
1591
- "time/elapsed_min": 50.00120544433594,
1592
- "train/t_can": 5.080739498138428,
1593
- "train/success_rate": 0.0,
1594
- "eval/t_can": 26.260000228881836,
1595
- "eval/success": 1.0
1596
- }
1597
- }
1598
- ]
1599
- }