Skip to content

Conversation

@vmoens
Copy link
Collaborator

@vmoens vmoens commented Oct 23, 2025

[ghstack-poisoned]
@pytorch-bot
Copy link

pytorch-bot bot commented Oct 23, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3219

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.

vmoens added a commit that referenced this pull request Oct 23, 2025
ghstack-source-id: d76d701
Pull-Request: #3219
@meta-cla meta-cla 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 23, 2025
[ghstack-poisoned]
@github-actions
Copy link

github-actions bot commented Oct 23, 2025

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

Total Benchmarks: 154. Improved: $\large\color{#35bf28}16$. Worsened: $\large\color{#d91a1a}14$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 84.2574μs 82.1191μs 12.1774 KOps/s 12.1168 KOps/s $\color{#35bf28}+0.50\%$
test_tensor_to_bytestream_speed[torch.save] 0.1430ms 0.1415ms 7.0681 KOps/s 7.0875 KOps/s $\color{#d91a1a}-0.27\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1212s 0.1207s 8.2851 Ops/s 8.2956 Ops/s $\color{#d91a1a}-0.13\%$
test_tensor_to_bytestream_speed[numpy] 2.8242μs 2.8160μs 355.1124 KOps/s 357.4283 KOps/s $\color{#d91a1a}-0.65\%$
test_tensor_to_bytestream_speed[safetensors] 44.4642μs 43.4560μs 23.0118 KOps/s 23.8278 KOps/s $\color{#d91a1a}-3.42\%$
test_simple 0.5534s 0.5527s 1.8094 Ops/s 1.7326 Ops/s $\color{#35bf28}+4.43\%$
test_transformed 1.1109s 1.1105s 0.9005 Ops/s 0.8692 Ops/s $\color{#35bf28}+3.60\%$
test_serial 1.6690s 1.6663s 0.6001 Ops/s 0.5887 Ops/s $\color{#35bf28}+1.94\%$
test_parallel 1.1684s 1.0929s 0.9150 Ops/s 0.9100 Ops/s $\color{#35bf28}+0.55\%$
test_step_mdp_speed[True-True-True-True-True] 0.1548ms 44.9514μs 22.2462 KOps/s 22.5084 KOps/s $\color{#d91a1a}-1.16\%$
test_step_mdp_speed[True-True-True-True-False] 60.7910μs 25.6203μs 39.0316 KOps/s 38.9805 KOps/s $\color{#35bf28}+0.13\%$
test_step_mdp_speed[True-True-True-False-True] 61.4010μs 25.5909μs 39.0764 KOps/s 39.4285 KOps/s $\color{#d91a1a}-0.89\%$
test_step_mdp_speed[True-True-True-False-False] 48.0110μs 14.1558μs 70.6426 KOps/s 70.2033 KOps/s $\color{#35bf28}+0.63\%$
test_step_mdp_speed[True-True-False-True-True] 0.1011ms 48.4129μs 20.6557 KOps/s 20.5307 KOps/s $\color{#35bf28}+0.61\%$
test_step_mdp_speed[True-True-False-True-False] 79.0910μs 28.1817μs 35.4840 KOps/s 34.6794 KOps/s $\color{#35bf28}+2.32\%$
test_step_mdp_speed[True-True-False-False-True] 64.6310μs 28.6878μs 34.8580 KOps/s 35.3307 KOps/s $\color{#d91a1a}-1.34\%$
test_step_mdp_speed[True-True-False-False-False] 46.0010μs 17.0262μs 58.7329 KOps/s 58.3046 KOps/s $\color{#35bf28}+0.73\%$
test_step_mdp_speed[True-False-True-True-True] 88.9220μs 51.1211μs 19.5614 KOps/s 19.3667 KOps/s $\color{#35bf28}+1.01\%$
test_step_mdp_speed[True-False-True-True-False] 68.9910μs 31.1661μs 32.0862 KOps/s 31.3715 KOps/s $\color{#35bf28}+2.28\%$
test_step_mdp_speed[True-False-True-False-True] 0.1052ms 28.4121μs 35.1963 KOps/s 36.0939 KOps/s $\color{#d91a1a}-2.49\%$
test_step_mdp_speed[True-False-True-False-False] 49.0200μs 17.4288μs 57.3764 KOps/s 59.2737 KOps/s $\color{#d91a1a}-3.20\%$
test_step_mdp_speed[True-False-False-True-True] 97.5510μs 53.8407μs 18.5733 KOps/s 18.3641 KOps/s $\color{#35bf28}+1.14\%$
test_step_mdp_speed[True-False-False-True-False] 82.8310μs 33.9670μs 29.4403 KOps/s 29.3424 KOps/s $\color{#35bf28}+0.33\%$
test_step_mdp_speed[True-False-False-False-True] 0.1066ms 30.9184μs 32.3432 KOps/s 32.8156 KOps/s $\color{#d91a1a}-1.44\%$
test_step_mdp_speed[True-False-False-False-False] 52.1910μs 19.6672μs 50.8461 KOps/s 50.5537 KOps/s $\color{#35bf28}+0.58\%$
test_step_mdp_speed[False-True-True-True-True] 0.1535ms 50.1097μs 19.9562 KOps/s 19.6436 KOps/s $\color{#35bf28}+1.59\%$
test_step_mdp_speed[False-True-True-True-False] 0.1277ms 30.8339μs 32.4318 KOps/s 32.1157 KOps/s $\color{#35bf28}+0.98\%$
test_step_mdp_speed[False-True-True-False-True] 2.3831ms 32.3255μs 30.9354 KOps/s 30.6789 KOps/s $\color{#35bf28}+0.84\%$
test_step_mdp_speed[False-True-True-False-False] 67.7810μs 18.7677μs 53.2829 KOps/s 53.4053 KOps/s $\color{#d91a1a}-0.23\%$
test_step_mdp_speed[False-True-False-True-True] 99.0110μs 54.0898μs 18.4878 KOps/s 18.5240 KOps/s $\color{#d91a1a}-0.20\%$
test_step_mdp_speed[False-True-False-True-False] 69.3210μs 33.0863μs 30.2240 KOps/s 29.1272 KOps/s $\color{#35bf28}+3.77\%$
test_step_mdp_speed[False-True-False-False-True] 73.7010μs 34.6235μs 28.8821 KOps/s 29.2152 KOps/s $\color{#d91a1a}-1.14\%$
test_step_mdp_speed[False-True-False-False-False] 54.7700μs 21.5436μs 46.4175 KOps/s 46.6458 KOps/s $\color{#d91a1a}-0.49\%$
test_step_mdp_speed[False-False-True-True-True] 0.1003ms 55.6693μs 17.9632 KOps/s 17.7424 KOps/s $\color{#35bf28}+1.24\%$
test_step_mdp_speed[False-False-True-True-False] 71.6300μs 35.8653μs 27.8821 KOps/s 27.1861 KOps/s $\color{#35bf28}+2.56\%$
test_step_mdp_speed[False-False-True-False-True] 71.3010μs 34.5800μs 28.9184 KOps/s 28.8689 KOps/s $\color{#35bf28}+0.17\%$
test_step_mdp_speed[False-False-True-False-False] 53.7410μs 21.1935μs 47.1844 KOps/s 46.2892 KOps/s $\color{#35bf28}+1.93\%$
test_step_mdp_speed[False-False-False-True-True] 96.3810μs 58.4253μs 17.1159 KOps/s 17.1818 KOps/s $\color{#d91a1a}-0.38\%$
test_step_mdp_speed[False-False-False-True-False] 70.1010μs 38.9975μs 25.6427 KOps/s 25.2387 KOps/s $\color{#35bf28}+1.60\%$
test_step_mdp_speed[False-False-False-False-True] 70.8010μs 36.4021μs 27.4709 KOps/s 27.0613 KOps/s $\color{#35bf28}+1.51\%$
test_step_mdp_speed[False-False-False-False-False] 0.1059ms 23.5760μs 42.4161 KOps/s 41.1663 KOps/s $\color{#35bf28}+3.04\%$
test_values[generalized_advantage_estimate-True-True] 10.5343ms 10.4171ms 95.9962 Ops/s 99.9367 Ops/s $\color{#d91a1a}-3.94\%$
test_values[vec_generalized_advantage_estimate-True-True] 19.5380ms 17.5614ms 56.9431 Ops/s 89.4448 Ops/s $\textbf{\color{#d91a1a}-36.34\%}$
test_values[td0_return_estimate-False-False] 0.2430ms 0.1262ms 7.9250 KOps/s 7.7573 KOps/s $\color{#35bf28}+2.16\%$
test_values[td1_return_estimate-False-False] 27.9779ms 27.5627ms 36.2809 Ops/s 36.1281 Ops/s $\color{#35bf28}+0.42\%$
test_values[vec_td1_return_estimate-False-False] 17.8975ms 17.6228ms 56.7445 Ops/s 88.6430 Ops/s $\textbf{\color{#d91a1a}-35.99\%}$
test_values[td_lambda_return_estimate-True-False] 41.6315ms 41.0161ms 24.3807 Ops/s 24.3948 Ops/s $\color{#d91a1a}-0.06\%$
test_values[vec_td_lambda_return_estimate-True-False] 18.3990ms 17.6564ms 56.6366 Ops/s 89.1246 Ops/s $\textbf{\color{#d91a1a}-36.45\%}$
test_gae_speed[generalized_advantage_estimate-False-1-512] 9.0308ms 8.8816ms 112.5925 Ops/s 114.7277 Ops/s $\color{#d91a1a}-1.86\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.7155ms 1.5419ms 648.5516 Ops/s 641.0141 Ops/s $\color{#35bf28}+1.18\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.5011ms 0.4259ms 2.3480 KOps/s 2.4018 KOps/s $\color{#d91a1a}-2.24\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 35.3429ms 35.1800ms 28.4252 Ops/s 36.0294 Ops/s $\textbf{\color{#d91a1a}-21.11\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 1.8537ms 1.7266ms 579.1717 Ops/s 582.1158 Ops/s $\color{#d91a1a}-0.51\%$
test_dqn_speed[False-None] 6.4694ms 1.4422ms 693.3798 Ops/s 702.2921 Ops/s $\color{#d91a1a}-1.27\%$
test_dqn_speed[False-backward] 1.9783ms 1.9300ms 518.1468 Ops/s 510.2732 Ops/s $\color{#35bf28}+1.54\%$
test_dqn_speed[True-None] 0.6733ms 0.5190ms 1.9267 KOps/s 1.9441 KOps/s $\color{#d91a1a}-0.90\%$
test_dqn_speed[True-backward] 1.0080ms 0.9652ms 1.0360 KOps/s 880.4854 Ops/s $\textbf{\color{#35bf28}+17.66\%}$
test_dqn_speed[reduce-overhead-None] 0.6282ms 0.5019ms 1.9924 KOps/s 1.8591 KOps/s $\textbf{\color{#35bf28}+7.17\%}$
test_dqn_speed[reduce-overhead-backward] 1.0058ms 0.9519ms 1.0505 KOps/s 919.6224 Ops/s $\textbf{\color{#35bf28}+14.23\%}$
test_ddpg_speed[False-None] 3.2051ms 2.8856ms 346.5509 Ops/s 346.2610 Ops/s $\color{#35bf28}+0.08\%$
test_ddpg_speed[False-backward] 4.3049ms 4.1237ms 242.5031 Ops/s 243.1277 Ops/s $\color{#d91a1a}-0.26\%$
test_ddpg_speed[True-None] 1.5151ms 1.3728ms 728.4214 Ops/s 697.2432 Ops/s $\color{#35bf28}+4.47\%$
test_ddpg_speed[True-backward] 2.3935ms 2.3530ms 424.9945 Ops/s 368.5253 Ops/s $\textbf{\color{#35bf28}+15.32\%}$
test_ddpg_speed[reduce-overhead-None] 1.4774ms 1.3636ms 733.3466 Ops/s 723.0475 Ops/s $\color{#35bf28}+1.42\%$
test_ddpg_speed[reduce-overhead-backward] 2.4168ms 2.3655ms 422.7376 Ops/s 423.0162 Ops/s $\color{#d91a1a}-0.07\%$
test_sac_speed[False-None] 8.3774ms 7.9158ms 126.3299 Ops/s 125.5243 Ops/s $\color{#35bf28}+0.64\%$
test_sac_speed[False-backward] 11.7340ms 11.2404ms 88.9646 Ops/s 89.0982 Ops/s $\color{#d91a1a}-0.15\%$
test_sac_speed[True-None] 2.2612ms 2.1265ms 470.2621 Ops/s 460.0807 Ops/s $\color{#35bf28}+2.21\%$
test_sac_speed[True-backward] 4.1844ms 4.0212ms 248.6814 Ops/s 240.0608 Ops/s $\color{#35bf28}+3.59\%$
test_sac_speed[reduce-overhead-None] 2.2767ms 2.1181ms 472.1226 Ops/s 458.4134 Ops/s $\color{#35bf28}+2.99\%$
test_sac_speed[reduce-overhead-backward] 4.1615ms 4.0515ms 246.8233 Ops/s 242.6738 Ops/s $\color{#35bf28}+1.71\%$
test_redq_speed[False-None] 14.6544ms 10.5172ms 95.0825 Ops/s 95.0526 Ops/s $\color{#35bf28}+0.03\%$
test_redq_speed[False-backward] 18.9485ms 18.2509ms 54.7918 Ops/s 55.9748 Ops/s $\color{#d91a1a}-2.11\%$
test_redq_speed[True-None] 4.7567ms 4.4275ms 225.8600 Ops/s 223.9391 Ops/s $\color{#35bf28}+0.86\%$
test_redq_speed[True-backward] 10.3855ms 10.0334ms 99.6673 Ops/s 103.2898 Ops/s $\color{#d91a1a}-3.51\%$
test_redq_speed[reduce-overhead-None] 4.9043ms 4.3956ms 227.4981 Ops/s 230.0527 Ops/s $\color{#d91a1a}-1.11\%$
test_redq_speed[reduce-overhead-backward] 10.6257ms 10.2597ms 97.4691 Ops/s 99.6694 Ops/s $\color{#d91a1a}-2.21\%$
test_redq_deprec_speed[False-None] 11.6200ms 11.1652ms 89.5641 Ops/s 88.9250 Ops/s $\color{#35bf28}+0.72\%$
test_redq_deprec_speed[False-backward] 16.7394ms 16.2072ms 61.7008 Ops/s 61.7020 Ops/s $-0.00\%$
test_redq_deprec_speed[True-None] 4.7383ms 3.7318ms 267.9658 Ops/s 273.0498 Ops/s $\color{#d91a1a}-1.86\%$
test_redq_deprec_speed[True-backward] 8.1038ms 7.7931ms 128.3182 Ops/s 137.6279 Ops/s $\textbf{\color{#d91a1a}-6.76\%}$
test_redq_deprec_speed[reduce-overhead-None] 3.8690ms 3.6306ms 275.4391 Ops/s 286.1499 Ops/s $\color{#d91a1a}-3.74\%$
test_redq_deprec_speed[reduce-overhead-backward] 7.9085ms 7.6974ms 129.9145 Ops/s 117.7325 Ops/s $\textbf{\color{#35bf28}+10.35\%}$
test_td3_speed[False-None] 8.3096ms 7.9866ms 125.2096 Ops/s 119.1343 Ops/s $\textbf{\color{#35bf28}+5.10\%}$
test_td3_speed[False-backward] 11.4773ms 10.9427ms 91.3855 Ops/s 90.7256 Ops/s $\color{#35bf28}+0.73\%$
test_td3_speed[True-None] 1.8380ms 1.7992ms 555.8027 Ops/s 554.9862 Ops/s $\color{#35bf28}+0.15\%$
test_td3_speed[True-backward] 3.7493ms 3.6579ms 273.3778 Ops/s 275.9096 Ops/s $\color{#d91a1a}-0.92\%$
test_td3_speed[reduce-overhead-None] 1.8419ms 1.7913ms 558.2665 Ops/s 554.4705 Ops/s $\color{#35bf28}+0.68\%$
test_td3_speed[reduce-overhead-backward] 3.7755ms 3.6615ms 273.1087 Ops/s 267.4302 Ops/s $\color{#35bf28}+2.12\%$
test_cql_speed[False-None] 29.1270ms 26.1967ms 38.1728 Ops/s 38.2078 Ops/s $\color{#d91a1a}-0.09\%$
test_cql_speed[False-backward] 38.5612ms 35.6162ms 28.0771 Ops/s 28.2383 Ops/s $\color{#d91a1a}-0.57\%$
test_cql_speed[True-None] 12.7277ms 12.3597ms 80.9081 Ops/s 80.2691 Ops/s $\color{#35bf28}+0.80\%$
test_cql_speed[True-backward] 19.5058ms 18.6282ms 53.6822 Ops/s 53.6354 Ops/s $\color{#35bf28}+0.09\%$
test_cql_speed[reduce-overhead-None] 13.0164ms 12.5554ms 79.6468 Ops/s 74.9650 Ops/s $\textbf{\color{#35bf28}+6.25\%}$
test_cql_speed[reduce-overhead-backward] 19.0318ms 18.6333ms 53.6673 Ops/s 55.8574 Ops/s $\color{#d91a1a}-3.92\%$
test_a2c_speed[False-None] 6.2351ms 5.4116ms 184.7875 Ops/s 182.7202 Ops/s $\color{#35bf28}+1.13\%$
test_a2c_speed[False-backward] 12.4169ms 12.0024ms 83.3170 Ops/s 82.6161 Ops/s $\color{#35bf28}+0.85\%$
test_a2c_speed[True-None] 3.8157ms 3.6790ms 271.8141 Ops/s 267.6696 Ops/s $\color{#35bf28}+1.55\%$
test_a2c_speed[True-backward] 9.4016ms 8.7981ms 113.6603 Ops/s 114.6284 Ops/s $\color{#d91a1a}-0.84\%$
test_a2c_speed[reduce-overhead-None] 3.8465ms 3.7279ms 268.2496 Ops/s 270.3855 Ops/s $\color{#d91a1a}-0.79\%$
test_a2c_speed[reduce-overhead-backward] 9.1937ms 8.9190ms 112.1202 Ops/s 98.2414 Ops/s $\textbf{\color{#35bf28}+14.13\%}$
test_ppo_speed[False-None] 6.1676ms 5.9124ms 169.1360 Ops/s 166.8129 Ops/s $\color{#35bf28}+1.39\%$
test_ppo_speed[False-backward] 13.0177ms 12.6641ms 78.9635 Ops/s 78.1944 Ops/s $\color{#35bf28}+0.98\%$
test_ppo_speed[True-None] 3.7965ms 3.6694ms 272.5218 Ops/s 268.8202 Ops/s $\color{#35bf28}+1.38\%$
test_ppo_speed[True-backward] 8.7215ms 8.5489ms 116.9743 Ops/s 116.4620 Ops/s $\color{#35bf28}+0.44\%$
test_ppo_speed[reduce-overhead-None] 4.1415ms 3.6504ms 273.9412 Ops/s 274.1358 Ops/s $\color{#d91a1a}-0.07\%$
test_ppo_speed[reduce-overhead-backward] 8.9684ms 8.8097ms 113.5118 Ops/s 111.8549 Ops/s $\color{#35bf28}+1.48\%$
test_reinforce_speed[False-None] 4.9247ms 4.6256ms 216.1864 Ops/s 211.4114 Ops/s $\color{#35bf28}+2.26\%$
test_reinforce_speed[False-backward] 7.6972ms 7.5207ms 132.9660 Ops/s 130.5488 Ops/s $\color{#35bf28}+1.85\%$
test_reinforce_speed[True-None] 3.1071ms 2.8952ms 345.4008 Ops/s 334.6618 Ops/s $\color{#35bf28}+3.21\%$
test_reinforce_speed[True-backward] 8.0393ms 7.8466ms 127.4433 Ops/s 128.7557 Ops/s $\color{#d91a1a}-1.02\%$
test_reinforce_speed[reduce-overhead-None] 2.9777ms 2.8602ms 349.6276 Ops/s 347.3964 Ops/s $\color{#35bf28}+0.64\%$
test_reinforce_speed[reduce-overhead-backward] 8.1603ms 7.8999ms 126.5841 Ops/s 116.6782 Ops/s $\textbf{\color{#35bf28}+8.49\%}$
test_iql_speed[False-None] 27.2610ms 20.7586ms 48.1727 Ops/s 47.9907 Ops/s $\color{#35bf28}+0.38\%$
test_iql_speed[False-backward] 37.3440ms 31.7215ms 31.5244 Ops/s 31.8801 Ops/s $\color{#d91a1a}-1.12\%$
test_iql_speed[True-None] 8.8753ms 8.5477ms 116.9903 Ops/s 113.4609 Ops/s $\color{#35bf28}+3.11\%$
test_iql_speed[True-backward] 17.4879ms 17.1173ms 58.4205 Ops/s 56.8769 Ops/s $\color{#35bf28}+2.71\%$
test_iql_speed[reduce-overhead-None] 9.0631ms 8.6563ms 115.5233 Ops/s 114.4643 Ops/s $\color{#35bf28}+0.93\%$
test_iql_speed[reduce-overhead-backward] 17.8270ms 17.4919ms 57.1695 Ops/s 58.0390 Ops/s $\color{#d91a1a}-1.50\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.6554ms 6.0646ms 164.8909 Ops/s 165.6406 Ops/s $\color{#d91a1a}-0.45\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.6872ms 0.3376ms 2.9617 KOps/s 3.0576 KOps/s $\color{#d91a1a}-3.14\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6301ms 0.3215ms 3.1106 KOps/s 3.3836 KOps/s $\textbf{\color{#d91a1a}-8.07\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0700ms 5.7543ms 173.7821 Ops/s 175.6033 Ops/s $\color{#d91a1a}-1.04\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.5801s 0.8364ms 1.1956 KOps/s 3.0650 KOps/s $\textbf{\color{#d91a1a}-60.99\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5987ms 0.2921ms 3.4234 KOps/s 3.5950 KOps/s $\color{#d91a1a}-4.77\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.5582ms 1.2807ms 780.8052 Ops/s 782.6309 Ops/s $\color{#d91a1a}-0.23\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.6023ms 1.2189ms 820.4073 Ops/s 806.3349 Ops/s $\color{#35bf28}+1.75\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.0467ms 5.9264ms 168.7377 Ops/s 171.3233 Ops/s $\color{#d91a1a}-1.51\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.2238ms 0.4258ms 2.3488 KOps/s 2.1056 KOps/s $\textbf{\color{#35bf28}+11.55\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6575ms 0.4322ms 2.3137 KOps/s 2.2267 KOps/s $\color{#35bf28}+3.91\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.2020ms 5.7917ms 172.6598 Ops/s 176.6738 Ops/s $\color{#d91a1a}-2.27\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 2.0386ms 0.3525ms 2.8365 KOps/s 811.1060 Ops/s $\textbf{\color{#35bf28}+249.70\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6252ms 0.3402ms 2.9394 KOps/s 3.8528 KOps/s $\textbf{\color{#d91a1a}-23.71\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 5.9991ms 5.7474ms 173.9926 Ops/s 174.7421 Ops/s $\color{#d91a1a}-0.43\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.7899ms 0.2908ms 3.4388 KOps/s 3.0701 KOps/s $\textbf{\color{#35bf28}+12.01\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5381ms 0.3151ms 3.1731 KOps/s 3.3306 KOps/s $\color{#d91a1a}-4.73\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 9.8871ms 6.1080ms 163.7193 Ops/s 168.1853 Ops/s $\color{#d91a1a}-2.66\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.0021ms 0.4690ms 2.1323 KOps/s 2.2598 KOps/s $\textbf{\color{#d91a1a}-5.64\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.8117ms 0.5033ms 1.9869 KOps/s 2.1917 KOps/s $\textbf{\color{#d91a1a}-9.34\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.4993s 14.9544ms 66.8697 Ops/s 193.0021 Ops/s $\textbf{\color{#d91a1a}-65.35\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 9.0481ms 2.0341ms 491.6268 Ops/s 434.7991 Ops/s $\textbf{\color{#35bf28}+13.07\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 7.7254ms 1.1979ms 834.7887 Ops/s 946.0471 Ops/s $\textbf{\color{#d91a1a}-11.76\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 6.8468ms 5.0865ms 196.6001 Ops/s 54.8715 Ops/s $\textbf{\color{#35bf28}+258.29\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 8.9682ms 2.0287ms 492.9334 Ops/s 704.7330 Ops/s $\textbf{\color{#d91a1a}-30.05\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 7.0288ms 1.2065ms 828.8243 Ops/s 749.4534 Ops/s $\textbf{\color{#35bf28}+10.59\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.4749s 14.6707ms 68.1632 Ops/s 183.8535 Ops/s $\textbf{\color{#d91a1a}-62.93\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 12.3023ms 2.1575ms 463.4972 Ops/s 447.8399 Ops/s $\color{#35bf28}+3.50\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 1.3211ms 1.0129ms 987.2795 Ops/s 700.2072 Ops/s $\textbf{\color{#35bf28}+41.00\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 36.8892ms 32.9534ms 30.3459 Ops/s 29.9234 Ops/s $\color{#35bf28}+1.41\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 22.2888ms 17.8665ms 55.9705 Ops/s 56.4902 Ops/s $\color{#d91a1a}-0.92\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 35.7824ms 33.7047ms 29.6694 Ops/s 29.0088 Ops/s $\color{#35bf28}+2.28\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.9952ms 17.8066ms 56.1590 Ops/s 55.9885 Ops/s $\color{#35bf28}+0.30\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 37.1868ms 35.3460ms 28.2918 Ops/s 27.6710 Ops/s $\color{#35bf28}+2.24\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 20.1667ms 19.1305ms 52.2725 Ops/s 51.3277 Ops/s $\color{#35bf28}+1.84\%$

[ghstack-poisoned]
vmoens added a commit that referenced this pull request Oct 23, 2025
ghstack-source-id: e5f1a7a
Pull-Request: #3219
@vmoens vmoens merged commit a16ee27 into gh/vmoens/168/base Oct 23, 2025
62 of 79 checks passed
@vmoens vmoens deleted the gh/vmoens/168/head branch October 23, 2025 18:47
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.

1 participant