Skip to content

Conversation

@vmoens
Copy link
Collaborator

@vmoens vmoens commented Dec 18, 2025

[ghstack-poisoned]
@pytorch-bot
Copy link

pytorch-bot bot commented Dec 18, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3261

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.

@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 Dec 18, 2025
[ghstack-poisoned]
@github-actions
Copy link

github-actions bot commented Dec 18, 2025

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

Total Benchmarks: 164. Improved: $\large\color{#35bf28}19$. Worsened: $\large\color{#d91a1a}8$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 82.5083μs 81.3665μs 12.2901 KOps/s 12.1071 KOps/s $\color{#35bf28}+1.51\%$
test_tensor_to_bytestream_speed[torch.save] 0.1399ms 0.1393ms 7.1773 KOps/s 7.0818 KOps/s $\color{#35bf28}+1.35\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1186s 0.1184s 8.4439 Ops/s 8.2856 Ops/s $\color{#35bf28}+1.91\%$
test_tensor_to_bytestream_speed[numpy] 2.7389μs 2.7265μs 366.7693 KOps/s 370.2247 KOps/s $\color{#d91a1a}-0.93\%$
test_tensor_to_bytestream_speed[safetensors] 41.0044μs 39.1232μs 25.5603 KOps/s 24.9435 KOps/s $\color{#35bf28}+2.47\%$
test_simple 0.5661s 0.5541s 1.8048 Ops/s 1.7216 Ops/s $\color{#35bf28}+4.84\%$
test_transformed 1.1211s 1.1200s 0.8929 Ops/s 0.8694 Ops/s $\color{#35bf28}+2.70\%$
test_serial 1.6864s 1.6775s 0.5961 Ops/s 0.5933 Ops/s $\color{#35bf28}+0.48\%$
test_parallel 1.2589s 1.1865s 0.8428 Ops/s 0.8993 Ops/s $\textbf{\color{#d91a1a}-6.28\%}$
test_step_mdp_speed[True-True-True-True-True] 0.1433ms 45.6602μs 21.9009 KOps/s 21.9225 KOps/s $\color{#d91a1a}-0.10\%$
test_step_mdp_speed[True-True-True-True-False] 64.0710μs 25.7329μs 38.8608 KOps/s 39.1314 KOps/s $\color{#d91a1a}-0.69\%$
test_step_mdp_speed[True-True-True-False-True] 0.1036ms 25.0966μs 39.8461 KOps/s 39.2666 KOps/s $\color{#35bf28}+1.48\%$
test_step_mdp_speed[True-True-True-False-False] 54.0310μs 13.8310μs 72.3015 KOps/s 70.8747 KOps/s $\color{#35bf28}+2.01\%$
test_step_mdp_speed[True-True-False-True-True] 80.1220μs 47.8653μs 20.8920 KOps/s 20.5128 KOps/s $\color{#35bf28}+1.85\%$
test_step_mdp_speed[True-True-False-True-False] 61.6610μs 28.3555μs 35.2665 KOps/s 35.4426 KOps/s $\color{#d91a1a}-0.50\%$
test_step_mdp_speed[True-True-False-False-True] 0.1445ms 27.8659μs 35.8862 KOps/s 35.1851 KOps/s $\color{#35bf28}+1.99\%$
test_step_mdp_speed[True-True-False-False-False] 73.4310μs 16.5994μs 60.2432 KOps/s 58.7012 KOps/s $\color{#35bf28}+2.63\%$
test_step_mdp_speed[True-False-True-True-True] 0.1047ms 50.6344μs 19.7494 KOps/s 19.1652 KOps/s $\color{#35bf28}+3.05\%$
test_step_mdp_speed[True-False-True-True-False] 71.3810μs 31.1346μs 32.1186 KOps/s 31.7774 KOps/s $\color{#35bf28}+1.07\%$
test_step_mdp_speed[True-False-True-False-True] 69.8120μs 28.2390μs 35.4120 KOps/s 35.0520 KOps/s $\color{#35bf28}+1.03\%$
test_step_mdp_speed[True-False-True-False-False] 46.7910μs 17.0733μs 58.5709 KOps/s 59.0598 KOps/s $\color{#d91a1a}-0.83\%$
test_step_mdp_speed[True-False-False-True-True] 96.9120μs 54.5429μs 18.3342 KOps/s 18.4338 KOps/s $\color{#d91a1a}-0.54\%$
test_step_mdp_speed[True-False-False-True-False] 96.1520μs 34.1206μs 29.3078 KOps/s 29.4054 KOps/s $\color{#d91a1a}-0.33\%$
test_step_mdp_speed[True-False-False-False-True] 82.5120μs 31.0497μs 32.2064 KOps/s 32.1442 KOps/s $\color{#35bf28}+0.19\%$
test_step_mdp_speed[True-False-False-False-False] 50.8910μs 19.8241μs 50.4437 KOps/s 50.5904 KOps/s $\color{#d91a1a}-0.29\%$
test_step_mdp_speed[False-True-True-True-True] 84.7020μs 51.3685μs 19.4672 KOps/s 19.3236 KOps/s $\color{#35bf28}+0.74\%$
test_step_mdp_speed[False-True-True-True-False] 79.3010μs 31.2995μs 31.9494 KOps/s 31.8690 KOps/s $\color{#35bf28}+0.25\%$
test_step_mdp_speed[False-True-True-False-True] 2.2833ms 32.1170μs 31.1361 KOps/s 30.4833 KOps/s $\color{#35bf28}+2.14\%$
test_step_mdp_speed[False-True-True-False-False] 50.2610μs 18.4889μs 54.0865 KOps/s 53.6787 KOps/s $\color{#35bf28}+0.76\%$
test_step_mdp_speed[False-True-False-True-True] 0.1434ms 53.7108μs 18.6182 KOps/s 18.3562 KOps/s $\color{#35bf28}+1.43\%$
test_step_mdp_speed[False-True-False-True-False] 80.5920μs 33.8503μs 29.5418 KOps/s 29.5217 KOps/s $\color{#35bf28}+0.07\%$
test_step_mdp_speed[False-True-False-False-True] 87.6620μs 34.2177μs 29.2246 KOps/s 28.7185 KOps/s $\color{#35bf28}+1.76\%$
test_step_mdp_speed[False-True-False-False-False] 64.8210μs 21.1514μs 47.2781 KOps/s 47.0048 KOps/s $\color{#35bf28}+0.58\%$
test_step_mdp_speed[False-False-True-True-True] 85.6810μs 56.5878μs 17.6716 KOps/s 17.4641 KOps/s $\color{#35bf28}+1.19\%$
test_step_mdp_speed[False-False-True-True-False] 91.1710μs 36.7711μs 27.1953 KOps/s 27.0802 KOps/s $\color{#35bf28}+0.42\%$
test_step_mdp_speed[False-False-True-False-True] 92.8620μs 34.4979μs 28.9873 KOps/s 28.5124 KOps/s $\color{#35bf28}+1.67\%$
test_step_mdp_speed[False-False-True-False-False] 56.5910μs 21.6651μs 46.1573 KOps/s 46.7290 KOps/s $\color{#d91a1a}-1.22\%$
test_step_mdp_speed[False-False-False-True-True] 0.1016ms 59.1165μs 16.9158 KOps/s 16.9578 KOps/s $\color{#d91a1a}-0.25\%$
test_step_mdp_speed[False-False-False-True-False] 0.1086ms 39.2702μs 25.4646 KOps/s 25.4055 KOps/s $\color{#35bf28}+0.23\%$
test_step_mdp_speed[False-False-False-False-True] 73.7210μs 36.6697μs 27.2705 KOps/s 26.7726 KOps/s $\color{#35bf28}+1.86\%$
test_step_mdp_speed[False-False-False-False-False] 57.5310μs 23.6721μs 42.2438 KOps/s 42.1562 KOps/s $\color{#35bf28}+0.21\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.8805s 0.7796s 1.2827 Ops/s 1.2862 Ops/s $\color{#d91a1a}-0.27\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7401s 0.6424s 1.5567 Ops/s 1.5590 Ops/s $\color{#d91a1a}-0.14\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7780s 1.6992s 0.5885 Ops/s 0.5939 Ops/s $\color{#d91a1a}-0.90\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5545s 1.4732s 0.6788 Ops/s 0.6839 Ops/s $\color{#d91a1a}-0.75\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0301s 1.9520s 0.5123 Ops/s 0.5144 Ops/s $\color{#d91a1a}-0.40\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.8076s 1.7263s 0.5793 Ops/s 0.5831 Ops/s $\color{#d91a1a}-0.66\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.7021s 4.6511s 0.2150 Ops/s 0.2122 Ops/s $\color{#35bf28}+1.34\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.6140s 4.4537s 0.2245 Ops/s 0.2223 Ops/s $\color{#35bf28}+1.01\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.1757s 2.0606s 0.4853 Ops/s 0.4997 Ops/s $\color{#d91a1a}-2.89\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7736s 1.6814s 0.5948 Ops/s 0.5942 Ops/s $\color{#35bf28}+0.09\%$
test_values[generalized_advantage_estimate-True-True] 10.5555ms 10.0818ms 99.1889 Ops/s 98.7248 Ops/s $\color{#35bf28}+0.47\%$
test_values[vec_generalized_advantage_estimate-True-True] 21.5334ms 18.0204ms 55.4928 Ops/s 57.0202 Ops/s $\color{#d91a1a}-2.68\%$
test_values[td0_return_estimate-False-False] 0.2519ms 0.1345ms 7.4348 KOps/s 7.7608 KOps/s $\color{#d91a1a}-4.20\%$
test_values[td1_return_estimate-False-False] 28.5676ms 27.7623ms 36.0200 Ops/s 36.2767 Ops/s $\color{#d91a1a}-0.71\%$
test_values[vec_td1_return_estimate-False-False] 22.9865ms 18.0226ms 55.4859 Ops/s 57.0873 Ops/s $\color{#d91a1a}-2.81\%$
test_values[td_lambda_return_estimate-True-False] 44.5895ms 40.8016ms 24.5089 Ops/s 24.6988 Ops/s $\color{#d91a1a}-0.77\%$
test_values[vec_td_lambda_return_estimate-True-False] 20.6330ms 17.8457ms 56.0360 Ops/s 56.4497 Ops/s $\color{#d91a1a}-0.73\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 9.3650ms 8.9265ms 112.0263 Ops/s 109.8210 Ops/s $\color{#35bf28}+2.01\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.9285ms 1.4911ms 670.6375 Ops/s 654.8156 Ops/s $\color{#35bf28}+2.42\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.4756ms 0.4103ms 2.4373 KOps/s 2.4101 KOps/s $\color{#35bf28}+1.13\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 38.7520ms 34.8112ms 28.7264 Ops/s 28.5644 Ops/s $\color{#35bf28}+0.57\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 5.5927ms 1.7327ms 577.1194 Ops/s 585.2374 Ops/s $\color{#d91a1a}-1.39\%$
test_dqn_speed[False-None] 1.8244ms 1.4422ms 693.4086 Ops/s 713.4923 Ops/s $\color{#d91a1a}-2.81\%$
test_dqn_speed[False-backward] 2.0374ms 1.9781ms 505.5297 Ops/s 518.0618 Ops/s $\color{#d91a1a}-2.42\%$
test_dqn_speed[True-None] 0.7513ms 0.5375ms 1.8605 KOps/s 1.8451 KOps/s $\color{#35bf28}+0.83\%$
test_dqn_speed[True-backward] 1.0343ms 0.9979ms 1.0021 KOps/s 1.0123 KOps/s $\color{#d91a1a}-1.01\%$
test_dqn_speed[reduce-overhead-None] 0.9136ms 0.5145ms 1.9437 KOps/s 1.8312 KOps/s $\textbf{\color{#35bf28}+6.14\%}$
test_dqn_speed[reduce-overhead-backward] 1.0328ms 0.9686ms 1.0324 KOps/s 939.3282 Ops/s $\textbf{\color{#35bf28}+9.91\%}$
test_ddpg_speed[False-None] 3.5329ms 2.8780ms 347.4587 Ops/s 344.5627 Ops/s $\color{#35bf28}+0.84\%$
test_ddpg_speed[False-backward] 4.3346ms 4.1322ms 242.0011 Ops/s 240.4173 Ops/s $\color{#35bf28}+0.66\%$
test_ddpg_speed[True-None] 1.7740ms 1.3799ms 724.6995 Ops/s 706.7951 Ops/s $\color{#35bf28}+2.53\%$
test_ddpg_speed[True-backward] 2.4281ms 2.3627ms 423.2363 Ops/s 348.7402 Ops/s $\textbf{\color{#35bf28}+21.36\%}$
test_ddpg_speed[reduce-overhead-None] 1.5338ms 1.3744ms 727.5900 Ops/s 667.5836 Ops/s $\textbf{\color{#35bf28}+8.99\%}$
test_ddpg_speed[reduce-overhead-backward] 2.3935ms 2.3421ms 426.9729 Ops/s 381.3999 Ops/s $\textbf{\color{#35bf28}+11.95\%}$
test_sac_speed[False-None] 9.5204ms 8.0940ms 123.5483 Ops/s 125.0822 Ops/s $\color{#d91a1a}-1.23\%$
test_sac_speed[False-backward] 11.9439ms 11.3556ms 88.0620 Ops/s 88.7238 Ops/s $\color{#d91a1a}-0.75\%$
test_sac_speed[True-None] 2.2873ms 2.1253ms 470.5180 Ops/s 467.5513 Ops/s $\color{#35bf28}+0.63\%$
test_sac_speed[True-backward] 4.0796ms 3.9830ms 251.0689 Ops/s 215.8306 Ops/s $\textbf{\color{#35bf28}+16.33\%}$
test_sac_speed[reduce-overhead-None] 2.3418ms 2.1399ms 467.3221 Ops/s 444.1412 Ops/s $\textbf{\color{#35bf28}+5.22\%}$
test_sac_speed[reduce-overhead-backward] 4.2056ms 4.0337ms 247.9112 Ops/s 249.5078 Ops/s $\color{#d91a1a}-0.64\%$
test_redq_speed[False-None] 14.8477ms 10.6689ms 93.7302 Ops/s 94.8194 Ops/s $\color{#d91a1a}-1.15\%$
test_redq_speed[False-backward] 21.8848ms 18.1098ms 55.2188 Ops/s 54.8818 Ops/s $\color{#35bf28}+0.61\%$
test_redq_speed[True-None] 4.8094ms 4.5118ms 221.6422 Ops/s 231.3719 Ops/s $\color{#d91a1a}-4.21\%$
test_redq_speed[True-backward] 10.6675ms 9.9515ms 100.4871 Ops/s 98.8237 Ops/s $\color{#35bf28}+1.68\%$
test_redq_speed[reduce-overhead-None] 5.8598ms 4.6025ms 217.2740 Ops/s 219.4893 Ops/s $\color{#d91a1a}-1.01\%$
test_redq_speed[reduce-overhead-backward] 10.2801ms 10.0399ms 99.6027 Ops/s 97.4520 Ops/s $\color{#35bf28}+2.21\%$
test_redq_deprec_speed[False-None] 13.9002ms 11.2390ms 88.9757 Ops/s 89.9003 Ops/s $\color{#d91a1a}-1.03\%$
test_redq_deprec_speed[False-backward] 16.7045ms 16.1528ms 61.9088 Ops/s 62.7285 Ops/s $\color{#d91a1a}-1.31\%$
test_redq_deprec_speed[True-None] 3.7747ms 3.6269ms 275.7157 Ops/s 271.0251 Ops/s $\color{#35bf28}+1.73\%$
test_redq_deprec_speed[True-backward] 7.7549ms 7.5964ms 131.6411 Ops/s 114.9408 Ops/s $\textbf{\color{#35bf28}+14.53\%}$
test_redq_deprec_speed[reduce-overhead-None] 3.8790ms 3.6361ms 275.0209 Ops/s 270.3250 Ops/s $\color{#35bf28}+1.74\%$
test_redq_deprec_speed[reduce-overhead-backward] 7.9440ms 7.7025ms 129.8276 Ops/s 127.5289 Ops/s $\color{#35bf28}+1.80\%$
test_td3_speed[False-None] 8.2882ms 8.0596ms 124.0759 Ops/s 116.6778 Ops/s $\textbf{\color{#35bf28}+6.34\%}$
test_td3_speed[False-backward] 11.4806ms 10.9500ms 91.3241 Ops/s 89.8105 Ops/s $\color{#35bf28}+1.69\%$
test_td3_speed[True-None] 1.8878ms 1.8379ms 544.1026 Ops/s 544.5490 Ops/s $\color{#d91a1a}-0.08\%$
test_td3_speed[True-backward] 3.7554ms 3.6253ms 275.8381 Ops/s 246.6023 Ops/s $\textbf{\color{#35bf28}+11.86\%}$
test_td3_speed[reduce-overhead-None] 1.8587ms 1.7974ms 556.3495 Ops/s 546.3489 Ops/s $\color{#35bf28}+1.83\%$
test_td3_speed[reduce-overhead-backward] 3.8336ms 3.6697ms 272.5015 Ops/s 263.8476 Ops/s $\color{#35bf28}+3.28\%$
test_cql_speed[False-None] 28.0738ms 26.0881ms 38.3316 Ops/s 38.0460 Ops/s $\color{#35bf28}+0.75\%$
test_cql_speed[False-backward] 36.3606ms 35.5278ms 28.1470 Ops/s 28.0643 Ops/s $\color{#35bf28}+0.29\%$
test_cql_speed[True-None] 12.9520ms 12.5518ms 79.6697 Ops/s 81.4178 Ops/s $\color{#d91a1a}-2.15\%$
test_cql_speed[True-backward] 18.9955ms 18.5811ms 53.8180 Ops/s 56.4984 Ops/s $\color{#d91a1a}-4.74\%$
test_cql_speed[reduce-overhead-None] 12.7828ms 12.5427ms 79.7274 Ops/s 80.0155 Ops/s $\color{#d91a1a}-0.36\%$
test_cql_speed[reduce-overhead-backward] 19.0122ms 18.5072ms 54.0330 Ops/s 58.3011 Ops/s $\textbf{\color{#d91a1a}-7.32\%}$
test_a2c_speed[False-None] 5.5131ms 5.3082ms 188.3872 Ops/s 180.7258 Ops/s $\color{#35bf28}+4.24\%$
test_a2c_speed[False-backward] 12.2168ms 11.8712ms 84.2378 Ops/s 83.3286 Ops/s $\color{#35bf28}+1.09\%$
test_a2c_speed[True-None] 3.8923ms 3.7391ms 267.4460 Ops/s 261.8626 Ops/s $\color{#35bf28}+2.13\%$
test_a2c_speed[True-backward] 8.8112ms 8.5747ms 116.6224 Ops/s 109.8281 Ops/s $\textbf{\color{#35bf28}+6.19\%}$
test_a2c_speed[reduce-overhead-None] 3.9503ms 3.7404ms 267.3482 Ops/s 262.6625 Ops/s $\color{#35bf28}+1.78\%$
test_a2c_speed[reduce-overhead-backward] 9.2450ms 8.7982ms 113.6600 Ops/s 110.8852 Ops/s $\color{#35bf28}+2.50\%$
test_ppo_speed[False-None] 6.2542ms 5.9493ms 168.0864 Ops/s 169.3768 Ops/s $\color{#d91a1a}-0.76\%$
test_ppo_speed[False-backward] 13.0646ms 12.7247ms 78.5875 Ops/s 80.1075 Ops/s $\color{#d91a1a}-1.90\%$
test_ppo_speed[True-None] 3.8222ms 3.6336ms 275.2058 Ops/s 271.1054 Ops/s $\color{#35bf28}+1.51\%$
test_ppo_speed[True-backward] 8.6740ms 8.4705ms 118.0573 Ops/s 118.4624 Ops/s $\color{#d91a1a}-0.34\%$
test_ppo_speed[reduce-overhead-None] 3.8308ms 3.6377ms 274.8962 Ops/s 271.7934 Ops/s $\color{#35bf28}+1.14\%$
test_ppo_speed[reduce-overhead-backward] 9.0029ms 8.7397ms 114.4207 Ops/s 111.0335 Ops/s $\color{#35bf28}+3.05\%$
test_reinforce_speed[False-None] 6.1017ms 4.5624ms 219.1849 Ops/s 212.2442 Ops/s $\color{#35bf28}+3.27\%$
test_reinforce_speed[False-backward] 7.7858ms 7.4980ms 133.3690 Ops/s 131.7382 Ops/s $\color{#35bf28}+1.24\%$
test_reinforce_speed[True-None] 3.1018ms 2.8914ms 345.8517 Ops/s 339.2625 Ops/s $\color{#35bf28}+1.94\%$
test_reinforce_speed[True-backward] 8.0574ms 7.7989ms 128.2234 Ops/s 128.7690 Ops/s $\color{#d91a1a}-0.42\%$
test_reinforce_speed[reduce-overhead-None] 3.1811ms 2.8984ms 345.0185 Ops/s 334.6862 Ops/s $\color{#35bf28}+3.09\%$
test_reinforce_speed[reduce-overhead-backward] 8.1746ms 7.9647ms 125.5538 Ops/s 112.0844 Ops/s $\textbf{\color{#35bf28}+12.02\%}$
test_iql_speed[False-None] 25.8446ms 20.6186ms 48.4998 Ops/s 49.9707 Ops/s $\color{#d91a1a}-2.94\%$
test_iql_speed[False-backward] 36.8042ms 31.0627ms 32.1930 Ops/s 32.4780 Ops/s $\color{#d91a1a}-0.88\%$
test_iql_speed[True-None] 9.1329ms 8.5981ms 116.3041 Ops/s 113.7882 Ops/s $\color{#35bf28}+2.21\%$
test_iql_speed[True-backward] 17.5117ms 16.8708ms 59.2740 Ops/s 59.9000 Ops/s $\color{#d91a1a}-1.05\%$
test_iql_speed[reduce-overhead-None] 9.2988ms 8.6763ms 115.2558 Ops/s 113.6152 Ops/s $\color{#35bf28}+1.44\%$
test_iql_speed[reduce-overhead-backward] 17.8228ms 17.2761ms 57.8834 Ops/s 58.3069 Ops/s $\color{#d91a1a}-0.73\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.1961ms 6.0401ms 165.5611 Ops/s 165.4593 Ops/s $\color{#35bf28}+0.06\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.8134ms 0.2810ms 3.5592 KOps/s 2.6120 KOps/s $\textbf{\color{#35bf28}+36.26\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6294ms 0.3274ms 3.0542 KOps/s 2.7801 KOps/s $\textbf{\color{#35bf28}+9.86\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0351ms 5.7899ms 172.7142 Ops/s 173.8020 Ops/s $\color{#d91a1a}-0.63\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.4259ms 0.3640ms 2.7470 KOps/s 2.8196 KOps/s $\color{#d91a1a}-2.58\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5691ms 0.3532ms 2.8310 KOps/s 2.9761 KOps/s $\color{#d91a1a}-4.88\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.7361ms 1.4132ms 707.6256 Ops/s 676.0135 Ops/s $\color{#35bf28}+4.68\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.5610ms 1.3376ms 747.5817 Ops/s 710.7775 Ops/s $\textbf{\color{#35bf28}+5.18\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.1856ms 5.9294ms 168.6520 Ops/s 170.2787 Ops/s $\color{#d91a1a}-0.96\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.1703ms 0.5119ms 1.9535 KOps/s 1.8939 KOps/s $\color{#35bf28}+3.14\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6489ms 0.4937ms 2.0254 KOps/s 2.0293 KOps/s $\color{#d91a1a}-0.19\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.9308ms 5.8309ms 171.5000 Ops/s 172.5889 Ops/s $\color{#d91a1a}-0.63\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.7051ms 0.3702ms 2.7013 KOps/s 3.1718 KOps/s $\textbf{\color{#d91a1a}-14.83\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5821ms 0.3617ms 2.7644 KOps/s 3.3824 KOps/s $\textbf{\color{#d91a1a}-18.27\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0621ms 5.7989ms 172.4465 Ops/s 170.4340 Ops/s $\color{#35bf28}+1.18\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.3603ms 0.3665ms 2.7287 KOps/s 3.1722 KOps/s $\textbf{\color{#d91a1a}-13.98\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4926ms 0.3521ms 2.8401 KOps/s 3.3767 KOps/s $\textbf{\color{#d91a1a}-15.89\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.0516ms 5.9519ms 168.0147 Ops/s 166.6004 Ops/s $\color{#35bf28}+0.85\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.1019ms 0.4382ms 2.2819 KOps/s 1.9956 KOps/s $\textbf{\color{#35bf28}+14.35\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6763ms 0.4571ms 2.1876 KOps/s 2.1636 KOps/s $\color{#35bf28}+1.11\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.7049s 19.1313ms 52.2704 Ops/s 195.1917 Ops/s $\textbf{\color{#d91a1a}-73.22\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 9.6398ms 2.1142ms 472.9908 Ops/s 430.0863 Ops/s $\textbf{\color{#35bf28}+9.98\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 7.6799ms 1.2130ms 824.4142 Ops/s 822.3727 Ops/s $\color{#35bf28}+0.25\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 7.4659ms 5.1873ms 192.7796 Ops/s 193.3152 Ops/s $\color{#d91a1a}-0.28\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 9.0985ms 2.0931ms 477.7622 Ops/s 427.7539 Ops/s $\textbf{\color{#35bf28}+11.69\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 7.1002ms 1.2366ms 808.6808 Ops/s 824.3216 Ops/s $\color{#d91a1a}-1.90\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 6.8648ms 5.2696ms 189.7664 Ops/s 49.1156 Ops/s $\textbf{\color{#35bf28}+286.37\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 12.3704ms 2.1528ms 464.5111 Ops/s 465.1799 Ops/s $\color{#d91a1a}-0.14\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 6.4355ms 1.3490ms 741.2784 Ops/s 969.6356 Ops/s $\textbf{\color{#d91a1a}-23.55\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 35.7465ms 33.6755ms 29.6952 Ops/s 29.2844 Ops/s $\color{#35bf28}+1.40\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 19.7510ms 17.8138ms 56.1363 Ops/s 56.6775 Ops/s $\color{#d91a1a}-0.96\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 38.4955ms 35.0772ms 28.5086 Ops/s 28.3619 Ops/s $\color{#35bf28}+0.52\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.4373ms 17.7132ms 56.4551 Ops/s 56.1717 Ops/s $\color{#35bf28}+0.50\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 41.7448ms 36.4706ms 27.4193 Ops/s 26.9812 Ops/s $\color{#35bf28}+1.62\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 21.1193ms 19.4896ms 51.3093 Ops/s 51.2160 Ops/s $\color{#35bf28}+0.18\%$

[ghstack-poisoned]
vmoens added a commit that referenced this pull request Dec 18, 2025
ghstack-source-id: 4f5d4f8
Pull-Request: #3261
@vmoens vmoens merged commit 714b596 into gh/vmoens/170/base Dec 18, 2025
51 of 69 checks passed
@vmoens vmoens deleted the gh/vmoens/170/head branch December 18, 2025 11:14
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