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.

Sign up for GitHub

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

Jump to bottom

[Feature] Send info dict to the storage device in RBs #2527

Merged
merged 1 commit into from
Oct 29, 2024
Merged

[Feature] Send info dict to the storage device in RBs #2527

merged 1 commit into from
Oct 29, 2024

Conversation

vmoens
Copy link
Contributor

@vmoens vmoens commented Oct 29, 2024 •
edited
Loading

@vmoens
Update
9534be2 
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Oct 29, 2024
@vmoens
[Feature] Send info dict to the storage device in RBs
304e802 
ghstack-source-id: 4ed60d649b17f96b49f90d234e679937c60a3c32
Pull Request resolved: #2527
@pytorch-bot PyTorch Bot
Copy link

pytorch-bot bot commented Oct 29, 2024 •
edited
Loading

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/2527

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

❌ 1 New Failure, 5 Unrelated Failures

As of commit 9534be2 with merge base 3e4b292 (image):

NEW FAILURE - The following job has failed:

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.

@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 29, 2024
@vmoens vmoens mentioned this pull request Oct 29, 2024
Open
3 tasks
@github-actions GitHub Actions
Copy link

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

Total Benchmarks: 145. Improved: $\large\color{#35bf28}13$. Worsened: $\large\color{#d91a1a}3$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_simple 0.4109s 0.4092s 2.4440 Ops/s 2.3117 Ops/s $\textbf{\color{#35bf28}+5.72\%}$
test_transformed 0.6724s 0.5957s 1.6788 Ops/s 1.6480 Ops/s $\color{#35bf28}+1.87\%$
test_serial 1.3884s 1.3222s 0.7563 Ops/s 0.7611 Ops/s $\color{#d91a1a}-0.63\%$
test_parallel 1.2473s 1.2413s 0.8056 Ops/s 0.7809 Ops/s $\color{#35bf28}+3.16\%$
test_step_mdp_speed[True-True-True-True-True] 0.2197ms 26.6928μs 37.4633 KOps/s 36.0256 KOps/s $\color{#35bf28}+3.99\%$
test_step_mdp_speed[True-True-True-True-False] 48.6410μs 15.4414μs 64.7611 KOps/s 63.8274 KOps/s $\color{#35bf28}+1.46\%$
test_step_mdp_speed[True-True-True-False-True] 0.1317ms 15.3338μs 65.2153 KOps/s 63.2694 KOps/s $\color{#35bf28}+3.08\%$
test_step_mdp_speed[True-True-True-False-False] 38.4320μs 8.7807μs 113.8866 KOps/s 110.3056 KOps/s $\color{#35bf28}+3.25\%$
test_step_mdp_speed[True-True-False-True-True] 62.8170μs 28.5839μs 34.9847 KOps/s 33.6793 KOps/s $\color{#35bf28}+3.88\%$
test_step_mdp_speed[True-True-False-True-False] 40.5960μs 17.3194μs 57.7386 KOps/s 57.0413 KOps/s $\color{#35bf28}+1.22\%$
test_step_mdp_speed[True-True-False-False-True] 51.5160μs 17.0536μs 58.6385 KOps/s 57.1593 KOps/s $\color{#35bf28}+2.59\%$
test_step_mdp_speed[True-True-False-False-False] 35.9370μs 10.4647μs 95.5596 KOps/s 92.5107 KOps/s $\color{#35bf28}+3.30\%$
test_step_mdp_speed[True-False-True-True-True] 82.8450μs 30.3996μs 32.8952 KOps/s 32.0770 KOps/s $\color{#35bf28}+2.55\%$
test_step_mdp_speed[True-False-True-True-False] 72.8670μs 18.9137μs 52.8719 KOps/s 51.7607 KOps/s $\color{#35bf28}+2.15\%$
test_step_mdp_speed[True-False-True-False-True] 45.8660μs 17.0149μs 58.7720 KOps/s 56.9593 KOps/s $\color{#35bf28}+3.18\%$
test_step_mdp_speed[True-False-True-False-False] 37.1590μs 10.4524μs 95.6720 KOps/s 94.4845 KOps/s $\color{#35bf28}+1.26\%$
test_step_mdp_speed[True-False-False-True-True] 0.1008ms 31.6887μs 31.5570 KOps/s 30.7303 KOps/s $\color{#35bf28}+2.69\%$
test_step_mdp_speed[True-False-False-True-False] 67.9670μs 20.4878μs 48.8096 KOps/s 48.1721 KOps/s $\color{#35bf28}+1.32\%$
test_step_mdp_speed[True-False-False-False-True] 48.0200μs 18.2689μs 54.7378 KOps/s 52.3962 KOps/s $\color{#35bf28}+4.47\%$
test_step_mdp_speed[True-False-False-False-False] 40.0350μs 12.0101μs 83.2632 KOps/s 81.7497 KOps/s $\color{#35bf28}+1.85\%$
test_step_mdp_speed[False-True-True-True-True] 67.7970μs 30.2176μs 33.0933 KOps/s 32.5358 KOps/s $\color{#35bf28}+1.71\%$
test_step_mdp_speed[False-True-True-True-False] 82.5510μs 18.6930μs 53.4959 KOps/s 52.2024 KOps/s $\color{#35bf28}+2.48\%$
test_step_mdp_speed[False-True-True-False-True] 39.7750μs 19.2550μs 51.9345 KOps/s 50.4654 KOps/s $\color{#35bf28}+2.91\%$
test_step_mdp_speed[False-True-True-False-False] 84.1560μs 11.6192μs 86.0642 KOps/s 83.4259 KOps/s $\color{#35bf28}+3.16\%$
test_step_mdp_speed[False-True-False-True-True] 96.1490μs 31.5229μs 31.7230 KOps/s 30.2340 KOps/s $\color{#35bf28}+4.92\%$
test_step_mdp_speed[False-True-False-True-False] 61.7850μs 20.2417μs 49.4029 KOps/s 48.3864 KOps/s $\color{#35bf28}+2.10\%$
test_step_mdp_speed[False-True-False-False-True] 2.7618ms 20.7385μs 48.2195 KOps/s 47.0076 KOps/s $\color{#35bf28}+2.58\%$
test_step_mdp_speed[False-True-False-False-False] 46.7870μs 13.1551μs 76.0160 KOps/s 74.5223 KOps/s $\color{#35bf28}+2.00\%$
test_step_mdp_speed[False-False-True-True-True] 91.4930μs 33.2585μs 30.0675 KOps/s 29.3012 KOps/s $\color{#35bf28}+2.62\%$
test_step_mdp_speed[False-False-True-True-False] 47.9500μs 22.0456μs 45.3604 KOps/s 44.6418 KOps/s $\color{#35bf28}+1.61\%$
test_step_mdp_speed[False-False-True-False-True] 56.1250μs 20.5907μs 48.5657 KOps/s 46.6896 KOps/s $\color{#35bf28}+4.02\%$
test_step_mdp_speed[False-False-True-False-False] 59.8100μs 12.8962μs 77.5424 KOps/s 73.4317 KOps/s $\textbf{\color{#35bf28}+5.60\%}$
test_step_mdp_speed[False-False-False-True-True] 80.6330μs 34.4587μs 29.0202 KOps/s 28.0006 KOps/s $\color{#35bf28}+3.64\%$
test_step_mdp_speed[False-False-False-True-False] 58.0690μs 23.4877μs 42.5754 KOps/s 41.7663 KOps/s $\color{#35bf28}+1.94\%$
test_step_mdp_speed[False-False-False-False-True] 58.1790μs 22.1496μs 45.1475 KOps/s 43.8334 KOps/s $\color{#35bf28}+3.00\%$
test_step_mdp_speed[False-False-False-False-False] 57.0370μs 14.6668μs 68.1814 KOps/s 67.3915 KOps/s $\color{#35bf28}+1.17\%$
test_values[generalized_advantage_estimate-True-True] 9.8213ms 9.4617ms 105.6893 Ops/s 106.4141 Ops/s $\color{#d91a1a}-0.68\%$
test_values[vec_generalized_advantage_estimate-True-True] 37.4862ms 35.4438ms 28.2137 Ops/s 27.7855 Ops/s $\color{#35bf28}+1.54\%$
test_values[td0_return_estimate-False-False] 0.2342ms 0.1739ms 5.7500 KOps/s 5.7123 KOps/s $\color{#35bf28}+0.66\%$
test_values[td1_return_estimate-False-False] 25.5359ms 23.8254ms 41.9720 Ops/s 41.4845 Ops/s $\color{#35bf28}+1.18\%$
test_values[vec_td1_return_estimate-False-False] 37.6566ms 35.6873ms 28.0212 Ops/s 27.7114 Ops/s $\color{#35bf28}+1.12\%$
test_values[td_lambda_return_estimate-True-False] 35.2862ms 34.4722ms 29.0089 Ops/s 28.6611 Ops/s $\color{#35bf28}+1.21\%$
test_values[vec_td_lambda_return_estimate-True-False] 38.0934ms 35.5810ms 28.1049 Ops/s 27.9226 Ops/s $\color{#35bf28}+0.65\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 10.1926ms 8.3625ms 119.5812 Ops/s 119.8936 Ops/s $\color{#d91a1a}-0.26\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 2.1938ms 1.7818ms 561.2408 Ops/s 453.9889 Ops/s $\textbf{\color{#35bf28}+23.62\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.4615ms 0.3533ms 2.8304 KOps/s 2.7895 KOps/s $\color{#35bf28}+1.46\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 48.2477ms 46.8703ms 21.3355 Ops/s 20.2175 Ops/s $\textbf{\color{#35bf28}+5.53\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 3.9237ms 3.0301ms 330.0215 Ops/s 329.3062 Ops/s $\color{#35bf28}+0.22\%$
test_dqn_speed[False-None] 5.9455ms 1.3336ms 749.8661 Ops/s 742.3061 Ops/s $\color{#35bf28}+1.02\%$
test_dqn_speed[False-backward] 1.8926ms 1.8131ms 551.5366 Ops/s 549.2455 Ops/s $\color{#35bf28}+0.42\%$
test_dqn_speed[True-None] 1.2532ms 0.4644ms 2.1534 KOps/s 2.1643 KOps/s $\color{#d91a1a}-0.50\%$
test_dqn_speed[True-backward] 1.0050ms 0.8903ms 1.1233 KOps/s 1.1326 KOps/s $\color{#d91a1a}-0.82\%$
test_dqn_speed[reduce-overhead-None] 0.6050ms 0.4711ms 2.1228 KOps/s 2.1505 KOps/s $\color{#d91a1a}-1.29\%$
test_dqn_speed[reduce-overhead-backward] 0.9491ms 0.8857ms 1.1291 KOps/s 1.1288 KOps/s $\color{#35bf28}+0.03\%$
test_ddpg_speed[False-None] 3.6735ms 2.7638ms 361.8206 Ops/s 358.4483 Ops/s $\color{#35bf28}+0.94\%$
test_ddpg_speed[False-backward] 4.1167ms 3.9048ms 256.0966 Ops/s 254.8918 Ops/s $\color{#35bf28}+0.47\%$
test_ddpg_speed[True-None] 1.3838ms 1.0134ms 986.8030 Ops/s 982.7805 Ops/s $\color{#35bf28}+0.41\%$
test_ddpg_speed[True-backward] 1.9518ms 1.8916ms 528.6497 Ops/s 528.8361 Ops/s $\color{#d91a1a}-0.04\%$
test_ddpg_speed[reduce-overhead-None] 1.1626ms 1.0129ms 987.2166 Ops/s 988.7529 Ops/s $\color{#d91a1a}-0.16\%$
test_ddpg_speed[reduce-overhead-backward] 1.9366ms 1.8964ms 527.3169 Ops/s 527.1304 Ops/s $\color{#35bf28}+0.04\%$
test_sac_speed[False-None] 8.2383ms 7.8089ms 128.0590 Ops/s 126.9505 Ops/s $\color{#35bf28}+0.87\%$
test_sac_speed[False-backward] 12.2476ms 10.5819ms 94.5014 Ops/s 94.9021 Ops/s $\color{#d91a1a}-0.42\%$
test_sac_speed[True-None] 2.1423ms 1.8320ms 545.8461 Ops/s 538.3485 Ops/s $\color{#35bf28}+1.39\%$
test_sac_speed[True-backward] 3.6503ms 3.5228ms 283.8690 Ops/s 278.8408 Ops/s $\color{#35bf28}+1.80\%$
test_sac_speed[reduce-overhead-None] 2.2534ms 1.8305ms 546.3071 Ops/s 541.0455 Ops/s $\color{#35bf28}+0.97\%$
test_sac_speed[reduce-overhead-backward] 3.8110ms 3.5344ms 282.9311 Ops/s 280.6653 Ops/s $\color{#35bf28}+0.81\%$
test_redq_speed[False-None] 14.1837ms 12.4691ms 80.1984 Ops/s 71.5816 Ops/s $\textbf{\color{#35bf28}+12.04\%}$
test_redq_speed[False-backward] 23.5869ms 21.7221ms 46.0361 Ops/s 44.1702 Ops/s $\color{#35bf28}+4.22\%$
test_redq_speed[True-None] 5.7105ms 4.4638ms 224.0239 Ops/s 207.5802 Ops/s $\textbf{\color{#35bf28}+7.92\%}$
test_redq_speed[True-backward] 12.9022ms 11.7926ms 84.7992 Ops/s 79.8016 Ops/s $\textbf{\color{#35bf28}+6.26\%}$
test_redq_speed[reduce-overhead-None] 5.5038ms 4.4856ms 222.9352 Ops/s 199.2055 Ops/s $\textbf{\color{#35bf28}+11.91\%}$
test_redq_speed[reduce-overhead-backward] 12.2307ms 11.7311ms 85.2438 Ops/s 77.4540 Ops/s $\textbf{\color{#35bf28}+10.06\%}$
test_redq_deprec_speed[False-None] 13.4510ms 12.4080ms 80.5930 Ops/s 76.6075 Ops/s $\textbf{\color{#35bf28}+5.20\%}$
test_redq_deprec_speed[False-backward] 20.2394ms 18.1111ms 55.2146 Ops/s 50.8634 Ops/s $\textbf{\color{#35bf28}+8.55\%}$
test_redq_deprec_speed[True-None] 4.5017ms 3.5612ms 280.8042 Ops/s 274.1604 Ops/s $\color{#35bf28}+2.42\%$
test_redq_deprec_speed[True-backward] 8.7983ms 8.0159ms 124.7515 Ops/s 122.0335 Ops/s $\color{#35bf28}+2.23\%$
test_redq_deprec_speed[reduce-overhead-None] 4.7178ms 3.5536ms 281.4028 Ops/s 273.0734 Ops/s $\color{#35bf28}+3.05\%$
test_redq_deprec_speed[reduce-overhead-backward] 8.2211ms 7.8892ms 126.7562 Ops/s 122.0617 Ops/s $\color{#35bf28}+3.85\%$
test_td3_speed[False-None] 8.0042ms 7.7367ms 129.2549 Ops/s 126.8785 Ops/s $\color{#35bf28}+1.87\%$
test_td3_speed[False-backward] 11.7184ms 10.1801ms 98.2305 Ops/s 96.3768 Ops/s $\color{#35bf28}+1.92\%$
test_td3_speed[True-None] 1.8620ms 1.7138ms 583.5114 Ops/s 565.8928 Ops/s $\color{#35bf28}+3.11\%$
test_td3_speed[True-backward] 3.3878ms 3.3270ms 300.5695 Ops/s 291.4255 Ops/s $\color{#35bf28}+3.14\%$
test_td3_speed[reduce-overhead-None] 1.8068ms 1.7181ms 582.0400 Ops/s 557.6295 Ops/s $\color{#35bf28}+4.38\%$
test_td3_speed[reduce-overhead-backward] 3.3943ms 3.3132ms 301.8214 Ops/s 285.8219 Ops/s $\textbf{\color{#35bf28}+5.60\%}$
test_cql_speed[False-None] 44.9391ms 36.2649ms 27.5749 Ops/s 27.3620 Ops/s $\color{#35bf28}+0.78\%$
test_cql_speed[False-backward] 0.2767s 51.1013ms 19.5690 Ops/s 21.4575 Ops/s $\textbf{\color{#d91a1a}-8.80\%}$
test_cql_speed[True-None] 16.9444ms 15.7981ms 63.2988 Ops/s 64.3498 Ops/s $\color{#d91a1a}-1.63\%$
test_cql_speed[True-backward] 23.4909ms 22.1310ms 45.1855 Ops/s 43.9347 Ops/s $\color{#35bf28}+2.85\%$
test_cql_speed[reduce-overhead-None] 16.2246ms 15.4229ms 64.8386 Ops/s 62.3311 Ops/s $\color{#35bf28}+4.02\%$
test_cql_speed[reduce-overhead-backward] 23.1687ms 21.9593ms 45.5389 Ops/s 44.5610 Ops/s $\color{#35bf28}+2.19\%$
test_a2c_speed[False-None] 8.5970ms 7.0628ms 141.5874 Ops/s 137.7662 Ops/s $\color{#35bf28}+2.77\%$
test_a2c_speed[False-backward] 16.1906ms 14.1487ms 70.6776 Ops/s 68.4180 Ops/s $\color{#35bf28}+3.30\%$
test_a2c_speed[True-None] 3.9132ms 3.2904ms 303.9107 Ops/s 300.0172 Ops/s $\color{#35bf28}+1.30\%$
test_a2c_speed[True-backward] 10.0701ms 9.6152ms 104.0017 Ops/s 102.3440 Ops/s $\color{#35bf28}+1.62\%$
test_a2c_speed[reduce-overhead-None] 4.3648ms 3.3622ms 297.4270 Ops/s 298.7917 Ops/s $\color{#d91a1a}-0.46\%$
test_a2c_speed[reduce-overhead-backward] 9.8964ms 9.5808ms 104.3755 Ops/s 102.9064 Ops/s $\color{#35bf28}+1.43\%$
test_ppo_speed[False-None] 9.0911ms 7.3445ms 136.1567 Ops/s 134.8228 Ops/s $\color{#35bf28}+0.99\%$
test_ppo_speed[False-backward] 15.9226ms 14.5647ms 68.6593 Ops/s 66.0596 Ops/s $\color{#35bf28}+3.94\%$
test_ppo_speed[True-None] 4.3473ms 3.6997ms 270.2929 Ops/s 267.3561 Ops/s $\color{#35bf28}+1.10\%$
test_ppo_speed[True-backward] 9.8361ms 9.4411ms 105.9197 Ops/s 104.7688 Ops/s $\color{#35bf28}+1.10\%$
test_ppo_speed[reduce-overhead-None] 4.3747ms 3.6859ms 271.3035 Ops/s 265.9987 Ops/s $\color{#35bf28}+1.99\%$
test_ppo_speed[reduce-overhead-backward] 9.7554ms 9.4770ms 105.5191 Ops/s 104.9060 Ops/s $\color{#35bf28}+0.58\%$
test_reinforce_speed[False-None] 7.3569ms 6.4067ms 156.0861 Ops/s 153.9277 Ops/s $\color{#35bf28}+1.40\%$
test_reinforce_speed[False-backward] 11.3029ms 9.6244ms 103.9030 Ops/s 102.7136 Ops/s $\color{#35bf28}+1.16\%$
test_reinforce_speed[True-None] 3.5578ms 2.9083ms 343.8387 Ops/s 375.4933 Ops/s $\textbf{\color{#d91a1a}-8.43\%}$
test_reinforce_speed[True-backward] 9.7562ms 8.5101ms 117.5078 Ops/s 117.5461 Ops/s $\color{#d91a1a}-0.03\%$
test_reinforce_speed[reduce-overhead-None] 3.0892ms 2.6212ms 381.5070 Ops/s 371.8722 Ops/s $\color{#35bf28}+2.59\%$
test_reinforce_speed[reduce-overhead-backward] 9.0909ms 8.4699ms 118.0650 Ops/s 116.7153 Ops/s $\color{#35bf28}+1.16\%$
test_iql_speed[False-None] 33.9529ms 31.8166ms 31.4301 Ops/s 30.7306 Ops/s $\color{#35bf28}+2.28\%$
test_iql_speed[False-backward] 47.0128ms 44.6892ms 22.3768 Ops/s 21.7683 Ops/s $\color{#35bf28}+2.80\%$
test_iql_speed[True-None] 11.6362ms 10.4632ms 95.5735 Ops/s 93.9577 Ops/s $\color{#35bf28}+1.72\%$
test_iql_speed[True-backward] 22.7498ms 21.1695ms 47.2377 Ops/s 46.6716 Ops/s $\color{#35bf28}+1.21\%$
test_iql_speed[reduce-overhead-None] 11.9718ms 10.4598ms 95.6044 Ops/s 93.3842 Ops/s $\color{#35bf28}+2.38\%$
test_iql_speed[reduce-overhead-backward] 22.9129ms 21.1937ms 47.1838 Ops/s 46.0657 Ops/s $\color{#35bf28}+2.43\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.4980ms 4.8201ms 207.4625 Ops/s 204.2726 Ops/s $\color{#35bf28}+1.56\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.7587ms 0.4977ms 2.0092 KOps/s 2.0762 KOps/s $\color{#d91a1a}-3.23\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.7900ms 0.4783ms 2.0909 KOps/s 2.1591 KOps/s $\color{#d91a1a}-3.16\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 7.1100ms 4.5881ms 217.9551 Ops/s 214.7107 Ops/s $\color{#35bf28}+1.51\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.6461ms 0.4950ms 2.0202 KOps/s 2.0873 KOps/s $\color{#d91a1a}-3.21\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6675ms 0.4667ms 2.1425 KOps/s 2.2159 KOps/s $\color{#d91a1a}-3.31\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 2.4295ms 1.6274ms 614.4582 Ops/s 612.6243 Ops/s $\color{#35bf28}+0.30\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 2.2780ms 1.5763ms 634.4029 Ops/s 634.9500 Ops/s $\color{#d91a1a}-0.09\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 7.2917ms 4.6780ms 213.7650 Ops/s 209.5792 Ops/s $\color{#35bf28}+2.00\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.4955ms 0.6383ms 1.5667 KOps/s 1.6060 KOps/s $\color{#d91a1a}-2.45\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.9909ms 0.6168ms 1.6213 KOps/s 1.6644 KOps/s $\color{#d91a1a}-2.59\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.0524ms 4.5159ms 221.4404 Ops/s 216.7544 Ops/s $\color{#35bf28}+2.16\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.9835ms 0.5012ms 1.9953 KOps/s 2.0661 KOps/s $\color{#d91a1a}-3.43\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6671ms 0.4748ms 2.1062 KOps/s 2.1748 KOps/s $\color{#d91a1a}-3.16\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 4.8527ms 4.5426ms 220.1405 Ops/s 221.0962 Ops/s $\color{#d91a1a}-0.43\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.6267ms 0.4946ms 2.0218 KOps/s 2.0463 KOps/s $\color{#d91a1a}-1.20\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.8227ms 0.4757ms 2.1020 KOps/s 2.2072 KOps/s $\color{#d91a1a}-4.76\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 4.9300ms 4.6915ms 213.1518 Ops/s 207.2505 Ops/s $\color{#35bf28}+2.85\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.2117ms 0.6475ms 1.5445 KOps/s 1.6102 KOps/s $\color{#d91a1a}-4.08\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7515ms 0.6140ms 1.6288 KOps/s 1.6304 KOps/s $\color{#d91a1a}-0.10\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 7.2267ms 4.2206ms 236.9359 Ops/s 230.5484 Ops/s $\color{#35bf28}+2.77\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 5.7785ms 2.2371ms 447.0165 Ops/s 443.1257 Ops/s $\color{#35bf28}+0.88\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 6.1001ms 1.2778ms 782.6181 Ops/s 784.9101 Ops/s $\color{#d91a1a}-0.29\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.3822s 11.7820ms 84.8752 Ops/s 221.7003 Ops/s $\textbf{\color{#d91a1a}-61.72\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 6.2995ms 2.3519ms 425.1925 Ops/s 445.1053 Ops/s $\color{#d91a1a}-4.47\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 5.4245ms 1.3281ms 752.9362 Ops/s 782.6513 Ops/s $\color{#d91a1a}-3.80\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 5.8949ms 4.3423ms 230.2950 Ops/s 224.7640 Ops/s $\color{#35bf28}+2.46\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 8.9776ms 2.4808ms 403.0939 Ops/s 413.2690 Ops/s $\color{#d91a1a}-2.46\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 4.8704ms 1.4510ms 689.1965 Ops/s 696.4398 Ops/s $\color{#d91a1a}-1.04\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000-100-True] 45.6838ms 40.3827ms 24.7631 Ops/s 22.0373 Ops/s $\textbf{\color{#35bf28}+12.37\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000-100-False] 9.2556ms 8.7897ms 113.7697 Ops/s 109.9488 Ops/s $\color{#35bf28}+3.48\%$

@github-actions GitHub Actions
Copy link

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

Total Benchmarks: 145. Improved: $\large\color{#35bf28}13$. Worsened: $\large\color{#d91a1a}15$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_simple 0.7263s 0.7262s 1.3771 Ops/s 1.3916 Ops/s $\color{#d91a1a}-1.04\%$
test_transformed 1.0612s 0.9846s 1.0156 Ops/s 1.0436 Ops/s $\color{#d91a1a}-2.68\%$
test_serial 2.1701s 2.0911s 0.4782 Ops/s 0.4804 Ops/s $\color{#d91a1a}-0.45\%$
test_parallel 2.0851s 1.9809s 0.5048 Ops/s 0.5113 Ops/s $\color{#d91a1a}-1.26\%$
test_step_mdp_speed[True-True-True-True-True] 0.2233ms 36.4449μs 27.4387 KOps/s 27.6337 KOps/s $\color{#d91a1a}-0.71\%$
test_step_mdp_speed[True-True-True-True-False] 63.4110μs 20.7100μs 48.2859 KOps/s 48.2264 KOps/s $\color{#35bf28}+0.12\%$
test_step_mdp_speed[True-True-True-False-True] 47.6600μs 20.1077μs 49.7322 KOps/s 50.8852 KOps/s $\color{#d91a1a}-2.27\%$
test_step_mdp_speed[True-True-True-False-False] 50.0610μs 11.7278μs 85.2675 KOps/s 86.4735 KOps/s $\color{#d91a1a}-1.39\%$
test_step_mdp_speed[True-True-False-True-True] 70.4010μs 38.8063μs 25.7690 KOps/s 25.9925 KOps/s $\color{#d91a1a}-0.86\%$
test_step_mdp_speed[True-True-False-True-False] 48.4510μs 22.8370μs 43.7886 KOps/s 44.8188 KOps/s $\color{#d91a1a}-2.30\%$
test_step_mdp_speed[True-True-False-False-True] 52.5000μs 21.6475μs 46.1948 KOps/s 46.1649 KOps/s $\color{#35bf28}+0.06\%$
test_step_mdp_speed[True-True-False-False-False] 42.7410μs 13.6220μs 73.4107 KOps/s 73.5544 KOps/s $\color{#d91a1a}-0.20\%$
test_step_mdp_speed[True-False-True-True-True] 67.6820μs 41.1139μs 24.3227 KOps/s 24.6154 KOps/s $\color{#d91a1a}-1.19\%$
test_step_mdp_speed[True-False-True-True-False] 54.0810μs 24.9296μs 40.1130 KOps/s 40.0467 KOps/s $\color{#35bf28}+0.17\%$
test_step_mdp_speed[True-False-True-False-True] 50.7810μs 22.0920μs 45.2652 KOps/s 45.3050 KOps/s $\color{#d91a1a}-0.09\%$
test_step_mdp_speed[True-False-True-False-False] 53.6310μs 13.8083μs 72.4201 KOps/s 74.1019 KOps/s $\color{#d91a1a}-2.27\%$
test_step_mdp_speed[True-False-False-True-True] 76.7710μs 43.0752μs 23.2152 KOps/s 24.0770 KOps/s $\color{#d91a1a}-3.58\%$
test_step_mdp_speed[True-False-False-True-False] 57.0210μs 27.0309μs 36.9946 KOps/s 37.7090 KOps/s $\color{#d91a1a}-1.89\%$
test_step_mdp_speed[True-False-False-False-True] 53.7610μs 24.0000μs 41.6667 KOps/s 41.6647 KOps/s $+0.00\%$
test_step_mdp_speed[True-False-False-False-False] 46.4800μs 15.7996μs 63.2928 KOps/s 63.7994 KOps/s $\color{#d91a1a}-0.79\%$
test_step_mdp_speed[False-True-True-True-True] 98.6420μs 40.3969μs 24.7544 KOps/s 25.1633 KOps/s $\color{#d91a1a}-1.62\%$
test_step_mdp_speed[False-True-True-True-False] 65.2110μs 24.4477μs 40.9036 KOps/s 40.6745 KOps/s $\color{#35bf28}+0.56\%$
test_step_mdp_speed[False-True-True-False-True] 49.4910μs 25.7454μs 38.8419 KOps/s 39.3588 KOps/s $\color{#d91a1a}-1.31\%$
test_step_mdp_speed[False-True-True-False-False] 42.7810μs 15.3189μs 65.2789 KOps/s 65.8416 KOps/s $\color{#d91a1a}-0.85\%$
test_step_mdp_speed[False-True-False-True-True] 82.9910μs 41.3356μs 24.1922 KOps/s 23.7163 KOps/s $\color{#35bf28}+2.01\%$
test_step_mdp_speed[False-True-False-True-False] 59.6410μs 26.2197μs 38.1393 KOps/s 37.8834 KOps/s $\color{#35bf28}+0.68\%$
test_step_mdp_speed[False-True-False-False-True] 3.4657ms 28.0636μs 35.6334 KOps/s 35.6190 KOps/s $\color{#35bf28}+0.04\%$
test_step_mdp_speed[False-True-False-False-False] 56.9010μs 17.3258μs 57.7175 KOps/s 57.9876 KOps/s $\color{#d91a1a}-0.47\%$
test_step_mdp_speed[False-False-True-True-True] 77.3010μs 45.1926μs 22.1275 KOps/s 22.2767 KOps/s $\color{#d91a1a}-0.67\%$
test_step_mdp_speed[False-False-True-True-False] 62.4610μs 29.0763μs 34.3923 KOps/s 34.6244 KOps/s $\color{#d91a1a}-0.67\%$
test_step_mdp_speed[False-False-True-False-True] 69.7910μs 28.0167μs 35.6930 KOps/s 36.1817 KOps/s $\color{#d91a1a}-1.35\%$
test_step_mdp_speed[False-False-True-False-False] 49.1710μs 17.5797μs 56.8837 KOps/s 58.0651 KOps/s $\color{#d91a1a}-2.03\%$
test_step_mdp_speed[False-False-False-True-True] 0.1409ms 45.6195μs 21.9204 KOps/s 22.1926 KOps/s $\color{#d91a1a}-1.23\%$
test_step_mdp_speed[False-False-False-True-False] 73.0010μs 30.3763μs 32.9204 KOps/s 32.6601 KOps/s $\color{#35bf28}+0.80\%$
test_step_mdp_speed[False-False-False-False-True] 53.8510μs 29.4178μs 33.9931 KOps/s 34.0869 KOps/s $\color{#d91a1a}-0.28\%$
test_step_mdp_speed[False-False-False-False-False] 46.8010μs 19.0435μs 52.5114 KOps/s 52.1693 KOps/s $\color{#35bf28}+0.66\%$
test_values[generalized_advantage_estimate-True-True] 25.1251ms 24.7334ms 40.4312 Ops/s 41.3879 Ops/s $\color{#d91a1a}-2.31\%$
test_values[vec_generalized_advantage_estimate-True-True] 0.1199s 3.2849ms 304.4205 Ops/s 346.4279 Ops/s $\textbf{\color{#d91a1a}-12.13\%}$
test_values[td0_return_estimate-False-False] 0.1398ms 66.5821μs 15.0191 KOps/s 15.6020 KOps/s $\color{#d91a1a}-3.74\%$
test_values[td1_return_estimate-False-False] 55.6121ms 55.2454ms 18.1011 Ops/s 18.5022 Ops/s $\color{#d91a1a}-2.17\%$
test_values[vec_td1_return_estimate-False-False] 1.2567ms 1.0676ms 936.7160 Ops/s 944.2852 Ops/s $\color{#d91a1a}-0.80\%$
test_values[td_lambda_return_estimate-True-False] 87.9760ms 87.5931ms 11.4164 Ops/s 11.6570 Ops/s $\color{#d91a1a}-2.06\%$
test_values[vec_td_lambda_return_estimate-True-False] 1.2312ms 1.0639ms 939.9374 Ops/s 947.5478 Ops/s $\color{#d91a1a}-0.80\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 24.5362ms 24.3687ms 41.0362 Ops/s 39.8636 Ops/s $\color{#35bf28}+2.94\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.0257ms 0.7407ms 1.3500 KOps/s 1.3692 KOps/s $\color{#d91a1a}-1.40\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.7635ms 0.6554ms 1.5257 KOps/s 1.5486 KOps/s $\color{#d91a1a}-1.48\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 1.5185ms 1.4685ms 680.9749 Ops/s 684.8512 Ops/s $\color{#d91a1a}-0.57\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 0.7155ms 0.6702ms 1.4921 KOps/s 1.4883 KOps/s $\color{#35bf28}+0.25\%$
test_dqn_speed[False-None] 6.8643ms 1.2916ms 774.2515 Ops/s 782.6185 Ops/s $\color{#d91a1a}-1.07\%$
test_dqn_speed[False-backward] 1.8587ms 1.7784ms 562.2967 Ops/s 556.9405 Ops/s $\color{#35bf28}+0.96\%$
test_dqn_speed[True-None] 0.9012ms 0.5524ms 1.8103 KOps/s 1.6972 KOps/s $\textbf{\color{#35bf28}+6.66\%}$
test_dqn_speed[True-backward] 1.1204ms 0.9937ms 1.0064 KOps/s 828.2356 Ops/s $\textbf{\color{#35bf28}+21.51\%}$
test_dqn_speed[reduce-overhead-None] 0.6200ms 0.5533ms 1.8073 KOps/s 1.7104 KOps/s $\textbf{\color{#35bf28}+5.67\%}$
test_dqn_speed[reduce-overhead-backward] 1.0927ms 1.0224ms 978.0989 Ops/s 1.0106 KOps/s $\color{#d91a1a}-3.22\%$
test_ddpg_speed[False-None] 3.4186ms 2.6645ms 375.2986 Ops/s 374.8383 Ops/s $\color{#35bf28}+0.12\%$
test_ddpg_speed[False-backward] 3.9713ms 3.8486ms 259.8340 Ops/s 261.8517 Ops/s $\color{#d91a1a}-0.77\%$
test_ddpg_speed[True-None] 1.3801ms 1.2431ms 804.4572 Ops/s 785.3124 Ops/s $\color{#35bf28}+2.44\%$
test_ddpg_speed[True-backward] 2.2368ms 2.1662ms 461.6328 Ops/s 454.3018 Ops/s $\color{#35bf28}+1.61\%$
test_ddpg_speed[reduce-overhead-None] 1.3353ms 1.2196ms 819.9220 Ops/s 801.7292 Ops/s $\color{#35bf28}+2.27\%$
test_ddpg_speed[reduce-overhead-backward] 2.2734ms 2.1620ms 462.5391 Ops/s 455.4313 Ops/s $\color{#35bf28}+1.56\%$
test_sac_speed[False-None] 8.6948ms 7.3285ms 136.4527 Ops/s 135.3828 Ops/s $\color{#35bf28}+0.79\%$
test_sac_speed[False-backward] 11.0928ms 10.4450ms 95.7393 Ops/s 96.1315 Ops/s $\color{#d91a1a}-0.41\%$
test_sac_speed[True-None] 2.1879ms 1.9859ms 503.5406 Ops/s 469.6459 Ops/s $\textbf{\color{#35bf28}+7.22\%}$
test_sac_speed[True-backward] 4.0009ms 3.9117ms 255.6434 Ops/s 249.7056 Ops/s $\color{#35bf28}+2.38\%$
test_sac_speed[reduce-overhead-None] 2.1054ms 1.9827ms 504.3596 Ops/s 498.5620 Ops/s $\color{#35bf28}+1.16\%$
test_sac_speed[reduce-overhead-backward] 3.9640ms 3.8433ms 260.1928 Ops/s 255.6252 Ops/s $\color{#35bf28}+1.79\%$
test_redq_speed[False-None] 15.0083ms 9.9675ms 100.3261 Ops/s 101.8415 Ops/s $\color{#d91a1a}-1.49\%$
test_redq_speed[False-backward] 17.7578ms 16.9014ms 59.1668 Ops/s 60.3767 Ops/s $\color{#d91a1a}-2.00\%$
test_redq_speed[True-None] 3.6974ms 3.4516ms 289.7228 Ops/s 297.6183 Ops/s $\color{#d91a1a}-2.65\%$
test_redq_speed[True-backward] 8.8649ms 8.5407ms 117.0860 Ops/s 122.6544 Ops/s $\color{#d91a1a}-4.54\%$
test_redq_speed[reduce-overhead-None] 3.7682ms 3.5508ms 281.6245 Ops/s 303.6865 Ops/s $\textbf{\color{#d91a1a}-7.26\%}$
test_redq_speed[reduce-overhead-backward] 8.8843ms 8.5294ms 117.2411 Ops/s 123.4396 Ops/s $\textbf{\color{#d91a1a}-5.02\%}$
test_redq_deprec_speed[False-None] 11.0514ms 10.3165ms 96.9325 Ops/s 96.3213 Ops/s $\color{#35bf28}+0.63\%$
test_redq_deprec_speed[False-backward] 15.5452ms 15.0763ms 66.3293 Ops/s 66.3618 Ops/s $\color{#d91a1a}-0.05\%$
test_redq_deprec_speed[True-None] 3.3975ms 3.1999ms 312.5135 Ops/s 306.4938 Ops/s $\color{#35bf28}+1.96\%$
test_redq_deprec_speed[True-backward] 7.5170ms 7.1858ms 139.1643 Ops/s 140.2049 Ops/s $\color{#d91a1a}-0.74\%$
test_redq_deprec_speed[reduce-overhead-None] 3.4166ms 3.2431ms 308.3462 Ops/s 304.0174 Ops/s $\color{#35bf28}+1.42\%$
test_redq_deprec_speed[reduce-overhead-backward] 7.4243ms 7.1820ms 139.2376 Ops/s 140.4250 Ops/s $\color{#d91a1a}-0.85\%$
test_td3_speed[False-None] 33.6415ms 7.5976ms 131.6202 Ops/s 138.2855 Ops/s $\color{#d91a1a}-4.82\%$
test_td3_speed[False-backward] 10.4993ms 10.1136ms 98.8767 Ops/s 99.8938 Ops/s $\color{#d91a1a}-1.02\%$
test_td3_speed[True-None] 1.9138ms 1.8863ms 530.1282 Ops/s 527.4511 Ops/s $\color{#35bf28}+0.51\%$
test_td3_speed[True-backward] 3.7491ms 3.6384ms 274.8427 Ops/s 274.5771 Ops/s $\color{#35bf28}+0.10\%$
test_td3_speed[reduce-overhead-None] 1.9018ms 1.8828ms 531.1143 Ops/s 525.8575 Ops/s $\color{#35bf28}+1.00\%$
test_td3_speed[reduce-overhead-backward] 3.8174ms 3.6307ms 275.4253 Ops/s 273.6260 Ops/s $\color{#35bf28}+0.66\%$
test_cql_speed[False-None] 27.2321ms 24.4187ms 40.9522 Ops/s 41.4235 Ops/s $\color{#d91a1a}-1.14\%$
test_cql_speed[False-backward] 37.4943ms 33.8320ms 29.5578 Ops/s 30.2939 Ops/s $\color{#d91a1a}-2.43\%$
test_cql_speed[True-None] 11.1230ms 10.7825ms 92.7427 Ops/s 96.8400 Ops/s $\color{#d91a1a}-4.23\%$
test_cql_speed[True-backward] 16.7136ms 16.4044ms 60.9593 Ops/s 66.9371 Ops/s $\textbf{\color{#d91a1a}-8.93\%}$
test_cql_speed[reduce-overhead-None] 11.1284ms 10.8159ms 92.4568 Ops/s 102.3257 Ops/s $\textbf{\color{#d91a1a}-9.64\%}$
test_cql_speed[reduce-overhead-backward] 16.6566ms 16.2376ms 61.5854 Ops/s 66.8687 Ops/s $\textbf{\color{#d91a1a}-7.90\%}$
test_a2c_speed[False-None] 5.3203ms 5.1188ms 195.3579 Ops/s 202.1801 Ops/s $\color{#d91a1a}-3.37\%$
test_a2c_speed[False-backward] 11.8230ms 11.3362ms 88.2131 Ops/s 87.6412 Ops/s $\color{#35bf28}+0.65\%$
test_a2c_speed[True-None] 3.3072ms 2.9864ms 334.8471 Ops/s 363.3952 Ops/s $\textbf{\color{#d91a1a}-7.86\%}$
test_a2c_speed[True-backward] 8.5906ms 8.2415ms 121.3375 Ops/s 119.7875 Ops/s $\color{#35bf28}+1.29\%$
test_a2c_speed[reduce-overhead-None] 3.1420ms 2.9973ms 333.6377 Ops/s 361.5097 Ops/s $\textbf{\color{#d91a1a}-7.71\%}$
test_a2c_speed[reduce-overhead-backward] 8.4671ms 8.1896ms 122.1067 Ops/s 127.0568 Ops/s $\color{#d91a1a}-3.90\%$
test_ppo_speed[False-None] 6.1804ms 5.5281ms 180.8943 Ops/s 192.9478 Ops/s $\textbf{\color{#d91a1a}-6.25\%}$
test_ppo_speed[False-backward] 12.3677ms 12.0388ms 83.0646 Ops/s 84.7256 Ops/s $\color{#d91a1a}-1.96\%$
test_ppo_speed[True-None] 3.5909ms 3.4345ms 291.1605 Ops/s 315.8091 Ops/s $\textbf{\color{#d91a1a}-7.80\%}$
test_ppo_speed[True-backward] 8.0854ms 7.9242ms 126.1963 Ops/s 122.3495 Ops/s $\color{#35bf28}+3.14\%$
test_ppo_speed[reduce-overhead-None] 3.5900ms 3.4069ms 293.5192 Ops/s 293.1329 Ops/s $\color{#35bf28}+0.13\%$
test_ppo_speed[reduce-overhead-backward] 8.3114ms 7.9976ms 125.0373 Ops/s 126.5402 Ops/s $\color{#d91a1a}-1.19\%$
test_reinforce_speed[False-None] 6.2010ms 4.3701ms 228.8298 Ops/s 229.8840 Ops/s $\color{#d91a1a}-0.46\%$
test_reinforce_speed[False-backward] 7.4849ms 7.1586ms 139.6923 Ops/s 140.4760 Ops/s $\color{#d91a1a}-0.56\%$
test_reinforce_speed[True-None] 2.3422ms 2.1759ms 459.5709 Ops/s 442.3845 Ops/s $\color{#35bf28}+3.88\%$
test_reinforce_speed[True-backward] 7.1182ms 6.9098ms 144.7210 Ops/s 134.8024 Ops/s $\textbf{\color{#35bf28}+7.36\%}$
test_reinforce_speed[reduce-overhead-None] 2.4570ms 2.2033ms 453.8618 Ops/s 456.6721 Ops/s $\color{#d91a1a}-0.62\%$
test_reinforce_speed[reduce-overhead-backward] 7.1219ms 6.9409ms 144.0729 Ops/s 147.6085 Ops/s $\color{#d91a1a}-2.40\%$
test_iql_speed[False-None] 19.8780ms 19.0751ms 52.4242 Ops/s 53.8734 Ops/s $\color{#d91a1a}-2.69\%$
test_iql_speed[False-backward] 30.1513ms 29.2161ms 34.2277 Ops/s 34.6587 Ops/s $\color{#d91a1a}-1.24\%$
test_iql_speed[True-None] 7.0462ms 6.6387ms 150.6325 Ops/s 150.0283 Ops/s $\color{#35bf28}+0.40\%$
test_iql_speed[True-backward] 15.3029ms 14.8967ms 67.1289 Ops/s 66.7119 Ops/s $\color{#35bf28}+0.63\%$
test_iql_speed[reduce-overhead-None] 6.9653ms 6.6257ms 150.9269 Ops/s 151.2150 Ops/s $\color{#d91a1a}-0.19\%$
test_iql_speed[reduce-overhead-backward] 15.5682ms 15.0364ms 66.5052 Ops/s 66.5518 Ops/s $\color{#d91a1a}-0.07\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.4456ms 6.2840ms 159.1348 Ops/s 155.3300 Ops/s $\color{#35bf28}+2.45\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.9102ms 0.2925ms 3.4185 KOps/s 3.8386 KOps/s $\textbf{\color{#d91a1a}-10.94\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5036ms 0.2695ms 3.7111 KOps/s 4.0333 KOps/s $\textbf{\color{#d91a1a}-7.99\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.2559ms 6.0344ms 165.7170 Ops/s 161.1526 Ops/s $\color{#35bf28}+2.83\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.6201ms 0.2511ms 3.9821 KOps/s 3.8664 KOps/s $\color{#35bf28}+2.99\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.3786ms 0.2310ms 4.3290 KOps/s 3.4980 KOps/s $\textbf{\color{#35bf28}+23.76\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.4130ms 1.2333ms 810.8270 Ops/s 787.0803 Ops/s $\color{#35bf28}+3.02\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.3700ms 1.1751ms 850.9764 Ops/s 814.2153 Ops/s $\color{#35bf28}+4.51\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.5344ms 6.3121ms 158.4248 Ops/s 155.4729 Ops/s $\color{#35bf28}+1.90\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 0.6823ms 0.4211ms 2.3747 KOps/s 2.3706 KOps/s $\color{#35bf28}+0.17\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7587ms 0.3945ms 2.5351 KOps/s 2.4466 KOps/s $\color{#35bf28}+3.62\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.1695ms 6.0610ms 164.9906 Ops/s 159.6355 Ops/s $\color{#35bf28}+3.35\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.5786ms 0.2541ms 3.9351 KOps/s 3.4663 KOps/s $\textbf{\color{#35bf28}+13.52\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5822ms 0.2355ms 4.2455 KOps/s 3.7100 KOps/s $\textbf{\color{#35bf28}+14.43\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.2367ms 6.0209ms 166.0877 Ops/s 163.0772 Ops/s $\color{#35bf28}+1.85\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.5676ms 0.2764ms 3.6182 KOps/s 2.4458 KOps/s $\textbf{\color{#35bf28}+47.94\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 6.7301ms 0.2404ms 4.1590 KOps/s 2.5481 KOps/s $\textbf{\color{#35bf28}+63.22\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.3949ms 6.2538ms 159.9025 Ops/s 158.0871 Ops/s $\color{#35bf28}+1.15\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.0567ms 0.4441ms 2.2519 KOps/s 2.4870 KOps/s $\textbf{\color{#d91a1a}-9.45\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.8122ms 0.4164ms 2.4017 KOps/s 2.6639 KOps/s $\textbf{\color{#d91a1a}-9.84\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.4040s 13.3042ms 75.1640 Ops/s 186.7863 Ops/s $\textbf{\color{#d91a1a}-59.76\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 9.3936ms 2.0071ms 498.2294 Ops/s 461.1388 Ops/s $\textbf{\color{#35bf28}+8.04\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 7.0521ms 1.1571ms 864.2184 Ops/s 829.4075 Ops/s $\color{#35bf28}+4.20\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 7.0082ms 5.3484ms 186.9731 Ops/s 188.3212 Ops/s $\color{#d91a1a}-0.72\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 9.5544ms 2.0333ms 491.8209 Ops/s 451.5002 Ops/s $\textbf{\color{#35bf28}+8.93\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 6.0986ms 1.1828ms 845.4777 Ops/s 830.6550 Ops/s $\color{#35bf28}+1.78\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.3729s 12.9092ms 77.4639 Ops/s 32.8149 Ops/s $\textbf{\color{#35bf28}+136.06\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 6.8601ms 2.1723ms 460.3494 Ops/s 465.1411 Ops/s $\color{#d91a1a}-1.03\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 9.2228ms 1.3534ms 738.8611 Ops/s 743.7961 Ops/s $\color{#d91a1a}-0.66\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000-100-True] 47.4081ms 45.7946ms 21.8366 Ops/s 21.6221 Ops/s $\color{#35bf28}+0.99\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000-100-False] 10.1031ms 9.6900ms 103.1988 Ops/s 100.8414 Ops/s $\color{#35bf28}+2.34\%$

@vmoens vmoens added the enhancement New feature or request label Oct 29, 2024
@vmoens vmoens merged commit 9534be2 into gh/vmoens/37/base Oct 29, 2024
72 of 74 checks passed
vmoens added a commit that referenced this pull request Oct 29, 2024
@vmoens
[Feature] Send info dict to the storage device in RBs
d524d0d 
ghstack-source-id: 4ed60d649b17f96b49f90d234e679937c60a3c32
Pull Request resolved: #2527
@vmoens vmoens deleted the gh/vmoens/37/head branch October 29, 2024 12:40
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
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
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@vmoens @facebook-github-bot