Skip to content

Conversation

[ghstack-poisoned]
@pytorch-bot
Copy link

pytorch-bot bot commented Oct 22, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3216

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

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

[ghstack-poisoned]
[ghstack-poisoned]
@github-actions
Copy link

github-actions bot commented Oct 22, 2025

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

Total Benchmarks: 154. Improved: $\large\color{#35bf28}17$. Worsened: $\large\color{#d91a1a}11$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 83.3515μs 82.1121μs 12.1785 KOps/s 12.1449 KOps/s $\color{#35bf28}+0.28\%$
test_tensor_to_bytestream_speed[torch.save] 0.1433ms 0.1422ms 7.0335 KOps/s 6.9836 KOps/s $\color{#35bf28}+0.71\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1218s 0.1216s 8.2248 Ops/s 8.2295 Ops/s $\color{#d91a1a}-0.06\%$
test_tensor_to_bytestream_speed[numpy] 2.8143μs 2.8091μs 355.9910 KOps/s 344.7018 KOps/s $\color{#35bf28}+3.28\%$
test_tensor_to_bytestream_speed[safetensors] 45.1265μs 42.9519μs 23.2819 KOps/s 23.5659 KOps/s $\color{#d91a1a}-1.21\%$
test_simple 0.5521s 0.5515s 1.8132 Ops/s 1.7296 Ops/s $\color{#35bf28}+4.83\%$
test_transformed 1.1378s 1.1192s 0.8935 Ops/s 0.8785 Ops/s $\color{#35bf28}+1.70\%$
test_serial 1.6951s 1.6828s 0.5942 Ops/s 0.5915 Ops/s $\color{#35bf28}+0.46\%$
test_parallel 1.1739s 1.0985s 0.9104 Ops/s 0.9331 Ops/s $\color{#d91a1a}-2.43\%$
test_step_mdp_speed[True-True-True-True-True] 0.2152ms 45.1847μs 22.1314 KOps/s 22.0735 KOps/s $\color{#35bf28}+0.26\%$
test_step_mdp_speed[True-True-True-True-False] 55.2810μs 25.4209μs 39.3377 KOps/s 39.0507 KOps/s $\color{#35bf28}+0.73\%$
test_step_mdp_speed[True-True-True-False-True] 54.2810μs 25.5265μs 39.1749 KOps/s 39.0572 KOps/s $\color{#35bf28}+0.30\%$
test_step_mdp_speed[True-True-True-False-False] 39.6010μs 14.1605μs 70.6187 KOps/s 71.1570 KOps/s $\color{#d91a1a}-0.76\%$
test_step_mdp_speed[True-True-False-True-True] 79.8410μs 48.9799μs 20.4166 KOps/s 20.5243 KOps/s $\color{#d91a1a}-0.53\%$
test_step_mdp_speed[True-True-False-True-False] 55.5410μs 28.2653μs 35.3791 KOps/s 35.1173 KOps/s $\color{#35bf28}+0.75\%$
test_step_mdp_speed[True-True-False-False-True] 56.8210μs 28.9305μs 34.5656 KOps/s 35.0463 KOps/s $\color{#d91a1a}-1.37\%$
test_step_mdp_speed[True-True-False-False-False] 45.0310μs 17.2609μs 57.9345 KOps/s 59.1200 KOps/s $\color{#d91a1a}-2.01\%$
test_step_mdp_speed[True-False-True-True-True] 87.8520μs 51.5164μs 19.4113 KOps/s 19.2961 KOps/s $\color{#35bf28}+0.60\%$
test_step_mdp_speed[True-False-True-True-False] 63.3910μs 30.9036μs 32.3587 KOps/s 31.9701 KOps/s $\color{#35bf28}+1.22\%$
test_step_mdp_speed[True-False-True-False-True] 61.7820μs 28.7570μs 34.7741 KOps/s 35.2533 KOps/s $\color{#d91a1a}-1.36\%$
test_step_mdp_speed[True-False-True-False-False] 46.7010μs 16.9844μs 58.8774 KOps/s 59.1787 KOps/s $\color{#d91a1a}-0.51\%$
test_step_mdp_speed[True-False-False-True-True] 96.2430μs 53.9634μs 18.5311 KOps/s 18.5383 KOps/s $\color{#d91a1a}-0.04\%$
test_step_mdp_speed[True-False-False-True-False] 93.8320μs 33.5595μs 29.7978 KOps/s 29.3423 KOps/s $\color{#35bf28}+1.55\%$
test_step_mdp_speed[True-False-False-False-True] 72.0120μs 31.0059μs 32.2520 KOps/s 32.4090 KOps/s $\color{#d91a1a}-0.48\%$
test_step_mdp_speed[True-False-False-False-False] 41.1110μs 19.4688μs 51.3644 KOps/s 51.4029 KOps/s $\color{#d91a1a}-0.07\%$
test_step_mdp_speed[False-True-True-True-True] 93.5620μs 50.9900μs 19.6117 KOps/s 19.6183 KOps/s $\color{#d91a1a}-0.03\%$
test_step_mdp_speed[False-True-True-True-False] 71.8520μs 31.2876μs 31.9615 KOps/s 32.8253 KOps/s $\color{#d91a1a}-2.63\%$
test_step_mdp_speed[False-True-True-False-True] 2.3410ms 32.5597μs 30.7128 KOps/s 30.8266 KOps/s $\color{#d91a1a}-0.37\%$
test_step_mdp_speed[False-True-True-False-False] 89.4020μs 18.8396μs 53.0797 KOps/s 52.7217 KOps/s $\color{#35bf28}+0.68\%$
test_step_mdp_speed[False-True-False-True-True] 90.6420μs 54.0499μs 18.5014 KOps/s 18.7646 KOps/s $\color{#d91a1a}-1.40\%$
test_step_mdp_speed[False-True-False-True-False] 65.7510μs 33.8585μs 29.5347 KOps/s 29.3184 KOps/s $\color{#35bf28}+0.74\%$
test_step_mdp_speed[False-True-False-False-True] 68.0810μs 34.8859μs 28.6649 KOps/s 28.6789 KOps/s $\color{#d91a1a}-0.05\%$
test_step_mdp_speed[False-True-False-False-False] 53.7620μs 21.3545μs 46.8285 KOps/s 46.6579 KOps/s $\color{#35bf28}+0.37\%$
test_step_mdp_speed[False-False-True-True-True] 93.7220μs 56.0118μs 17.8534 KOps/s 17.3486 KOps/s $\color{#35bf28}+2.91\%$
test_step_mdp_speed[False-False-True-True-False] 71.1510μs 36.5788μs 27.3383 KOps/s 27.1550 KOps/s $\color{#35bf28}+0.67\%$
test_step_mdp_speed[False-False-True-False-True] 69.9120μs 34.6722μs 28.8415 KOps/s 28.1791 KOps/s $\color{#35bf28}+2.35\%$
test_step_mdp_speed[False-False-True-False-False] 53.2010μs 21.3749μs 46.7838 KOps/s 46.3567 KOps/s $\color{#35bf28}+0.92\%$
test_step_mdp_speed[False-False-False-True-True] 92.7520μs 58.6515μs 17.0499 KOps/s 16.8663 KOps/s $\color{#35bf28}+1.09\%$
test_step_mdp_speed[False-False-False-True-False] 84.6920μs 39.1905μs 25.5164 KOps/s 25.3463 KOps/s $\color{#35bf28}+0.67\%$
test_step_mdp_speed[False-False-False-False-True] 67.5410μs 37.1866μs 26.8914 KOps/s 26.4411 KOps/s $\color{#35bf28}+1.70\%$
test_step_mdp_speed[False-False-False-False-False] 52.5510μs 23.7356μs 42.1308 KOps/s 41.3323 KOps/s $\color{#35bf28}+1.93\%$
test_values[generalized_advantage_estimate-True-True] 11.0669ms 10.2220ms 97.8281 Ops/s 98.1235 Ops/s $\color{#d91a1a}-0.30\%$
test_values[vec_generalized_advantage_estimate-True-True] 15.1408ms 11.1370ms 89.7912 Ops/s 90.6179 Ops/s $\color{#d91a1a}-0.91\%$
test_values[td0_return_estimate-False-False] 0.2322ms 0.1316ms 7.5978 KOps/s 7.4866 KOps/s $\color{#35bf28}+1.48\%$
test_values[td1_return_estimate-False-False] 29.7836ms 28.4141ms 35.1938 Ops/s 36.2841 Ops/s $\color{#d91a1a}-3.00\%$
test_values[vec_td1_return_estimate-False-False] 11.6053ms 11.1085ms 90.0208 Ops/s 89.3446 Ops/s $\color{#35bf28}+0.76\%$
test_values[td_lambda_return_estimate-True-False] 44.2798ms 42.6692ms 23.4361 Ops/s 24.3948 Ops/s $\color{#d91a1a}-3.93\%$
test_values[vec_td_lambda_return_estimate-True-False] 12.2655ms 11.1699ms 89.5266 Ops/s 89.6090 Ops/s $\color{#d91a1a}-0.09\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 8.7783ms 8.7344ms 114.4896 Ops/s 114.2765 Ops/s $\color{#35bf28}+0.19\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.7536ms 1.5330ms 652.3090 Ops/s 642.9795 Ops/s $\color{#35bf28}+1.45\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.8261ms 0.4332ms 2.3081 KOps/s 2.4076 KOps/s $\color{#d91a1a}-4.13\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 29.3158ms 22.8910ms 43.6852 Ops/s 33.7217 Ops/s $\textbf{\color{#35bf28}+29.55\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 2.1138ms 1.7273ms 578.9387 Ops/s 581.3077 Ops/s $\color{#d91a1a}-0.41\%$
test_dqn_speed[False-None] 6.5795ms 1.4420ms 693.4637 Ops/s 701.8537 Ops/s $\color{#d91a1a}-1.20\%$
test_dqn_speed[False-backward] 2.0714ms 1.9722ms 507.0375 Ops/s 518.2926 Ops/s $\color{#d91a1a}-2.17\%$
test_dqn_speed[True-None] 0.6990ms 0.5170ms 1.9343 KOps/s 1.9382 KOps/s $\color{#d91a1a}-0.20\%$
test_dqn_speed[True-backward] 0.9826ms 0.9530ms 1.0493 KOps/s 930.1028 Ops/s $\textbf{\color{#35bf28}+12.82\%}$
test_dqn_speed[reduce-overhead-None] 0.9082ms 0.5215ms 1.9176 KOps/s 1.9474 KOps/s $\color{#d91a1a}-1.53\%$
test_dqn_speed[reduce-overhead-backward] 0.9727ms 0.9381ms 1.0660 KOps/s 876.5477 Ops/s $\textbf{\color{#35bf28}+21.61\%}$
test_ddpg_speed[False-None] 3.2581ms 2.8659ms 348.9366 Ops/s 335.7213 Ops/s $\color{#35bf28}+3.94\%$
test_ddpg_speed[False-backward] 4.1714ms 4.0514ms 246.8262 Ops/s 243.8259 Ops/s $\color{#35bf28}+1.23\%$
test_ddpg_speed[True-None] 1.7686ms 1.3661ms 731.9874 Ops/s 718.7678 Ops/s $\color{#35bf28}+1.84\%$
test_ddpg_speed[True-backward] 2.7046ms 2.3282ms 429.5228 Ops/s 354.5025 Ops/s $\textbf{\color{#35bf28}+21.16\%}$
test_ddpg_speed[reduce-overhead-None] 1.7617ms 1.3626ms 733.8820 Ops/s 729.7902 Ops/s $\color{#35bf28}+0.56\%$
test_ddpg_speed[reduce-overhead-backward] 2.4631ms 2.3070ms 433.4566 Ops/s 347.9591 Ops/s $\textbf{\color{#35bf28}+24.57\%}$
test_sac_speed[False-None] 8.3199ms 7.8818ms 126.8738 Ops/s 123.6180 Ops/s $\color{#35bf28}+2.63\%$
test_sac_speed[False-backward] 11.5695ms 11.1272ms 89.8703 Ops/s 89.4932 Ops/s $\color{#35bf28}+0.42\%$
test_sac_speed[True-None] 2.5087ms 2.0898ms 478.5241 Ops/s 457.3416 Ops/s $\color{#35bf28}+4.63\%$
test_sac_speed[True-backward] 4.4146ms 3.9936ms 250.3984 Ops/s 246.9412 Ops/s $\color{#35bf28}+1.40\%$
test_sac_speed[reduce-overhead-None] 2.2409ms 2.0888ms 478.7499 Ops/s 449.3359 Ops/s $\textbf{\color{#35bf28}+6.55\%}$
test_sac_speed[reduce-overhead-backward] 4.4419ms 4.0062ms 249.6122 Ops/s 230.1602 Ops/s $\textbf{\color{#35bf28}+8.45\%}$
test_redq_speed[False-None] 10.6550ms 10.1695ms 98.3336 Ops/s 98.1271 Ops/s $\color{#35bf28}+0.21\%$
test_redq_speed[False-backward] 18.1486ms 17.4407ms 57.3373 Ops/s 57.2130 Ops/s $\color{#35bf28}+0.22\%$
test_redq_speed[True-None] 4.5705ms 4.3126ms 231.8802 Ops/s 232.6002 Ops/s $\color{#d91a1a}-0.31\%$
test_redq_speed[True-backward] 9.8154ms 9.5400ms 104.8221 Ops/s 100.7920 Ops/s $\color{#35bf28}+4.00\%$
test_redq_speed[reduce-overhead-None] 4.4304ms 4.2765ms 233.8371 Ops/s 226.6851 Ops/s $\color{#35bf28}+3.16\%$
test_redq_speed[reduce-overhead-backward] 10.1661ms 9.7675ms 102.3804 Ops/s 98.1937 Ops/s $\color{#35bf28}+4.26\%$
test_redq_deprec_speed[False-None] 11.0532ms 10.7516ms 93.0094 Ops/s 90.7842 Ops/s $\color{#35bf28}+2.45\%$
test_redq_deprec_speed[False-backward] 16.2043ms 15.6154ms 64.0394 Ops/s 62.8407 Ops/s $\color{#35bf28}+1.91\%$
test_redq_deprec_speed[True-None] 3.7935ms 3.4440ms 290.3597 Ops/s 274.0395 Ops/s $\textbf{\color{#35bf28}+5.96\%}$
test_redq_deprec_speed[True-backward] 7.5407ms 7.1917ms 139.0497 Ops/s 135.6744 Ops/s $\color{#35bf28}+2.49\%$
test_redq_deprec_speed[reduce-overhead-None] 3.7812ms 3.4394ms 290.7481 Ops/s 282.7471 Ops/s $\color{#35bf28}+2.83\%$
test_redq_deprec_speed[reduce-overhead-backward] 7.5138ms 7.2463ms 138.0006 Ops/s 114.6330 Ops/s $\textbf{\color{#35bf28}+20.38\%}$
test_td3_speed[False-None] 8.1969ms 7.9373ms 125.9868 Ops/s 124.6586 Ops/s $\color{#35bf28}+1.07\%$
test_td3_speed[False-backward] 11.3940ms 10.8238ms 92.3890 Ops/s 91.5943 Ops/s $\color{#35bf28}+0.87\%$
test_td3_speed[True-None] 1.8236ms 1.7671ms 565.8842 Ops/s 554.0888 Ops/s $\color{#35bf28}+2.13\%$
test_td3_speed[True-backward] 3.7654ms 3.5414ms 282.3726 Ops/s 274.4501 Ops/s $\color{#35bf28}+2.89\%$
test_td3_speed[reduce-overhead-None] 1.7629ms 1.7302ms 577.9617 Ops/s 564.3090 Ops/s $\color{#35bf28}+2.42\%$
test_td3_speed[reduce-overhead-backward] 3.9259ms 3.5707ms 280.0544 Ops/s 274.5392 Ops/s $\color{#35bf28}+2.01\%$
test_cql_speed[False-None] 29.5414ms 25.5251ms 39.1771 Ops/s 38.3693 Ops/s $\color{#35bf28}+2.11\%$
test_cql_speed[False-backward] 38.8990ms 35.3790ms 28.2654 Ops/s 28.4343 Ops/s $\color{#d91a1a}-0.59\%$
test_cql_speed[True-None] 12.5006ms 11.9817ms 83.4609 Ops/s 80.6359 Ops/s $\color{#35bf28}+3.50\%$
test_cql_speed[True-backward] 17.8104ms 17.4728ms 57.2318 Ops/s 55.8816 Ops/s $\color{#35bf28}+2.42\%$
test_cql_speed[reduce-overhead-None] 12.8091ms 12.1056ms 82.6064 Ops/s 81.6303 Ops/s $\color{#35bf28}+1.20\%$
test_cql_speed[reduce-overhead-backward] 17.8950ms 17.5660ms 56.9283 Ops/s 56.3506 Ops/s $\color{#35bf28}+1.03\%$
test_a2c_speed[False-None] 5.7758ms 5.3797ms 185.8831 Ops/s 185.1784 Ops/s $\color{#35bf28}+0.38\%$
test_a2c_speed[False-backward] 12.0269ms 11.7805ms 84.8863 Ops/s 84.1868 Ops/s $\color{#35bf28}+0.83\%$
test_a2c_speed[True-None] 3.9929ms 3.6123ms 276.8349 Ops/s 272.1732 Ops/s $\color{#35bf28}+1.71\%$
test_a2c_speed[True-backward] 8.7813ms 8.5462ms 117.0108 Ops/s 107.2891 Ops/s $\textbf{\color{#35bf28}+9.06\%}$
test_a2c_speed[reduce-overhead-None] 4.0996ms 3.6772ms 271.9461 Ops/s 272.1013 Ops/s $\color{#d91a1a}-0.06\%$
test_a2c_speed[reduce-overhead-backward] 9.2829ms 8.7394ms 114.4245 Ops/s 115.4241 Ops/s $\color{#d91a1a}-0.87\%$
test_ppo_speed[False-None] 6.2314ms 5.8699ms 170.3593 Ops/s 169.1986 Ops/s $\color{#35bf28}+0.69\%$
test_ppo_speed[False-backward] 12.9193ms 12.5608ms 79.6127 Ops/s 79.9532 Ops/s $\color{#d91a1a}-0.43\%$
test_ppo_speed[True-None] 4.0013ms 3.6089ms 277.0917 Ops/s 275.0427 Ops/s $\color{#35bf28}+0.74\%$
test_ppo_speed[True-backward] 8.5666ms 8.3122ms 120.3049 Ops/s 109.2328 Ops/s $\textbf{\color{#35bf28}+10.14\%}$
test_ppo_speed[reduce-overhead-None] 3.7268ms 3.5869ms 278.7950 Ops/s 276.1709 Ops/s $\color{#35bf28}+0.95\%$
test_ppo_speed[reduce-overhead-backward] 8.9005ms 8.6726ms 115.3052 Ops/s 109.6676 Ops/s $\textbf{\color{#35bf28}+5.14\%}$
test_reinforce_speed[False-None] 5.0218ms 4.5391ms 220.3067 Ops/s 215.0657 Ops/s $\color{#35bf28}+2.44\%$
test_reinforce_speed[False-backward] 7.5314ms 7.3317ms 136.3934 Ops/s 133.7449 Ops/s $\color{#35bf28}+1.98\%$
test_reinforce_speed[True-None] 2.9640ms 2.8082ms 356.0984 Ops/s 349.7049 Ops/s $\color{#35bf28}+1.83\%$
test_reinforce_speed[True-backward] 7.8616ms 7.6251ms 131.1466 Ops/s 130.9991 Ops/s $\color{#35bf28}+0.11\%$
test_reinforce_speed[reduce-overhead-None] 3.1999ms 2.8281ms 353.5928 Ops/s 354.5456 Ops/s $\color{#d91a1a}-0.27\%$
test_reinforce_speed[reduce-overhead-backward] 8.0159ms 7.8070ms 128.0900 Ops/s 123.4890 Ops/s $\color{#35bf28}+3.73\%$
test_iql_speed[False-None] 25.3954ms 20.1807ms 49.5522 Ops/s 47.8322 Ops/s $\color{#35bf28}+3.60\%$
test_iql_speed[False-backward] 36.1396ms 30.4535ms 32.8370 Ops/s 32.6900 Ops/s $\color{#35bf28}+0.45\%$
test_iql_speed[True-None] 8.7562ms 8.3990ms 119.0613 Ops/s 116.0022 Ops/s $\color{#35bf28}+2.64\%$
test_iql_speed[True-backward] 16.8986ms 16.4587ms 60.7583 Ops/s 60.7580 Ops/s $+0.00\%$
test_iql_speed[reduce-overhead-None] 8.7543ms 8.4087ms 118.9245 Ops/s 113.2652 Ops/s $\color{#35bf28}+5.00\%$
test_iql_speed[reduce-overhead-backward] 17.2561ms 16.8846ms 59.2254 Ops/s 57.5483 Ops/s $\color{#35bf28}+2.91\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.5625ms 6.0282ms 165.8881 Ops/s 166.3523 Ops/s $\color{#d91a1a}-0.28\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.5585ms 0.2803ms 3.5676 KOps/s 3.5678 KOps/s $-0.01\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6812ms 0.2588ms 3.8641 KOps/s 3.8372 KOps/s $\color{#35bf28}+0.70\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0008ms 5.7535ms 173.8083 Ops/s 175.4509 Ops/s $\color{#d91a1a}-0.94\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.5801s 0.7092ms 1.4100 KOps/s 3.6551 KOps/s $\textbf{\color{#d91a1a}-61.42\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5642ms 0.3104ms 3.2212 KOps/s 3.9566 KOps/s $\textbf{\color{#d91a1a}-18.59\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.6451ms 1.4169ms 705.7678 Ops/s 804.9533 Ops/s $\textbf{\color{#d91a1a}-12.32\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.6311ms 1.3193ms 757.9565 Ops/s 859.2684 Ops/s $\textbf{\color{#d91a1a}-11.79\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.0723ms 5.9434ms 168.2546 Ops/s 170.6699 Ops/s $\color{#d91a1a}-1.42\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 0.8107ms 0.4454ms 2.2451 KOps/s 2.3637 KOps/s $\textbf{\color{#d91a1a}-5.02\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7299ms 0.4365ms 2.2910 KOps/s 2.4837 KOps/s $\textbf{\color{#d91a1a}-7.76\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.8816ms 5.8018ms 172.3597 Ops/s 175.3358 Ops/s $\color{#d91a1a}-1.70\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 2.0141ms 0.2840ms 3.5208 KOps/s 817.4006 Ops/s $\textbf{\color{#35bf28}+330.73\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.4688ms 0.2910ms 3.4361 KOps/s 3.9068 KOps/s $\textbf{\color{#d91a1a}-12.05\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 5.9017ms 5.7165ms 174.9320 Ops/s 173.7219 Ops/s $\color{#35bf28}+0.70\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.9112ms 0.2724ms 3.6715 KOps/s 3.6481 KOps/s $\color{#35bf28}+0.64\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4920ms 0.2641ms 3.7868 KOps/s 3.9205 KOps/s $\color{#d91a1a}-3.41\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 9.9555ms 6.0964ms 164.0301 Ops/s 167.8244 Ops/s $\color{#d91a1a}-2.26\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.7575ms 0.4294ms 2.3286 KOps/s 2.3424 KOps/s $\color{#d91a1a}-0.59\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6992ms 0.4513ms 2.2156 KOps/s 2.4591 KOps/s $\textbf{\color{#d91a1a}-9.90\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.4927s 14.8160ms 67.4947 Ops/s 194.3828 Ops/s $\textbf{\color{#d91a1a}-65.28\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 3.8177ms 1.7029ms 587.2280 Ops/s 410.5582 Ops/s $\textbf{\color{#35bf28}+43.03\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 2.0154ms 0.9994ms 1.0006 KOps/s 937.1570 Ops/s $\textbf{\color{#35bf28}+6.77\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 7.7904ms 5.0537ms 197.8754 Ops/s 55.5335 Ops/s $\textbf{\color{#35bf28}+256.32\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 7.7034ms 2.0132ms 496.7158 Ops/s 593.4979 Ops/s $\textbf{\color{#d91a1a}-16.31\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 7.9992ms 1.2033ms 831.0333 Ops/s 814.7136 Ops/s $\color{#35bf28}+2.00\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.4683s 14.5328ms 68.8099 Ops/s 187.2255 Ops/s $\textbf{\color{#d91a1a}-63.25\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 10.3913ms 2.0938ms 477.6053 Ops/s 484.5926 Ops/s $\color{#d91a1a}-1.44\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 1.1674ms 0.9789ms 1.0215 KOps/s 751.8691 Ops/s $\textbf{\color{#35bf28}+35.86\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 37.4765ms 32.9664ms 30.3339 Ops/s 29.5803 Ops/s $\color{#35bf28}+2.55\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 19.1332ms 17.5668ms 56.9257 Ops/s 56.8484 Ops/s $\color{#35bf28}+0.14\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 36.6901ms 33.7176ms 29.6581 Ops/s 28.9581 Ops/s $\color{#35bf28}+2.42\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.0957ms 17.6291ms 56.7243 Ops/s 55.6070 Ops/s $\color{#35bf28}+2.01\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 37.9927ms 35.5118ms 28.1597 Ops/s 27.1326 Ops/s $\color{#35bf28}+3.79\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 21.0167ms 19.0803ms 52.4102 Ops/s 51.0876 Ops/s $\color{#35bf28}+2.59\%$

[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Oct 23, 2025
ghstack-source-id: 6362a74
Pull-Request: #3216
@vmoens vmoens merged commit 119d766 into gh/vmoens/166/base Oct 23, 2025
34 of 38 checks passed
@vmoens vmoens deleted the gh/vmoens/166/head branch October 23, 2025 18:50
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. llm/ci

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant