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] Composite.separates #2599

Merged
merged 2 commits into from
Nov 24, 2024
Merged

[Feature] Composite.separates #2599

merged 2 commits into from
Nov 24, 2024

Conversation

[ghstack-poisoned]
Copy link

pytorch-bot bot commented Nov 23, 2024

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/2599

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

❌ 18 New Failures, 2 Pending

As of commit 6210832 with merge base 152bc81 (image):

NEW FAILURES - The following jobs have failed:

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

@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 Nov 23, 2024
[ghstack-poisoned]
@vmoens vmoens merged commit 6210832 into gh/vmoens/46/base Nov 24, 2024
38 of 58 checks passed
vmoens added a commit that referenced this pull request Nov 24, 2024
ghstack-source-id: fbfc4308a81cd96ecc61723df8c0eb870c442def
Pull Request resolved: #2599
@vmoens vmoens deleted the gh/vmoens/46/head branch November 24, 2024 08:17
Copy link

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

Total Benchmarks: 149. Improved: $\large\color{#35bf28}5$. Worsened: $\large\color{#d91a1a}13$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_simple 0.4233s 0.4215s 2.3726 Ops/s 2.2495 Ops/s $\textbf{\color{#35bf28}+5.47\%}$
test_transformed 0.5945s 0.5915s 1.6907 Ops/s 1.6806 Ops/s $\color{#35bf28}+0.60\%$
test_serial 1.3397s 1.3360s 0.7485 Ops/s 0.7397 Ops/s $\color{#35bf28}+1.20\%$
test_parallel 1.3699s 1.2914s 0.7743 Ops/s 0.7683 Ops/s $\color{#35bf28}+0.78\%$
test_step_mdp_speed[True-True-True-True-True] 0.3086ms 26.2482μs 38.0978 KOps/s 36.6817 KOps/s $\color{#35bf28}+3.86\%$
test_step_mdp_speed[True-True-True-True-False] 54.0880μs 15.3441μs 65.1715 KOps/s 63.6823 KOps/s $\color{#35bf28}+2.34\%$
test_step_mdp_speed[True-True-True-False-True] 42.8200μs 14.9696μs 66.8019 KOps/s 64.2038 KOps/s $\color{#35bf28}+4.05\%$
test_step_mdp_speed[True-True-True-False-False] 52.2680μs 8.7013μs 114.9248 KOps/s 110.8577 KOps/s $\color{#35bf28}+3.67\%$
test_step_mdp_speed[True-True-False-True-True] 66.6640μs 28.1322μs 35.5465 KOps/s 34.4161 KOps/s $\color{#35bf28}+3.28\%$
test_step_mdp_speed[True-True-False-True-False] 44.1020μs 16.8691μs 59.2801 KOps/s 57.5728 KOps/s $\color{#35bf28}+2.97\%$
test_step_mdp_speed[True-True-False-False-True] 57.4870μs 16.0581μs 62.2737 KOps/s 58.3313 KOps/s $\textbf{\color{#35bf28}+6.76\%}$
test_step_mdp_speed[True-True-False-False-False] 38.6320μs 10.4670μs 95.5387 KOps/s 93.5816 KOps/s $\color{#35bf28}+2.09\%$
test_step_mdp_speed[True-False-True-True-True] 72.2340μs 29.0775μs 34.3909 KOps/s 32.8247 KOps/s $\color{#35bf28}+4.77\%$
test_step_mdp_speed[True-False-True-True-False] 51.6970μs 18.4650μs 54.1565 KOps/s 51.8485 KOps/s $\color{#35bf28}+4.45\%$
test_step_mdp_speed[True-False-True-False-True] 46.8580μs 16.4284μs 60.8703 KOps/s 57.7829 KOps/s $\textbf{\color{#35bf28}+5.34\%}$
test_step_mdp_speed[True-False-True-False-False] 50.4750μs 10.3430μs 96.6838 KOps/s 94.7291 KOps/s $\color{#35bf28}+2.06\%$
test_step_mdp_speed[True-False-False-True-True] 77.1340μs 31.2796μs 31.9698 KOps/s 31.5503 KOps/s $\color{#35bf28}+1.33\%$
test_step_mdp_speed[True-False-False-True-False] 51.1450μs 20.2983μs 49.2651 KOps/s 48.6956 KOps/s $\color{#35bf28}+1.17\%$
test_step_mdp_speed[True-False-False-False-True] 51.6660μs 18.0409μs 55.4296 KOps/s 54.4872 KOps/s $\color{#35bf28}+1.73\%$
test_step_mdp_speed[True-False-False-False-False] 38.4820μs 11.9376μs 83.7691 KOps/s 82.4695 KOps/s $\color{#35bf28}+1.58\%$
test_step_mdp_speed[False-True-True-True-True] 73.7570μs 29.8324μs 33.5206 KOps/s 32.9487 KOps/s $\color{#35bf28}+1.74\%$
test_step_mdp_speed[False-True-True-True-False] 81.2890μs 18.7406μs 53.3602 KOps/s 52.3301 KOps/s $\color{#35bf28}+1.97\%$
test_step_mdp_speed[False-True-True-False-True] 51.7970μs 18.6467μs 53.6288 KOps/s 51.2543 KOps/s $\color{#35bf28}+4.63\%$
test_step_mdp_speed[False-True-True-False-False] 32.6710μs 11.4416μs 87.4000 KOps/s 83.7049 KOps/s $\color{#35bf28}+4.41\%$
test_step_mdp_speed[False-True-False-True-True] 63.2680μs 31.0282μs 32.2287 KOps/s 31.2544 KOps/s $\color{#35bf28}+3.12\%$
test_step_mdp_speed[False-True-False-True-False] 71.8470μs 20.0709μs 49.8234 KOps/s 48.7089 KOps/s $\color{#35bf28}+2.29\%$
test_step_mdp_speed[False-True-False-False-True] 3.0867ms 20.2584μs 49.3622 KOps/s 47.2750 KOps/s $\color{#35bf28}+4.41\%$
test_step_mdp_speed[False-True-False-False-False] 43.9920μs 13.0398μs 76.6881 KOps/s 73.2569 KOps/s $\color{#35bf28}+4.68\%$
test_step_mdp_speed[False-False-True-True-True] 75.6710μs 32.4462μs 30.8203 KOps/s 29.8360 KOps/s $\color{#35bf28}+3.30\%$
test_step_mdp_speed[False-False-True-True-False] 63.3880μs 22.0452μs 45.3613 KOps/s 44.9687 KOps/s $\color{#35bf28}+0.87\%$
test_step_mdp_speed[False-False-True-False-True] 48.7510μs 20.5878μs 48.5725 KOps/s 48.2721 KOps/s $\color{#35bf28}+0.62\%$
test_step_mdp_speed[False-False-True-False-False] 43.2810μs 13.0556μs 76.5957 KOps/s 74.3095 KOps/s $\color{#35bf28}+3.08\%$
test_step_mdp_speed[False-False-False-True-True] 75.2700μs 33.2269μs 30.0961 KOps/s 28.7378 KOps/s $\color{#35bf28}+4.73\%$
test_step_mdp_speed[False-False-False-True-False] 53.5510μs 22.9255μs 43.6195 KOps/s 42.4837 KOps/s $\color{#35bf28}+2.67\%$
test_step_mdp_speed[False-False-False-False-True] 53.0990μs 21.5192μs 46.4701 KOps/s 45.1770 KOps/s $\color{#35bf28}+2.86\%$
test_step_mdp_speed[False-False-False-False-False] 52.5480μs 14.3776μs 69.5527 KOps/s 67.4900 KOps/s $\color{#35bf28}+3.06\%$
test_values[generalized_advantage_estimate-True-True] 13.7425ms 9.7498ms 102.5667 Ops/s 103.8049 Ops/s $\color{#d91a1a}-1.19\%$
test_values[vec_generalized_advantage_estimate-True-True] 38.8387ms 35.5406ms 28.1368 Ops/s 27.6762 Ops/s $\color{#35bf28}+1.66\%$
test_values[td0_return_estimate-False-False] 0.2480ms 0.1755ms 5.6976 KOps/s 5.7256 KOps/s $\color{#d91a1a}-0.49\%$
test_values[td1_return_estimate-False-False] 26.8690ms 24.1243ms 41.4519 Ops/s 41.5125 Ops/s $\color{#d91a1a}-0.15\%$
test_values[vec_td1_return_estimate-False-False] 38.0445ms 35.7706ms 27.9559 Ops/s 27.4598 Ops/s $\color{#35bf28}+1.81\%$
test_values[td_lambda_return_estimate-True-False] 35.5171ms 34.9116ms 28.6438 Ops/s 28.7994 Ops/s $\color{#d91a1a}-0.54\%$
test_values[vec_td_lambda_return_estimate-True-False] 38.0388ms 35.8163ms 27.9203 Ops/s 27.4865 Ops/s $\color{#35bf28}+1.58\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 11.2159ms 8.5183ms 117.3949 Ops/s 117.0544 Ops/s $\color{#35bf28}+0.29\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 2.3056ms 1.9043ms 525.1217 Ops/s 522.1952 Ops/s $\color{#35bf28}+0.56\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.5845ms 0.3671ms 2.7240 KOps/s 2.7699 KOps/s $\color{#d91a1a}-1.66\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 50.0009ms 46.2748ms 21.6100 Ops/s 23.3603 Ops/s $\textbf{\color{#d91a1a}-7.49\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 3.9999ms 3.0289ms 330.1483 Ops/s 328.3594 Ops/s $\color{#35bf28}+0.54\%$
test_dqn_speed[False-None] 6.6501ms 1.3489ms 741.3214 Ops/s 748.3450 Ops/s $\color{#d91a1a}-0.94\%$
test_dqn_speed[False-backward] 1.8805ms 1.8213ms 549.0694 Ops/s 548.6911 Ops/s $\color{#35bf28}+0.07\%$
test_dqn_speed[True-None] 0.6421ms 0.4614ms 2.1675 KOps/s 2.1268 KOps/s $\color{#35bf28}+1.91\%$
test_dqn_speed[True-backward] 0.9369ms 0.8911ms 1.1222 KOps/s 1.0188 KOps/s $\textbf{\color{#35bf28}+10.15\%}$
test_dqn_speed[reduce-overhead-None] 0.5688ms 0.4641ms 2.1547 KOps/s 2.1623 KOps/s $\color{#d91a1a}-0.35\%$
test_dqn_speed[reduce-overhead-backward] 0.9788ms 0.8893ms 1.1244 KOps/s 1.1240 KOps/s $\color{#35bf28}+0.04\%$
test_ddpg_speed[False-None] 0.1877s 3.3327ms 300.0593 Ops/s 359.0086 Ops/s $\textbf{\color{#d91a1a}-16.42\%}$
test_ddpg_speed[False-backward] 4.9156ms 4.0097ms 249.3935 Ops/s 255.5205 Ops/s $\color{#d91a1a}-2.40\%$
test_ddpg_speed[True-None] 1.2724ms 1.0120ms 988.1267 Ops/s 990.4115 Ops/s $\color{#d91a1a}-0.23\%$
test_ddpg_speed[True-backward] 1.9689ms 1.9077ms 524.1809 Ops/s 517.8566 Ops/s $\color{#35bf28}+1.22\%$
test_ddpg_speed[reduce-overhead-None] 1.3809ms 1.0153ms 984.9701 Ops/s 987.2001 Ops/s $\color{#d91a1a}-0.23\%$
test_ddpg_speed[reduce-overhead-backward] 1.9758ms 1.8983ms 526.7878 Ops/s 517.4128 Ops/s $\color{#35bf28}+1.81\%$
test_sac_speed[False-None] 0.2123s 9.3668ms 106.7598 Ops/s 127.3863 Ops/s $\textbf{\color{#d91a1a}-16.19\%}$
test_sac_speed[False-backward] 10.9951ms 10.5334ms 94.9362 Ops/s 94.7499 Ops/s $\color{#35bf28}+0.20\%$
test_sac_speed[True-None] 2.5251ms 1.8755ms 533.1861 Ops/s 531.1988 Ops/s $\color{#35bf28}+0.37\%$
test_sac_speed[True-backward] 4.4705ms 3.6822ms 271.5735 Ops/s 279.8995 Ops/s $\color{#d91a1a}-2.97\%$
test_sac_speed[reduce-overhead-None] 2.4752ms 1.8895ms 529.2405 Ops/s 527.4489 Ops/s $\color{#35bf28}+0.34\%$
test_sac_speed[reduce-overhead-backward] 3.6899ms 3.5615ms 280.7835 Ops/s 271.2029 Ops/s $\color{#35bf28}+3.53\%$
test_redq_speed[False-None] 13.9157ms 12.9142ms 77.4343 Ops/s 76.3845 Ops/s $\color{#35bf28}+1.37\%$
test_redq_speed[False-backward] 23.2585ms 22.1062ms 45.2361 Ops/s 44.4546 Ops/s $\color{#35bf28}+1.76\%$
test_redq_speed[True-None] 6.1095ms 5.0702ms 197.2318 Ops/s 218.8762 Ops/s $\textbf{\color{#d91a1a}-9.89\%}$
test_redq_speed[True-backward] 13.1952ms 12.2801ms 81.4323 Ops/s 82.7024 Ops/s $\color{#d91a1a}-1.54\%$
test_redq_speed[reduce-overhead-None] 5.3160ms 4.6644ms 214.3912 Ops/s 218.1654 Ops/s $\color{#d91a1a}-1.73\%$
test_redq_speed[reduce-overhead-backward] 13.0608ms 12.0484ms 82.9985 Ops/s 76.7099 Ops/s $\textbf{\color{#35bf28}+8.20\%}$
test_redq_deprec_speed[False-None] 15.0473ms 12.6389ms 79.1211 Ops/s 76.0606 Ops/s $\color{#35bf28}+4.02\%$
test_redq_deprec_speed[False-backward] 20.0570ms 19.0766ms 52.4202 Ops/s 53.2356 Ops/s $\color{#d91a1a}-1.53\%$
test_redq_deprec_speed[True-None] 4.0182ms 3.6044ms 277.4351 Ops/s 271.8554 Ops/s $\color{#35bf28}+2.05\%$
test_redq_deprec_speed[True-backward] 8.4865ms 8.0654ms 123.9864 Ops/s 123.7253 Ops/s $\color{#35bf28}+0.21\%$
test_redq_deprec_speed[reduce-overhead-None] 4.2616ms 3.7739ms 264.9794 Ops/s 277.7571 Ops/s $\color{#d91a1a}-4.60\%$
test_redq_deprec_speed[reduce-overhead-backward] 8.4743ms 8.0507ms 124.2133 Ops/s 123.5366 Ops/s $\color{#35bf28}+0.55\%$
test_td3_speed[False-None] 8.1650ms 7.8536ms 127.3305 Ops/s 129.3393 Ops/s $\color{#d91a1a}-1.55\%$
test_td3_speed[False-backward] 12.5715ms 10.3415ms 96.6981 Ops/s 98.2529 Ops/s $\color{#d91a1a}-1.58\%$
test_td3_speed[True-None] 2.0527ms 1.7691ms 565.2649 Ops/s 565.0730 Ops/s $\color{#35bf28}+0.03\%$
test_td3_speed[True-backward] 4.2485ms 3.4235ms 292.0958 Ops/s 298.5637 Ops/s $\color{#d91a1a}-2.17\%$
test_td3_speed[reduce-overhead-None] 2.0553ms 1.7623ms 567.4299 Ops/s 575.4828 Ops/s $\color{#d91a1a}-1.40\%$
test_td3_speed[reduce-overhead-backward] 3.7965ms 3.5090ms 284.9817 Ops/s 297.5825 Ops/s $\color{#d91a1a}-4.23\%$
test_cql_speed[False-None] 41.2391ms 36.0456ms 27.7426 Ops/s 28.3471 Ops/s $\color{#d91a1a}-2.13\%$
test_cql_speed[False-backward] 50.5139ms 46.8487ms 21.3453 Ops/s 21.9719 Ops/s $\color{#d91a1a}-2.85\%$
test_cql_speed[True-None] 17.2480ms 15.5760ms 64.2013 Ops/s 63.5759 Ops/s $\color{#35bf28}+0.98\%$
test_cql_speed[True-backward] 24.3361ms 22.3576ms 44.7275 Ops/s 44.0610 Ops/s $\color{#35bf28}+1.51\%$
test_cql_speed[reduce-overhead-None] 17.9608ms 16.1201ms 62.0342 Ops/s 63.6402 Ops/s $\color{#d91a1a}-2.52\%$
test_cql_speed[reduce-overhead-backward] 24.2812ms 23.2369ms 43.0350 Ops/s 43.9894 Ops/s $\color{#d91a1a}-2.17\%$
test_a2c_speed[False-None] 9.0830ms 7.4005ms 135.1265 Ops/s 140.7245 Ops/s $\color{#d91a1a}-3.98\%$
test_a2c_speed[False-backward] 15.8014ms 14.9899ms 66.7116 Ops/s 70.3058 Ops/s $\textbf{\color{#d91a1a}-5.11\%}$
test_a2c_speed[True-None] 5.0865ms 4.3502ms 229.8721 Ops/s 235.1208 Ops/s $\color{#d91a1a}-2.23\%$
test_a2c_speed[True-backward] 11.7534ms 11.3475ms 88.1249 Ops/s 92.5562 Ops/s $\color{#d91a1a}-4.79\%$
test_a2c_speed[reduce-overhead-None] 5.1030ms 4.3362ms 230.6160 Ops/s 235.4112 Ops/s $\color{#d91a1a}-2.04\%$
test_a2c_speed[reduce-overhead-backward] 11.3691ms 10.7809ms 92.7565 Ops/s 91.9775 Ops/s $\color{#35bf28}+0.85\%$
test_ppo_speed[False-None] 8.9911ms 7.3963ms 135.2029 Ops/s 136.0476 Ops/s $\color{#d91a1a}-0.62\%$
test_ppo_speed[False-backward] 17.0326ms 14.7937ms 67.5965 Ops/s 65.6564 Ops/s $\color{#35bf28}+2.95\%$
test_ppo_speed[True-None] 4.0405ms 3.7862ms 264.1142 Ops/s 267.4354 Ops/s $\color{#d91a1a}-1.24\%$
test_ppo_speed[True-backward] 10.3162ms 9.9202ms 100.8047 Ops/s 103.5780 Ops/s $\color{#d91a1a}-2.68\%$
test_ppo_speed[reduce-overhead-None] 4.1868ms 3.7805ms 264.5147 Ops/s 265.4255 Ops/s $\color{#d91a1a}-0.34\%$
test_ppo_speed[reduce-overhead-backward] 11.1581ms 10.1163ms 98.8506 Ops/s 103.4788 Ops/s $\color{#d91a1a}-4.47\%$
test_reinforce_speed[False-None] 8.0409ms 6.6558ms 150.2449 Ops/s 154.5931 Ops/s $\color{#d91a1a}-2.81\%$
test_reinforce_speed[False-backward] 11.8140ms 10.1612ms 98.4138 Ops/s 102.0180 Ops/s $\color{#d91a1a}-3.53\%$
test_reinforce_speed[True-None] 3.1170ms 2.7855ms 359.0057 Ops/s 369.4245 Ops/s $\color{#d91a1a}-2.82\%$
test_reinforce_speed[True-backward] 9.5119ms 8.7530ms 114.2461 Ops/s 115.8300 Ops/s $\color{#d91a1a}-1.37\%$
test_reinforce_speed[reduce-overhead-None] 3.2776ms 2.8126ms 355.5455 Ops/s 361.6578 Ops/s $\color{#d91a1a}-1.69\%$
test_reinforce_speed[reduce-overhead-backward] 10.9874ms 9.5336ms 104.8919 Ops/s 115.2090 Ops/s $\textbf{\color{#d91a1a}-8.96\%}$
test_iql_speed[False-None] 35.7424ms 33.6668ms 29.7029 Ops/s 31.3130 Ops/s $\textbf{\color{#d91a1a}-5.14\%}$
test_iql_speed[False-backward] 50.6110ms 47.1961ms 21.1882 Ops/s 22.1492 Ops/s $\color{#d91a1a}-4.34\%$
test_iql_speed[True-None] 12.7473ms 11.0973ms 90.1123 Ops/s 93.3831 Ops/s $\color{#d91a1a}-3.50\%$
test_iql_speed[True-backward] 23.8819ms 22.4649ms 44.5139 Ops/s 46.0930 Ops/s $\color{#d91a1a}-3.43\%$
test_iql_speed[reduce-overhead-None] 12.3704ms 11.1034ms 90.0623 Ops/s 92.0855 Ops/s $\color{#d91a1a}-2.20\%$
test_iql_speed[reduce-overhead-backward] 22.9632ms 22.0940ms 45.2611 Ops/s 46.4766 Ops/s $\color{#d91a1a}-2.62\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.8025ms 5.1559ms 193.9521 Ops/s 207.4989 Ops/s $\textbf{\color{#d91a1a}-6.53\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.3619ms 0.5316ms 1.8812 KOps/s 1.9563 KOps/s $\color{#d91a1a}-3.84\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.8098ms 0.5065ms 1.9742 KOps/s 2.0450 KOps/s $\color{#d91a1a}-3.46\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 7.2366ms 4.5989ms 217.4425 Ops/s 218.4537 Ops/s $\color{#d91a1a}-0.46\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.6487ms 0.5035ms 1.9862 KOps/s 1.9948 KOps/s $\color{#d91a1a}-0.43\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.7686ms 0.4820ms 2.0745 KOps/s 2.0921 KOps/s $\color{#d91a1a}-0.84\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 2.2048ms 1.6465ms 607.3399 Ops/s 610.1245 Ops/s $\color{#d91a1a}-0.46\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 2.0756ms 1.5970ms 626.1747 Ops/s 631.8704 Ops/s $\color{#d91a1a}-0.90\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 7.6431ms 4.7437ms 210.8045 Ops/s 210.6768 Ops/s $\color{#35bf28}+0.06\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 0.9745ms 0.6506ms 1.5371 KOps/s 1.5515 KOps/s $\color{#d91a1a}-0.93\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.8539ms 0.6263ms 1.5968 KOps/s 1.6234 KOps/s $\color{#d91a1a}-1.64\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 4.9350ms 4.6185ms 216.5189 Ops/s 220.9100 Ops/s $\color{#d91a1a}-1.99\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.1066ms 0.5196ms 1.9246 KOps/s 1.9369 KOps/s $\color{#d91a1a}-0.63\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6743ms 0.4973ms 2.0107 KOps/s 1.9308 KOps/s $\color{#35bf28}+4.14\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 7.6864ms 4.6481ms 215.1420 Ops/s 216.9720 Ops/s $\color{#d91a1a}-0.84\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.6074ms 0.5094ms 1.9632 KOps/s 1.9944 KOps/s $\color{#d91a1a}-1.56\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.7295ms 0.4808ms 2.0798 KOps/s 2.1008 KOps/s $\color{#d91a1a}-1.00\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 5.4930ms 4.7418ms 210.8907 Ops/s 207.2524 Ops/s $\color{#35bf28}+1.76\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.6710ms 0.6492ms 1.5404 KOps/s 1.5522 KOps/s $\color{#d91a1a}-0.76\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.8486ms 0.6290ms 1.5899 KOps/s 1.6025 KOps/s $\color{#d91a1a}-0.79\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.4614s 13.3242ms 75.0515 Ops/s 243.9727 Ops/s $\textbf{\color{#d91a1a}-69.24\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 8.6947ms 2.3146ms 432.0437 Ops/s 457.0120 Ops/s $\textbf{\color{#d91a1a}-5.46\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 3.8515ms 1.2851ms 778.1223 Ops/s 775.4082 Ops/s $\color{#35bf28}+0.35\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 5.5243ms 4.2051ms 237.8073 Ops/s 243.8782 Ops/s $\color{#d91a1a}-2.49\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 4.9762ms 2.2578ms 442.9091 Ops/s 434.3008 Ops/s $\color{#35bf28}+1.98\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 7.6337ms 1.3532ms 738.9987 Ops/s 721.1380 Ops/s $\color{#35bf28}+2.48\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 5.8112ms 4.2815ms 233.5637 Ops/s 252.4855 Ops/s $\textbf{\color{#d91a1a}-7.49\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 0.4331s 10.9533ms 91.2965 Ops/s 405.4669 Ops/s $\textbf{\color{#d91a1a}-77.48\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 7.5930ms 1.5223ms 656.9217 Ops/s 694.0911 Ops/s $\textbf{\color{#d91a1a}-5.36\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 11.5171ms 11.1164ms 89.9575 Ops/s 89.8797 Ops/s $\color{#35bf28}+0.09\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 14.5143ms 14.2787ms 70.0343 Ops/s 70.1553 Ops/s $\color{#d91a1a}-0.17\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 20.9607ms 20.0096ms 49.9760 Ops/s 50.0114 Ops/s $\color{#d91a1a}-0.07\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 14.6378ms 14.4365ms 69.2687 Ops/s 69.6035 Ops/s $\color{#d91a1a}-0.48\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 22.7283ms 20.0663ms 49.8348 Ops/s 49.7505 Ops/s $\color{#35bf28}+0.17\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 15.9087ms 15.6827ms 63.7645 Ops/s 63.5518 Ops/s $\color{#35bf28}+0.33\%$

Copy link

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

Total Benchmarks: 149. Improved: $\large\color{#35bf28}17$. Worsened: $\large\color{#d91a1a}12$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_simple 0.7380s 0.7374s 1.3562 Ops/s 1.3463 Ops/s $\color{#35bf28}+0.73\%$
test_transformed 0.9782s 0.9778s 1.0227 Ops/s 1.0113 Ops/s $\color{#35bf28}+1.13\%$
test_serial 2.1161s 2.1128s 0.4733 Ops/s 0.4670 Ops/s $\color{#35bf28}+1.35\%$
test_parallel 2.0243s 1.9932s 0.5017 Ops/s 0.5020 Ops/s $\color{#d91a1a}-0.05\%$
test_step_mdp_speed[True-True-True-True-True] 0.1813ms 36.0971μs 27.7031 KOps/s 27.8781 KOps/s $\color{#d91a1a}-0.63\%$
test_step_mdp_speed[True-True-True-True-False] 58.9200μs 20.4985μs 48.7841 KOps/s 48.2908 KOps/s $\color{#35bf28}+1.02\%$
test_step_mdp_speed[True-True-True-False-True] 55.5810μs 19.8400μs 50.4033 KOps/s 49.3005 KOps/s $\color{#35bf28}+2.24\%$
test_step_mdp_speed[True-True-True-False-False] 41.3400μs 11.5422μs 86.6389 KOps/s 86.3022 KOps/s $\color{#35bf28}+0.39\%$
test_step_mdp_speed[True-True-False-True-True] 78.0210μs 37.9443μs 26.3544 KOps/s 25.9919 KOps/s $\color{#35bf28}+1.39\%$
test_step_mdp_speed[True-True-False-True-False] 54.0100μs 22.4789μs 44.4861 KOps/s 44.5591 KOps/s $\color{#d91a1a}-0.16\%$
test_step_mdp_speed[True-True-False-False-True] 54.1010μs 21.9008μs 45.6604 KOps/s 43.6487 KOps/s $\color{#35bf28}+4.61\%$
test_step_mdp_speed[True-True-False-False-False] 49.0500μs 13.7595μs 72.6770 KOps/s 73.0828 KOps/s $\color{#d91a1a}-0.56\%$
test_step_mdp_speed[True-False-True-True-True] 71.9100μs 40.2596μs 24.8388 KOps/s 24.4858 KOps/s $\color{#35bf28}+1.44\%$
test_step_mdp_speed[True-False-True-True-False] 61.8010μs 24.6974μs 40.4900 KOps/s 41.7405 KOps/s $\color{#d91a1a}-3.00\%$
test_step_mdp_speed[True-False-True-False-True] 65.3110μs 21.7717μs 45.9312 KOps/s 44.5905 KOps/s $\color{#35bf28}+3.01\%$
test_step_mdp_speed[True-False-True-False-False] 40.9800μs 13.6307μs 73.3638 KOps/s 72.1568 KOps/s $\color{#35bf28}+1.67\%$
test_step_mdp_speed[True-False-False-True-True] 75.7910μs 41.7043μs 23.9783 KOps/s 23.5656 KOps/s $\color{#35bf28}+1.75\%$
test_step_mdp_speed[True-False-False-True-False] 61.6510μs 26.3010μs 38.0214 KOps/s 37.5502 KOps/s $\color{#35bf28}+1.25\%$
test_step_mdp_speed[True-False-False-False-True] 70.6100μs 23.7941μs 42.0272 KOps/s 40.9657 KOps/s $\color{#35bf28}+2.59\%$
test_step_mdp_speed[True-False-False-False-False] 50.6200μs 15.5793μs 64.1877 KOps/s 62.9623 KOps/s $\color{#35bf28}+1.95\%$
test_step_mdp_speed[False-True-True-True-True] 86.5610μs 40.1529μs 24.9048 KOps/s 24.5268 KOps/s $\color{#35bf28}+1.54\%$
test_step_mdp_speed[False-True-True-True-False] 55.4300μs 24.7463μs 40.4101 KOps/s 40.2668 KOps/s $\color{#35bf28}+0.36\%$
test_step_mdp_speed[False-True-True-False-True] 79.6210μs 24.9411μs 40.0945 KOps/s 38.4460 KOps/s $\color{#35bf28}+4.29\%$
test_step_mdp_speed[False-True-True-False-False] 0.3795ms 15.2536μs 65.5584 KOps/s 64.0469 KOps/s $\color{#35bf28}+2.36\%$
test_step_mdp_speed[False-True-False-True-True] 0.4133ms 42.4059μs 23.5817 KOps/s 23.2473 KOps/s $\color{#35bf28}+1.44\%$
test_step_mdp_speed[False-True-False-True-False] 0.4064ms 26.5022μs 37.7327 KOps/s 37.9886 KOps/s $\color{#d91a1a}-0.67\%$
test_step_mdp_speed[False-True-False-False-True] 3.3980ms 27.3242μs 36.5975 KOps/s 36.0442 KOps/s $\color{#35bf28}+1.54\%$
test_step_mdp_speed[False-True-False-False-False] 0.4058ms 17.2955μs 57.8185 KOps/s 57.1351 KOps/s $\color{#35bf28}+1.20\%$
test_step_mdp_speed[False-False-True-True-True] 0.1009ms 44.4631μs 22.4906 KOps/s 22.6690 KOps/s $\color{#d91a1a}-0.79\%$
test_step_mdp_speed[False-False-True-True-False] 0.4059ms 28.8714μs 34.6363 KOps/s 35.0274 KOps/s $\color{#d91a1a}-1.12\%$
test_step_mdp_speed[False-False-True-False-True] 88.9610μs 27.7301μs 36.0619 KOps/s 36.2198 KOps/s $\color{#d91a1a}-0.44\%$
test_step_mdp_speed[False-False-True-False-False] 0.4005ms 17.2843μs 57.8559 KOps/s 58.1585 KOps/s $\color{#d91a1a}-0.52\%$
test_step_mdp_speed[False-False-False-True-True] 0.4271ms 45.8427μs 21.8137 KOps/s 21.8031 KOps/s $\color{#35bf28}+0.05\%$
test_step_mdp_speed[False-False-False-True-False] 0.4147ms 30.6096μs 32.6695 KOps/s 33.0323 KOps/s $\color{#d91a1a}-1.10\%$
test_step_mdp_speed[False-False-False-False-True] 0.4234ms 29.1596μs 34.2940 KOps/s 33.9903 KOps/s $\color{#35bf28}+0.89\%$
test_step_mdp_speed[False-False-False-False-False] 41.6410μs 18.9258μs 52.8379 KOps/s 51.9185 KOps/s $\color{#35bf28}+1.77\%$
test_values[generalized_advantage_estimate-True-True] 25.8137ms 25.2641ms 39.5819 Ops/s 38.3337 Ops/s $\color{#35bf28}+3.26\%$
test_values[vec_generalized_advantage_estimate-True-True] 0.1158s 3.2147ms 311.0700 Ops/s 334.8520 Ops/s $\textbf{\color{#d91a1a}-7.10\%}$
test_values[td0_return_estimate-False-False] 0.1039ms 79.6256μs 12.5588 KOps/s 12.5137 KOps/s $\color{#35bf28}+0.36\%$
test_values[td1_return_estimate-False-False] 60.0162ms 57.2180ms 17.4770 Ops/s 17.2371 Ops/s $\color{#35bf28}+1.39\%$
test_values[vec_td1_return_estimate-False-False] 1.3034ms 1.0860ms 920.8441 Ops/s 914.2971 Ops/s $\color{#35bf28}+0.72\%$
test_values[td_lambda_return_estimate-True-False] 94.5410ms 89.4112ms 11.1843 Ops/s 10.9108 Ops/s $\color{#35bf28}+2.51\%$
test_values[vec_td_lambda_return_estimate-True-False] 1.2076ms 1.0789ms 926.8500 Ops/s 911.4201 Ops/s $\color{#35bf28}+1.69\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 27.0295ms 26.2491ms 38.0965 Ops/s 38.9483 Ops/s $\color{#d91a1a}-2.19\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.0622ms 0.7528ms 1.3284 KOps/s 1.3177 KOps/s $\color{#35bf28}+0.81\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 1.0516ms 0.6688ms 1.4951 KOps/s 1.4784 KOps/s $\color{#35bf28}+1.13\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 1.5533ms 1.4891ms 671.5383 Ops/s 671.8323 Ops/s $\color{#d91a1a}-0.04\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 1.0848ms 0.6805ms 1.4695 KOps/s 1.4380 KOps/s $\color{#35bf28}+2.20\%$
test_dqn_speed[False-None] 7.2663ms 1.4246ms 701.9654 Ops/s 690.8268 Ops/s $\color{#35bf28}+1.61\%$
test_dqn_speed[False-backward] 2.0472ms 2.0112ms 497.2122 Ops/s 484.4018 Ops/s $\color{#35bf28}+2.64\%$
test_dqn_speed[True-None] 0.6606ms 0.5327ms 1.8771 KOps/s 1.8584 KOps/s $\color{#35bf28}+1.01\%$
test_dqn_speed[True-backward] 1.3592ms 1.1966ms 835.7125 Ops/s 897.4504 Ops/s $\textbf{\color{#d91a1a}-6.88\%}$
test_dqn_speed[reduce-overhead-None] 0.6566ms 0.5488ms 1.8221 KOps/s 1.7938 KOps/s $\color{#35bf28}+1.57\%$
test_dqn_speed[reduce-overhead-backward] 1.0902ms 1.0583ms 944.8953 Ops/s 932.7837 Ops/s $\color{#35bf28}+1.30\%$
test_ddpg_speed[False-None] 3.0439ms 2.6838ms 372.6018 Ops/s 365.7176 Ops/s $\color{#35bf28}+1.88\%$
test_ddpg_speed[False-backward] 4.4738ms 4.0210ms 248.6923 Ops/s 239.6718 Ops/s $\color{#35bf28}+3.76\%$
test_ddpg_speed[True-None] 1.2286ms 1.0825ms 923.7847 Ops/s 914.7274 Ops/s $\color{#35bf28}+0.99\%$
test_ddpg_speed[True-backward] 2.3550ms 2.2953ms 435.6754 Ops/s 453.9386 Ops/s $\color{#d91a1a}-4.02\%$
test_ddpg_speed[reduce-overhead-None] 1.4890ms 1.0911ms 916.4774 Ops/s 899.2311 Ops/s $\color{#35bf28}+1.92\%$
test_ddpg_speed[reduce-overhead-backward] 1.8066ms 1.7603ms 568.0836 Ops/s 604.2651 Ops/s $\textbf{\color{#d91a1a}-5.99\%}$
test_sac_speed[False-None] 8.1835ms 7.6557ms 130.6215 Ops/s 127.1307 Ops/s $\color{#35bf28}+2.75\%$
test_sac_speed[False-backward] 11.1965ms 10.7818ms 92.7491 Ops/s 92.3553 Ops/s $\color{#35bf28}+0.43\%$
test_sac_speed[True-None] 1.9350ms 1.5351ms 651.4396 Ops/s 631.1361 Ops/s $\color{#35bf28}+3.22\%$
test_sac_speed[True-backward] 3.4724ms 3.3384ms 299.5412 Ops/s 292.7618 Ops/s $\color{#35bf28}+2.32\%$
test_sac_speed[reduce-overhead-None] 22.4880ms 12.4817ms 80.1174 Ops/s 79.3795 Ops/s $\color{#35bf28}+0.93\%$
test_sac_speed[reduce-overhead-backward] 1.5540ms 1.4826ms 674.4729 Ops/s 670.0496 Ops/s $\color{#35bf28}+0.66\%$
test_redq_speed[False-None] 8.0511ms 7.2401ms 138.1192 Ops/s 133.2116 Ops/s $\color{#35bf28}+3.68\%$
test_redq_speed[False-backward] 12.0200ms 11.3139ms 88.3872 Ops/s 86.1146 Ops/s $\color{#35bf28}+2.64\%$
test_redq_speed[True-None] 2.3564ms 1.9762ms 506.0310 Ops/s 486.5661 Ops/s $\color{#35bf28}+4.00\%$
test_redq_speed[True-backward] 4.2721ms 3.8084ms 262.5783 Ops/s 273.4583 Ops/s $\color{#d91a1a}-3.98\%$
test_redq_speed[reduce-overhead-None] 2.4548ms 1.9742ms 506.5312 Ops/s 500.2546 Ops/s $\color{#35bf28}+1.25\%$
test_redq_speed[reduce-overhead-backward] 3.9079ms 3.7905ms 263.8203 Ops/s 270.9643 Ops/s $\color{#d91a1a}-2.64\%$
test_redq_deprec_speed[False-None] 9.1300ms 8.6062ms 116.1960 Ops/s 112.4276 Ops/s $\color{#35bf28}+3.35\%$
test_redq_deprec_speed[False-backward] 12.3215ms 11.7895ms 84.8212 Ops/s 84.3368 Ops/s $\color{#35bf28}+0.57\%$
test_redq_deprec_speed[True-None] 2.8473ms 2.4121ms 414.5689 Ops/s 412.5753 Ops/s $\color{#35bf28}+0.48\%$
test_redq_deprec_speed[True-backward] 4.6067ms 4.1531ms 240.7824 Ops/s 235.6870 Ops/s $\color{#35bf28}+2.16\%$
test_redq_deprec_speed[reduce-overhead-None] 2.7187ms 2.3095ms 432.9923 Ops/s 424.3613 Ops/s $\color{#35bf28}+2.03\%$
test_redq_deprec_speed[reduce-overhead-backward] 4.3900ms 3.9300ms 254.4529 Ops/s 238.5572 Ops/s $\textbf{\color{#35bf28}+6.66\%}$
test_td3_speed[False-None] 7.6765ms 7.4945ms 133.4313 Ops/s 127.8086 Ops/s $\color{#35bf28}+4.40\%$
test_td3_speed[False-backward] 10.6604ms 9.8799ms 101.2155 Ops/s 97.9309 Ops/s $\color{#35bf28}+3.35\%$
test_td3_speed[True-None] 1.6146ms 1.5624ms 640.0494 Ops/s 627.2109 Ops/s $\color{#35bf28}+2.05\%$
test_td3_speed[True-backward] 3.1227ms 3.0606ms 326.7321 Ops/s 304.4941 Ops/s $\textbf{\color{#35bf28}+7.30\%}$
test_td3_speed[reduce-overhead-None] 79.8632ms 25.4709ms 39.2604 Ops/s 38.0353 Ops/s $\color{#35bf28}+3.22\%$
test_td3_speed[reduce-overhead-backward] 1.3352ms 1.2694ms 787.7669 Ops/s 700.2496 Ops/s $\textbf{\color{#35bf28}+12.50\%}$
test_cql_speed[False-None] 15.8736ms 15.4704ms 64.6397 Ops/s 63.1556 Ops/s $\color{#35bf28}+2.35\%$
test_cql_speed[False-backward] 21.2380ms 20.4500ms 48.8998 Ops/s 46.8516 Ops/s $\color{#35bf28}+4.37\%$
test_cql_speed[True-None] 3.4947ms 2.9674ms 336.9941 Ops/s 341.4058 Ops/s $\color{#d91a1a}-1.29\%$
test_cql_speed[True-backward] 5.5597ms 5.0687ms 197.2910 Ops/s 196.5824 Ops/s $\color{#35bf28}+0.36\%$
test_cql_speed[reduce-overhead-None] 21.5581ms 13.1044ms 76.3100 Ops/s 75.7698 Ops/s $\color{#35bf28}+0.71\%$
test_cql_speed[reduce-overhead-backward] 1.6022ms 1.5284ms 654.2982 Ops/s 637.9776 Ops/s $\color{#35bf28}+2.56\%$
test_a2c_speed[False-None] 3.3268ms 3.1421ms 318.2547 Ops/s 314.7998 Ops/s $\color{#35bf28}+1.10\%$
test_a2c_speed[False-backward] 6.8646ms 6.1842ms 161.7019 Ops/s 159.2138 Ops/s $\color{#35bf28}+1.56\%$
test_a2c_speed[True-None] 1.0960ms 1.0181ms 982.2024 Ops/s 979.4824 Ops/s $\color{#35bf28}+0.28\%$
test_a2c_speed[True-backward] 2.8767ms 2.7822ms 359.4334 Ops/s 379.6106 Ops/s $\textbf{\color{#d91a1a}-5.32\%}$
test_a2c_speed[reduce-overhead-None] 0.3784s 12.3769ms 80.7954 Ops/s 86.6484 Ops/s $\textbf{\color{#d91a1a}-6.75\%}$
test_a2c_speed[reduce-overhead-backward] 1.1240ms 1.0764ms 929.0035 Ops/s 880.9754 Ops/s $\textbf{\color{#35bf28}+5.45\%}$
test_ppo_speed[False-None] 3.6270ms 3.5245ms 283.7311 Ops/s 275.3185 Ops/s $\color{#35bf28}+3.06\%$
test_ppo_speed[False-backward] 6.9666ms 6.8486ms 146.0146 Ops/s 143.6691 Ops/s $\color{#35bf28}+1.63\%$
test_ppo_speed[True-None] 1.0206ms 0.9378ms 1.0664 KOps/s 1.0231 KOps/s $\color{#35bf28}+4.23\%$
test_ppo_speed[True-backward] 2.9096ms 2.6295ms 380.3022 Ops/s 383.6143 Ops/s $\color{#d91a1a}-0.86\%$
test_ppo_speed[reduce-overhead-None] 0.5636ms 0.4892ms 2.0443 KOps/s 1.9063 KOps/s $\textbf{\color{#35bf28}+7.24\%}$
test_ppo_speed[reduce-overhead-backward] 1.0010ms 0.9579ms 1.0440 KOps/s 1.0031 KOps/s $\color{#35bf28}+4.08\%$
test_reinforce_speed[False-None] 2.2832ms 2.1232ms 470.9882 Ops/s 437.0148 Ops/s $\textbf{\color{#35bf28}+7.77\%}$
test_reinforce_speed[False-backward] 3.5628ms 3.1144ms 321.0873 Ops/s 309.0531 Ops/s $\color{#35bf28}+3.89\%$
test_reinforce_speed[True-None] 0.9672ms 0.8413ms 1.1886 KOps/s 1.2003 KOps/s $\color{#d91a1a}-0.98\%$
test_reinforce_speed[True-backward] 2.8627ms 2.4881ms 401.9126 Ops/s 407.9658 Ops/s $\color{#d91a1a}-1.48\%$
test_reinforce_speed[reduce-overhead-None] 22.5023ms 11.9119ms 83.9496 Ops/s 86.4933 Ops/s $\color{#d91a1a}-2.94\%$
test_reinforce_speed[reduce-overhead-backward] 1.1293ms 1.0645ms 939.4233 Ops/s 949.4683 Ops/s $\color{#d91a1a}-1.06\%$
test_iql_speed[False-None] 9.2951ms 8.8856ms 112.5420 Ops/s 112.2473 Ops/s $\color{#35bf28}+0.26\%$
test_iql_speed[False-backward] 12.9886ms 12.4272ms 80.4688 Ops/s 79.6227 Ops/s $\color{#35bf28}+1.06\%$
test_iql_speed[True-None] 1.9575ms 1.8315ms 546.0145 Ops/s 573.2263 Ops/s $\color{#d91a1a}-4.75\%$
test_iql_speed[True-backward] 4.6617ms 4.2807ms 233.6079 Ops/s 234.5353 Ops/s $\color{#d91a1a}-0.40\%$
test_iql_speed[reduce-overhead-None] 20.2035ms 11.4730ms 87.1609 Ops/s 87.3470 Ops/s $\color{#d91a1a}-0.21\%$
test_iql_speed[reduce-overhead-backward] 1.4452ms 1.4068ms 710.8577 Ops/s 711.4838 Ops/s $\color{#d91a1a}-0.09\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.8809ms 6.2827ms 159.1671 Ops/s 157.1614 Ops/s $\color{#35bf28}+1.28\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.5049ms 0.3087ms 3.2391 KOps/s 3.0330 KOps/s $\textbf{\color{#35bf28}+6.80\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.4757ms 0.2823ms 3.5428 KOps/s 3.0823 KOps/s $\textbf{\color{#35bf28}+14.94\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.2980ms 6.0736ms 164.6482 Ops/s 163.1222 Ops/s $\color{#35bf28}+0.94\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.8364ms 0.2793ms 3.5804 KOps/s 3.2899 KOps/s $\textbf{\color{#35bf28}+8.83\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4843ms 0.2697ms 3.7080 KOps/s 3.4777 KOps/s $\textbf{\color{#35bf28}+6.62\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.7391ms 1.3182ms 758.6254 Ops/s 703.9203 Ops/s $\textbf{\color{#35bf28}+7.77\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.6028ms 1.2066ms 828.8068 Ops/s 743.3454 Ops/s $\textbf{\color{#35bf28}+11.50\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.4148ms 6.2480ms 160.0513 Ops/s 160.1358 Ops/s $\color{#d91a1a}-0.05\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.1075ms 0.4259ms 2.3481 KOps/s 2.4015 KOps/s $\color{#d91a1a}-2.22\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7749ms 0.4635ms 2.1574 KOps/s 2.5390 KOps/s $\textbf{\color{#d91a1a}-15.03\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.2275ms 6.0952ms 164.0629 Ops/s 163.6577 Ops/s $\color{#35bf28}+0.25\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.9848ms 0.3194ms 3.1312 KOps/s 3.3868 KOps/s $\textbf{\color{#d91a1a}-7.55\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6510ms 0.3300ms 3.0301 KOps/s 3.2708 KOps/s $\textbf{\color{#d91a1a}-7.36\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.3592ms 6.0331ms 165.7520 Ops/s 164.6374 Ops/s $\color{#35bf28}+0.68\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 1.4344ms 0.2742ms 3.6464 KOps/s 3.4071 KOps/s $\textbf{\color{#35bf28}+7.02\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4947ms 0.2707ms 3.6938 KOps/s 3.4403 KOps/s $\textbf{\color{#35bf28}+7.37\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.3582ms 6.2641ms 159.6393 Ops/s 160.7577 Ops/s $\color{#d91a1a}-0.70\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.2052ms 0.4539ms 2.2033 KOps/s 2.2346 KOps/s $\color{#d91a1a}-1.40\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6716ms 0.4502ms 2.2210 KOps/s 2.2349 KOps/s $\color{#d91a1a}-0.62\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 6.9214ms 5.3075ms 188.4110 Ops/s 185.9769 Ops/s $\color{#35bf28}+1.31\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 10.0801ms 2.0525ms 487.2091 Ops/s 438.3958 Ops/s $\textbf{\color{#35bf28}+11.13\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 6.6921ms 1.1578ms 863.6771 Ops/s 835.3136 Ops/s $\color{#35bf28}+3.40\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.4876s 14.9514ms 66.8832 Ops/s 186.8608 Ops/s $\textbf{\color{#d91a1a}-64.21\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 6.0170ms 1.9859ms 503.5553 Ops/s 439.6711 Ops/s $\textbf{\color{#35bf28}+14.53\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 9.8637ms 1.2449ms 803.2665 Ops/s 890.6433 Ops/s $\textbf{\color{#d91a1a}-9.81\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 9.1748ms 5.6112ms 178.2135 Ops/s 33.5562 Ops/s $\textbf{\color{#35bf28}+431.09\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 8.4247ms 2.2030ms 453.9357 Ops/s 501.2045 Ops/s $\textbf{\color{#d91a1a}-9.43\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 8.4581ms 1.3817ms 723.7361 Ops/s 828.3150 Ops/s $\textbf{\color{#d91a1a}-12.63\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 13.4502ms 13.0534ms 76.6081 Ops/s 75.0213 Ops/s $\color{#35bf28}+2.12\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 16.9828ms 16.5100ms 60.5695 Ops/s 58.3574 Ops/s $\color{#35bf28}+3.79\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 17.8350ms 17.4149ms 57.4221 Ops/s 54.9241 Ops/s $\color{#35bf28}+4.55\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 17.2372ms 16.7720ms 59.6233 Ops/s 59.0975 Ops/s $\color{#35bf28}+0.89\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 17.5698ms 17.3467ms 57.6479 Ops/s 55.4323 Ops/s $\color{#35bf28}+4.00\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 18.8832ms 18.0876ms 55.2866 Ops/s 55.6732 Ops/s $\color{#d91a1a}-0.69\%$

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.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants