Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Feature] Consistent dropout #1587

Draft
wants to merge 4 commits into
base: main
Choose a base branch
from
Draft

[Feature] Consistent dropout #1587

wants to merge 4 commits into from

Conversation

vmoens
Copy link
Contributor

@vmoens vmoens commented Oct 1, 2023

@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 1, 2023
@vmoens vmoens added the enhancement New feature or request label Oct 1, 2023
@vmoens vmoens marked this pull request as draft October 1, 2023 05:21
@github-actions
Copy link

github-actions bot commented Oct 1, 2023

$\color{#D29922}\textsf{\Large⚠\kern{0.2cm}\normalsize Warning}$ Result of CPU Benchmark Tests

Total Benchmarks: 89. Improved: $\large\color{#35bf28}4$. Worsened: $\large\color{#d91a1a}5$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_single 65.7558ms 64.4588ms 15.5138 Ops/s 14.8073 Ops/s $\color{#35bf28}+4.77\%$
test_sync 43.4873ms 39.4983ms 25.3175 Ops/s 27.3641 Ops/s $\textbf{\color{#d91a1a}-7.48\%}$
test_async 55.8261ms 32.8799ms 30.4137 Ops/s 30.5909 Ops/s $\color{#d91a1a}-0.58\%$
test_simple 0.5096s 0.4468s 2.2381 Ops/s 2.2747 Ops/s $\color{#d91a1a}-1.61\%$
test_transformed 0.6834s 0.6216s 1.6088 Ops/s 1.6080 Ops/s $\color{#35bf28}+0.05\%$
test_serial 1.4323s 1.3712s 0.7293 Ops/s 0.7408 Ops/s $\color{#d91a1a}-1.56\%$
test_parallel 1.4076s 1.3582s 0.7363 Ops/s 0.7414 Ops/s $\color{#d91a1a}-0.69\%$
test_step_mdp_speed[True-True-True-True-True] 0.1530ms 23.0447μs 43.3939 KOps/s 43.7747 KOps/s $\color{#d91a1a}-0.87\%$
test_step_mdp_speed[True-True-True-True-False] 36.9980μs 14.0508μs 71.1701 KOps/s 73.3651 KOps/s $\color{#d91a1a}-2.99\%$
test_step_mdp_speed[True-True-True-False-True] 47.8990μs 14.1061μs 70.8913 KOps/s 71.7074 KOps/s $\color{#d91a1a}-1.14\%$
test_step_mdp_speed[True-True-True-False-False] 35.9670μs 8.3592μs 119.6280 KOps/s 121.3002 KOps/s $\color{#d91a1a}-1.38\%$
test_step_mdp_speed[True-True-False-True-True] 48.7100μs 24.7367μs 40.4258 KOps/s 41.4708 KOps/s $\color{#d91a1a}-2.52\%$
test_step_mdp_speed[True-True-False-True-False] 56.2550μs 15.4596μs 64.6846 KOps/s 66.5844 KOps/s $\color{#d91a1a}-2.85\%$
test_step_mdp_speed[True-True-False-False-True] 50.1140μs 15.2951μs 65.3803 KOps/s 66.2548 KOps/s $\color{#d91a1a}-1.32\%$
test_step_mdp_speed[True-True-False-False-False] 49.1110μs 9.7574μs 102.4863 KOps/s 105.0698 KOps/s $\color{#d91a1a}-2.46\%$
test_step_mdp_speed[True-False-True-True-True] 55.3330μs 26.0589μs 38.3746 KOps/s 39.1162 KOps/s $\color{#d91a1a}-1.90\%$
test_step_mdp_speed[True-False-True-True-False] 67.2450μs 16.8442μs 59.3674 KOps/s 61.3338 KOps/s $\color{#d91a1a}-3.21\%$
test_step_mdp_speed[True-False-True-False-True] 72.3350μs 15.4411μs 64.7623 KOps/s 65.9830 KOps/s $\color{#d91a1a}-1.85\%$
test_step_mdp_speed[True-False-True-False-False] 33.6630μs 9.8570μs 101.4507 KOps/s 103.8782 KOps/s $\color{#d91a1a}-2.34\%$
test_step_mdp_speed[True-False-False-True-True] 65.8620μs 27.1774μs 36.7953 KOps/s 37.2329 KOps/s $\color{#d91a1a}-1.18\%$
test_step_mdp_speed[True-False-False-True-False] 46.5470μs 18.1000μs 55.2485 KOps/s 57.6798 KOps/s $\color{#d91a1a}-4.22\%$
test_step_mdp_speed[True-False-False-False-True] 47.9190μs 16.6069μs 60.2158 KOps/s 61.0246 KOps/s $\color{#d91a1a}-1.33\%$
test_step_mdp_speed[True-False-False-False-False] 42.3690μs 11.0414μs 90.5681 KOps/s 92.7722 KOps/s $\color{#d91a1a}-2.38\%$
test_step_mdp_speed[False-True-True-True-True] 59.1100μs 26.0239μs 38.4262 KOps/s 39.1469 KOps/s $\color{#d91a1a}-1.84\%$
test_step_mdp_speed[False-True-True-True-False] 42.6990μs 16.9126μs 59.1274 KOps/s 61.2996 KOps/s $\color{#d91a1a}-3.54\%$
test_step_mdp_speed[False-True-True-False-True] 61.6340μs 17.8126μs 56.1402 KOps/s 57.5454 KOps/s $\color{#d91a1a}-2.44\%$
test_step_mdp_speed[False-True-True-False-False] 46.3770μs 11.1587μs 89.6165 KOps/s 91.1136 KOps/s $\color{#d91a1a}-1.64\%$
test_step_mdp_speed[False-True-False-True-True] 64.8800μs 27.0060μs 37.0288 KOps/s 37.2060 KOps/s $\color{#d91a1a}-0.48\%$
test_step_mdp_speed[False-True-False-True-False] 67.4550μs 17.9679μs 55.6549 KOps/s 56.8197 KOps/s $\color{#d91a1a}-2.05\%$
test_step_mdp_speed[False-True-False-False-True] 45.6750μs 18.9265μs 52.8361 KOps/s 53.5455 KOps/s $\color{#d91a1a}-1.32\%$
test_step_mdp_speed[False-True-False-False-False] 39.2830μs 12.3644μs 80.8771 KOps/s 83.2325 KOps/s $\color{#d91a1a}-2.83\%$
test_step_mdp_speed[False-False-True-True-True] 67.7970μs 28.4722μs 35.1220 KOps/s 35.2427 KOps/s $\color{#d91a1a}-0.34\%$
test_step_mdp_speed[False-False-True-True-False] 61.0230μs 19.3878μs 51.5789 KOps/s 52.7586 KOps/s $\color{#d91a1a}-2.24\%$
test_step_mdp_speed[False-False-True-False-True] 50.5550μs 19.1205μs 52.3000 KOps/s 53.0955 KOps/s $\color{#d91a1a}-1.50\%$
test_step_mdp_speed[False-False-True-False-False] 55.3930μs 12.5705μs 79.5512 KOps/s 83.2521 KOps/s $\color{#d91a1a}-4.45\%$
test_step_mdp_speed[False-False-False-True-True] 69.9600μs 29.4497μs 33.9562 KOps/s 34.3815 KOps/s $\color{#d91a1a}-1.24\%$
test_step_mdp_speed[False-False-False-True-False] 79.6490μs 20.6497μs 48.4269 KOps/s 50.2886 KOps/s $\color{#d91a1a}-3.70\%$
test_step_mdp_speed[False-False-False-False-True] 47.3280μs 19.8802μs 50.3014 KOps/s 51.5016 KOps/s $\color{#d91a1a}-2.33\%$
test_step_mdp_speed[False-False-False-False-False] 40.9660μs 13.4634μs 74.2755 KOps/s 76.3415 KOps/s $\color{#d91a1a}-2.71\%$
test_values[generalized_advantage_estimate-True-True] 17.1997ms 12.3933ms 80.6886 Ops/s 81.5070 Ops/s $\color{#d91a1a}-1.00\%$
test_values[vec_generalized_advantage_estimate-True-True] 33.6469ms 26.4621ms 37.7899 Ops/s 37.7800 Ops/s $\color{#35bf28}+0.03\%$
test_values[td0_return_estimate-False-False] 0.2480ms 0.1934ms 5.1700 KOps/s 5.4040 KOps/s $\color{#d91a1a}-4.33\%$
test_values[td1_return_estimate-False-False] 26.6334ms 26.2068ms 38.1580 Ops/s 38.1107 Ops/s $\color{#35bf28}+0.12\%$
test_values[vec_td1_return_estimate-False-False] 0.1094s 28.1754ms 35.4920 Ops/s 37.7370 Ops/s $\textbf{\color{#d91a1a}-5.95\%}$
test_values[td_lambda_return_estimate-True-False] 37.2673ms 36.2709ms 27.5703 Ops/s 27.3335 Ops/s $\color{#35bf28}+0.87\%$
test_values[vec_td_lambda_return_estimate-True-False] 33.4426ms 26.5953ms 37.6006 Ops/s 37.8989 Ops/s $\color{#d91a1a}-0.79\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 8.4059ms 8.1909ms 122.0869 Ops/s 122.7016 Ops/s $\color{#d91a1a}-0.50\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 9.9070ms 1.9285ms 518.5362 Ops/s 537.8400 Ops/s $\color{#d91a1a}-3.59\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.5777ms 0.4332ms 2.3084 KOps/s 2.3277 KOps/s $\color{#d91a1a}-0.83\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 48.9629ms 40.0820ms 24.9488 Ops/s 25.0599 Ops/s $\color{#d91a1a}-0.44\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 10.1133ms 2.5393ms 393.8043 Ops/s 390.6440 Ops/s $\color{#35bf28}+0.81\%$
test_dqn_speed 8.9575ms 1.6980ms 588.9453 Ops/s 558.3805 Ops/s $\textbf{\color{#35bf28}+5.47\%}$
test_ddpg_speed 10.8626ms 3.8727ms 258.2197 Ops/s 266.8785 Ops/s $\color{#d91a1a}-3.24\%$
test_sac_speed 18.1297ms 11.0384ms 90.5929 Ops/s 90.9825 Ops/s $\color{#d91a1a}-0.43\%$
test_redq_speed 28.1998ms 20.2622ms 49.3531 Ops/s 49.5943 Ops/s $\color{#d91a1a}-0.49\%$
test_redq_deprec_speed 0.1007s 17.9123ms 55.8275 Ops/s 61.5902 Ops/s $\textbf{\color{#d91a1a}-9.36\%}$
test_td3_speed 20.3837ms 11.5663ms 86.4582 Ops/s 91.7307 Ops/s $\textbf{\color{#d91a1a}-5.75\%}$
test_cql_speed 50.0496ms 40.4104ms 24.7461 Ops/s 22.8654 Ops/s $\textbf{\color{#35bf28}+8.22\%}$
test_a2c_speed 17.3493ms 9.2878ms 107.6679 Ops/s 107.4849 Ops/s $\color{#35bf28}+0.17\%$
test_ppo_speed 20.9833ms 9.8749ms 101.2673 Ops/s 106.8170 Ops/s $\textbf{\color{#d91a1a}-5.20\%}$
test_reinforce_speed 16.4438ms 8.3434ms 119.8552 Ops/s 106.4364 Ops/s $\textbf{\color{#35bf28}+12.61\%}$
test_iql_speed 45.2255ms 36.5803ms 27.3372 Ops/s 26.6801 Ops/s $\color{#35bf28}+2.46\%$
test_sample_rb[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 2.9900ms 2.0190ms 495.2834 Ops/s 495.8020 Ops/s $\color{#d91a1a}-0.10\%$
test_sample_rb[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 3.8161ms 2.1740ms 459.9746 Ops/s 467.6987 Ops/s $\color{#d91a1a}-1.65\%$
test_sample_rb[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 4.0965ms 2.1812ms 458.4699 Ops/s 467.7703 Ops/s $\color{#d91a1a}-1.99\%$
test_sample_rb[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 3.1349ms 2.0432ms 489.4389 Ops/s 498.3004 Ops/s $\color{#d91a1a}-1.78\%$
test_sample_rb[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 3.4732ms 2.1679ms 461.2735 Ops/s 475.6391 Ops/s $\color{#d91a1a}-3.02\%$
test_sample_rb[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 5.5061ms 2.1356ms 468.2456 Ops/s 476.0815 Ops/s $\color{#d91a1a}-1.65\%$
test_sample_rb[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 3.3603ms 2.0049ms 498.7885 Ops/s 495.5294 Ops/s $\color{#35bf28}+0.66\%$
test_sample_rb[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 3.6916ms 2.1250ms 470.5836 Ops/s 472.9755 Ops/s $\color{#d91a1a}-0.51\%$
test_sample_rb[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 4.1935ms 2.1423ms 466.7794 Ops/s 469.5794 Ops/s $\color{#d91a1a}-0.60\%$
test_iterate_rb[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 2.9547ms 1.9832ms 504.2249 Ops/s 505.0755 Ops/s $\color{#d91a1a}-0.17\%$
test_iterate_rb[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 4.3742ms 2.1610ms 462.7460 Ops/s 473.1661 Ops/s $\color{#d91a1a}-2.20\%$
test_iterate_rb[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 3.0440ms 2.1584ms 463.3037 Ops/s 461.2771 Ops/s $\color{#35bf28}+0.44\%$
test_iterate_rb[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 3.0117ms 2.0157ms 496.1068 Ops/s 487.6447 Ops/s $\color{#35bf28}+1.74\%$
test_iterate_rb[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 3.8433ms 2.1596ms 463.0413 Ops/s 468.5596 Ops/s $\color{#d91a1a}-1.18\%$
test_iterate_rb[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 3.5270ms 2.1520ms 464.6805 Ops/s 472.3757 Ops/s $\color{#d91a1a}-1.63\%$
test_iterate_rb[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 2.2780ms 2.0241ms 494.0524 Ops/s 498.3835 Ops/s $\color{#d91a1a}-0.87\%$
test_iterate_rb[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 4.0059ms 2.1682ms 461.2212 Ops/s 473.7695 Ops/s $\color{#d91a1a}-2.65\%$
test_iterate_rb[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 3.2334ms 2.1635ms 462.2083 Ops/s 472.0826 Ops/s $\color{#d91a1a}-2.09\%$
test_populate_rb[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.2051s 18.8636ms 53.0121 Ops/s 54.1080 Ops/s $\color{#d91a1a}-2.03\%$
test_populate_rb[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 0.1164s 17.0063ms 58.8019 Ops/s 57.8444 Ops/s $\color{#35bf28}+1.66\%$
test_populate_rb[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 0.1176s 17.1378ms 58.3506 Ops/s 59.0854 Ops/s $\color{#d91a1a}-1.24\%$
test_populate_rb[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.1232s 17.1177ms 58.4191 Ops/s 56.5102 Ops/s $\color{#35bf28}+3.38\%$
test_populate_rb[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 0.1228s 17.5075ms 57.1184 Ops/s 50.8615 Ops/s $\textbf{\color{#35bf28}+12.30\%}$
test_populate_rb[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 0.1229s 17.3998ms 57.4720 Ops/s 57.8024 Ops/s $\color{#d91a1a}-0.57\%$
test_populate_rb[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.1210s 15.0724ms 66.3463 Ops/s 65.3910 Ops/s $\color{#35bf28}+1.46\%$
test_populate_rb[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 0.1176s 17.1180ms 58.4181 Ops/s 57.6726 Ops/s $\color{#35bf28}+1.29\%$
test_populate_rb[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 0.1205s 17.4023ms 57.4636 Ops/s 57.4471 Ops/s $\color{#35bf28}+0.03\%$

Copy link

pytorch-bot bot commented Nov 27, 2023

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/1587

Note: Links to docs will display an error until the docs builds have been completed.

✅ You can merge normally! (8 Unrelated Failures)

As of commit 70ab0f8 with merge base bc7595f (image):

FLAKY - The following jobs failed but were likely due to flakiness present on trunk:

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

This comment was automatically generated by Dr. CI and updates every 15 minutes.

Copy link

$\color{#D29922}\textsf{\Large⚠\kern{0.2cm}\normalsize Warning}$ Result of GPU Benchmark Tests

Total Benchmarks: 92. Improved: $\large\color{#35bf28}5$. Worsened: $\large\color{#d91a1a}3$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_single 0.1230s 0.1200s 8.3301 Ops/s 8.2933 Ops/s $\color{#35bf28}+0.44\%$
test_sync 0.1019s 0.1017s 9.8286 Ops/s 9.7555 Ops/s $\color{#35bf28}+0.75\%$
test_async 0.2697s 98.7397ms 10.1276 Ops/s 10.0007 Ops/s $\color{#35bf28}+1.27\%$
test_single_pixels 0.1431s 0.1427s 7.0083 Ops/s 7.0002 Ops/s $\color{#35bf28}+0.12\%$
test_sync_pixels 99.6984ms 94.9585ms 10.5309 Ops/s 10.5436 Ops/s $\color{#d91a1a}-0.12\%$
test_async_pixels 0.1841s 90.6840ms 11.0273 Ops/s 10.9570 Ops/s $\color{#35bf28}+0.64\%$
test_simple 0.9308s 0.8631s 1.1586 Ops/s 1.1460 Ops/s $\color{#35bf28}+1.10\%$
test_transformed 1.1670s 1.1058s 0.9043 Ops/s 0.8941 Ops/s $\color{#35bf28}+1.15\%$
test_serial 2.4695s 2.4070s 0.4155 Ops/s 0.4110 Ops/s $\color{#35bf28}+1.08\%$
test_parallel 2.5846s 2.5243s 0.3962 Ops/s 0.4024 Ops/s $\color{#d91a1a}-1.56\%$
test_step_mdp_speed[True-True-True-True-True] 71.9610μs 35.1156μs 28.4774 KOps/s 27.8686 KOps/s $\color{#35bf28}+2.18\%$
test_step_mdp_speed[True-True-True-True-False] 45.8410μs 20.9929μs 47.6351 KOps/s 47.1576 KOps/s $\color{#35bf28}+1.01\%$
test_step_mdp_speed[True-True-True-False-True] 48.3110μs 21.2030μs 47.1632 KOps/s 47.4371 KOps/s $\color{#d91a1a}-0.58\%$
test_step_mdp_speed[True-True-True-False-False] 36.7800μs 12.1936μs 82.0104 KOps/s 80.1133 KOps/s $\color{#35bf28}+2.37\%$
test_step_mdp_speed[True-True-False-True-True] 65.0310μs 36.1393μs 27.6707 KOps/s 26.8441 KOps/s $\color{#35bf28}+3.08\%$
test_step_mdp_speed[True-True-False-True-False] 47.0010μs 22.3476μs 44.7475 KOps/s 44.2484 KOps/s $\color{#35bf28}+1.13\%$
test_step_mdp_speed[True-True-False-False-True] 47.4810μs 21.7309μs 46.0175 KOps/s 44.2794 KOps/s $\color{#35bf28}+3.93\%$
test_step_mdp_speed[True-True-False-False-False] 38.3810μs 13.9012μs 71.9364 KOps/s 69.5820 KOps/s $\color{#35bf28}+3.38\%$
test_step_mdp_speed[True-False-True-True-True] 77.5320μs 38.8675μs 25.7285 KOps/s 25.2231 KOps/s $\color{#35bf28}+2.00\%$
test_step_mdp_speed[True-False-True-True-False] 58.3110μs 25.2335μs 39.6298 KOps/s 40.7271 KOps/s $\color{#d91a1a}-2.69\%$
test_step_mdp_speed[True-False-True-False-True] 57.0610μs 23.1922μs 43.1180 KOps/s 43.7937 KOps/s $\color{#d91a1a}-1.54\%$
test_step_mdp_speed[True-False-True-False-False] 37.4210μs 14.2340μs 70.2541 KOps/s 70.9671 KOps/s $\color{#d91a1a}-1.00\%$
test_step_mdp_speed[True-False-False-True-True] 65.8110μs 40.4899μs 24.6975 KOps/s 23.8441 KOps/s $\color{#35bf28}+3.58\%$
test_step_mdp_speed[True-False-False-True-False] 49.5610μs 25.8592μs 38.6710 KOps/s 38.1628 KOps/s $\color{#35bf28}+1.33\%$
test_step_mdp_speed[True-False-False-False-True] 50.0210μs 24.3085μs 41.1379 KOps/s 40.6969 KOps/s $\color{#35bf28}+1.08\%$
test_step_mdp_speed[True-False-False-False-False] 38.5200μs 15.6146μs 64.0428 KOps/s 61.8681 KOps/s $\color{#35bf28}+3.52\%$
test_step_mdp_speed[False-True-True-True-True] 90.2410μs 38.3699μs 26.0621 KOps/s 25.3182 KOps/s $\color{#35bf28}+2.94\%$
test_step_mdp_speed[False-True-True-True-False] 50.1600μs 24.6754μs 40.5262 KOps/s 40.4371 KOps/s $\color{#35bf28}+0.22\%$
test_step_mdp_speed[False-True-True-False-True] 42.7710μs 26.4194μs 37.8510 KOps/s 37.2505 KOps/s $\color{#35bf28}+1.61\%$
test_step_mdp_speed[False-True-True-False-False] 36.8010μs 15.9288μs 62.7794 KOps/s 60.8671 KOps/s $\color{#35bf28}+3.14\%$
test_step_mdp_speed[False-True-False-True-True] 69.0110μs 41.6889μs 23.9872 KOps/s 23.9642 KOps/s $\color{#35bf28}+0.10\%$
test_step_mdp_speed[False-True-False-True-False] 50.3900μs 26.1431μs 38.2511 KOps/s 37.4685 KOps/s $\color{#35bf28}+2.09\%$
test_step_mdp_speed[False-True-False-False-True] 51.4300μs 28.5880μs 34.9797 KOps/s 34.5428 KOps/s $\color{#35bf28}+1.26\%$
test_step_mdp_speed[False-True-False-False-False] 35.8610μs 17.9592μs 55.6817 KOps/s 54.9325 KOps/s $\color{#35bf28}+1.36\%$
test_step_mdp_speed[False-False-True-True-True] 0.1067ms 42.4081μs 23.5804 KOps/s 22.8306 KOps/s $\color{#35bf28}+3.28\%$
test_step_mdp_speed[False-False-True-True-False] 56.8910μs 27.9906μs 35.7262 KOps/s 34.6880 KOps/s $\color{#35bf28}+2.99\%$
test_step_mdp_speed[False-False-True-False-True] 55.1010μs 28.3231μs 35.3068 KOps/s 34.4308 KOps/s $\color{#35bf28}+2.54\%$
test_step_mdp_speed[False-False-True-False-False] 56.0810μs 18.0295μs 55.4645 KOps/s 55.2939 KOps/s $\color{#35bf28}+0.31\%$
test_step_mdp_speed[False-False-False-True-True] 0.1054ms 43.6132μs 22.9288 KOps/s 22.0254 KOps/s $\color{#35bf28}+4.10\%$
test_step_mdp_speed[False-False-False-True-False] 64.8420μs 29.8421μs 33.5097 KOps/s 32.3080 KOps/s $\color{#35bf28}+3.72\%$
test_step_mdp_speed[False-False-False-False-True] 58.3310μs 28.8449μs 34.6682 KOps/s 32.9065 KOps/s $\textbf{\color{#35bf28}+5.35\%}$
test_step_mdp_speed[False-False-False-False-False] 41.2010μs 19.4394μs 51.4420 KOps/s 48.6864 KOps/s $\textbf{\color{#35bf28}+5.66\%}$
test_values[generalized_advantage_estimate-True-True] 25.7269ms 24.7857ms 40.3458 Ops/s 40.4340 Ops/s $\color{#d91a1a}-0.22\%$
test_values[vec_generalized_advantage_estimate-True-True] 86.9101ms 3.2791ms 304.9621 Ops/s 312.6387 Ops/s $\color{#d91a1a}-2.46\%$
test_values[td0_return_estimate-False-False] 94.9610μs 61.3559μs 16.2984 KOps/s 15.9680 KOps/s $\color{#35bf28}+2.07\%$
test_values[td1_return_estimate-False-False] 55.7182ms 54.3577ms 18.3967 Ops/s 18.7622 Ops/s $\color{#d91a1a}-1.95\%$
test_values[vec_td1_return_estimate-False-False] 1.9596ms 1.7179ms 582.1105 Ops/s 586.7441 Ops/s $\color{#d91a1a}-0.79\%$
test_values[td_lambda_return_estimate-True-False] 89.9322ms 88.3096ms 11.3238 Ops/s 11.6331 Ops/s $\color{#d91a1a}-2.66\%$
test_values[vec_td_lambda_return_estimate-True-False] 1.9843ms 1.7162ms 582.6682 Ops/s 585.4342 Ops/s $\color{#d91a1a}-0.47\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 24.7016ms 23.9508ms 41.7523 Ops/s 42.0637 Ops/s $\color{#d91a1a}-0.74\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 0.8431ms 0.6960ms 1.4368 KOps/s 1.4214 KOps/s $\color{#35bf28}+1.08\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.7600ms 0.6789ms 1.4729 KOps/s 1.5094 KOps/s $\color{#d91a1a}-2.42\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 1.5168ms 1.4626ms 683.7223 Ops/s 683.0126 Ops/s $\color{#35bf28}+0.10\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 0.9520ms 0.6818ms 1.4668 KOps/s 1.4312 KOps/s $\color{#35bf28}+2.49\%$
test_dqn_speed 7.7903ms 1.4546ms 687.4747 Ops/s 668.8062 Ops/s $\color{#35bf28}+2.79\%$
test_ddpg_speed 4.7032ms 3.3237ms 300.8732 Ops/s 300.8100 Ops/s $\color{#35bf28}+0.02\%$
test_sac_speed 94.2086ms 10.0866ms 99.1411 Ops/s 108.7263 Ops/s $\textbf{\color{#d91a1a}-8.82\%}$
test_redq_speed 16.9641ms 16.4793ms 60.6821 Ops/s 60.1442 Ops/s $\color{#35bf28}+0.89\%$
test_redq_deprec_speed 14.2135ms 12.9746ms 77.0739 Ops/s 77.4876 Ops/s $\color{#d91a1a}-0.53\%$
test_td3_speed 19.2103ms 9.4817ms 105.4669 Ops/s 105.9975 Ops/s $\color{#d91a1a}-0.50\%$
test_cql_speed 32.6804ms 31.3788ms 31.8686 Ops/s 30.4686 Ops/s $\color{#35bf28}+4.59\%$
test_a2c_speed 8.6134ms 7.1283ms 140.2868 Ops/s 141.7486 Ops/s $\color{#d91a1a}-1.03\%$
test_ppo_speed 8.8716ms 7.4452ms 134.3145 Ops/s 136.6788 Ops/s $\color{#d91a1a}-1.73\%$
test_reinforce_speed 7.5503ms 6.1457ms 162.7160 Ops/s 165.2620 Ops/s $\color{#d91a1a}-1.54\%$
test_iql_speed 28.6921ms 27.1953ms 36.7711 Ops/s 37.0608 Ops/s $\color{#d91a1a}-0.78\%$
test_sample_rb[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 3.1846ms 2.5026ms 399.5868 Ops/s 396.2916 Ops/s $\color{#35bf28}+0.83\%$
test_sample_rb[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 3.9185ms 2.6861ms 372.2848 Ops/s 326.1991 Ops/s $\textbf{\color{#35bf28}+14.13\%}$
test_sample_rb[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 3.9450ms 2.6902ms 371.7231 Ops/s 372.9594 Ops/s $\color{#d91a1a}-0.33\%$
test_sample_rb[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 3.1463ms 2.4867ms 402.1442 Ops/s 398.9222 Ops/s $\color{#35bf28}+0.81\%$
test_sample_rb[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 3.7607ms 2.6830ms 372.7223 Ops/s 370.4507 Ops/s $\color{#35bf28}+0.61\%$
test_sample_rb[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 3.9494ms 2.6937ms 371.2352 Ops/s 373.2656 Ops/s $\color{#d91a1a}-0.54\%$
test_sample_rb[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 2.7234ms 2.5048ms 399.2321 Ops/s 397.2757 Ops/s $\color{#35bf28}+0.49\%$
test_sample_rb[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 4.0177ms 2.6917ms 371.5145 Ops/s 370.1825 Ops/s $\color{#35bf28}+0.36\%$
test_sample_rb[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 3.8349ms 2.6903ms 371.7022 Ops/s 372.9640 Ops/s $\color{#d91a1a}-0.34\%$
test_iterate_rb[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 2.7218ms 2.5148ms 397.6387 Ops/s 397.1890 Ops/s $\color{#35bf28}+0.11\%$
test_iterate_rb[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 4.2305ms 2.6971ms 370.7629 Ops/s 371.6203 Ops/s $\color{#d91a1a}-0.23\%$
test_iterate_rb[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 4.0843ms 2.6988ms 370.5404 Ops/s 371.2687 Ops/s $\color{#d91a1a}-0.20\%$
test_iterate_rb[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 3.0685ms 2.4898ms 401.6404 Ops/s 396.7588 Ops/s $\color{#35bf28}+1.23\%$
test_iterate_rb[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 4.0895ms 2.7137ms 368.5059 Ops/s 370.9746 Ops/s $\color{#d91a1a}-0.67\%$
test_iterate_rb[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 4.1527ms 2.6942ms 371.1740 Ops/s 370.1471 Ops/s $\color{#35bf28}+0.28\%$
test_iterate_rb[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 3.0908ms 2.5128ms 397.9676 Ops/s 396.4535 Ops/s $\color{#35bf28}+0.38\%$
test_iterate_rb[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 4.3490ms 2.9596ms 337.8813 Ops/s 368.9570 Ops/s $\textbf{\color{#d91a1a}-8.42\%}$
test_iterate_rb[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 3.5711ms 2.6931ms 371.3154 Ops/s 370.5835 Ops/s $\color{#35bf28}+0.20\%$
test_populate_rb[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.2133s 19.4106ms 51.5182 Ops/s 51.8002 Ops/s $\color{#d91a1a}-0.54\%$
test_populate_rb[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 0.1250s 15.3120ms 65.3081 Ops/s 56.6158 Ops/s $\textbf{\color{#35bf28}+15.35\%}$
test_populate_rb[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 0.1268s 17.5941ms 56.8372 Ops/s 56.3637 Ops/s $\color{#35bf28}+0.84\%$
test_populate_rb[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.1272s 17.4248ms 57.3896 Ops/s 64.7084 Ops/s $\textbf{\color{#d91a1a}-11.31\%}$
test_populate_rb[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 0.1252s 17.6077ms 56.7933 Ops/s 56.4270 Ops/s $\color{#35bf28}+0.65\%$
test_populate_rb[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 0.1258s 17.6151ms 56.7696 Ops/s 56.6489 Ops/s $\color{#35bf28}+0.21\%$
test_populate_rb[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.1264s 17.4903ms 57.1745 Ops/s 56.1512 Ops/s $\color{#35bf28}+1.82\%$
test_populate_rb[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 0.1256s 17.5036ms 57.1311 Ops/s 56.0507 Ops/s $\color{#35bf28}+1.93\%$
test_populate_rb[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 0.1258s 15.3529ms 65.1344 Ops/s 56.2626 Ops/s $\textbf{\color{#35bf28}+15.77\%}$

@N00bcak N00bcak mentioned this pull request Aug 15, 2024
9 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. enhancement New feature or request
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants